Method for determining an epicenter of dental conditions

Abstract

This disclosure relates to a computer-implemented method comprising the steps of obtaining a three-dimensional (3D) model of a dental site, the 3D model comprising a plurality of data points describing a surface of the dental site in 3D space; estimating a presence of one or more dental conditions defining one or more inspection sites in the 3D model, wherein each inspection site comprises one or more of the plurality of data points; determining an aggregated severity level of the one or more dental conditions for each of the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site; and determining an epicenter of the one or more aggregated severity levels of each inspection site by determining a maximum aggregated severity level of each inspection site and associating the one or more data points of the maximum aggregated severity level with a location of the epicenter.

Description

[0001]METHOD FOR DETERMINING AN EPICENTER OF DENTAL CONDITIONS Technical fieldThe present disclosure relates to the field of dental diagnostics and, in particular, to a computer-implemented method and system for determining an epicenter of one or more dental conditions of a dental site. BackgroundNowadays, most dental health conditions are preventable, and many of them are even treatableif detected in the early stages of their development. The detection of a health condition of adental site of a patient requires an examination of the dental site by a dental practitioner, whichmay be carried out during a clinical visit of the patient in a dental studio. During the clinicalvisit, the dental practitioner may look for possible problem areas of the dental site of the patientto assess the overall health status of the patient’s dental site, the latter including at least thepatient’s dentition and gingiva. Typically, the dental practitioner looks for possible cavities,tooth cracks, tooth wear or other conditions in the patient’s teeth, then they may review existingrestorations, and finally they may proceed with examining the health status of the patient’sgingiva.Additional diagnostic tools may be used by dental practitioners to support and optimize theexamination process. Among these tools, digital dentistry is certainly one of the most widelyused techniques nowadays due to its advantages over non-digital techniques. Digital dentistryenables dental practitioners to perform a scan of the patient’s dental site by means of anintraoral scanner which can acquire scan data relative to the dental site. The scan data may beused to generate a three-dimensional (3D) digital representation of the dental site of the patient,also known as a 3D model of the dental site, which may be input to a number of softwaresolutions, including, but not limited to, patient monitoring systems, and dental software systems configured to estimate a presence of possible problem areas in the patient’s dental site,namely automated diagnostic tools. Said automated diagnostic tools offer solutions to dentalpractitioners to smoothly detect, classify and / or quantify the severity of the conditions presentin the patient’s dental site.A continuous effort in digital dentistry is on improving automated diagnostic tools for a varietyof dental conditions of the patient’s dental site, with the goal of optimizing both the diagnosticprocess and the conveyance of information about the detected dental conditions from the dental practitioner to the patient. Currently existing solutions enable the dental practitioner tovisualize the output of a diagnostic tool for a specific dental condition on a display, e.g. thedisplay of a computer in the dental studio used by the dental practitioner during the clinicalvisit. Typically, a graphical representation of the output is displayed on a rendering of the 3Dmodel of the patient’s dental site. For example, the graphical representation of the output of adiagnostic tool for a specific dental condition may be a color or a heatmap overlayed on therendering of the 3D model of the dental site, or the graphical representation may be a pointerto each detected problem area of the 3D model of the dental site. However, current solutionsonly allow the dental practitioner to visualize the graphical representation of one dentalcondition at a time on the 3D model of the dental site, and accordingly to assess a presenceand / or a severity of one dental condition at a time. If the dental practitioner wants to examinemore than one dental condition, e.g. caries, tooth crack and plaque, they have to manually selectthe dental condition they want to inspect, and manually switch between the visualization of thegraphical representation of different dental conditions on the displayed 3D model of the dentalsite.Current diagnostic tools do not enable visualizing the graphical representation of all thedetected dental conditions on a same rendering of the 3D model of the dental site. Accordingly,existing diagnostic tools suffer from dispersing information relative to the overall health statusof the dental site, and dental practitioners do not have any tool to objectively determine thelocations of the dental site which are affected by more dental conditions. Furthermore,manually switching between the graphical representation of different dental conditions maycause the dental practitioner to miss problem areas of the dental site. Another drawback ofexisting diagnostic tools is that it may be challenging for the dental practitioner to convey tothe patient the information that one or more locations of their dental site necessitate a treatmentplan due to their health status, as switching between the graphical representation of differentdental conditions does not convey to the patient a clear and effective visualization of the overallhealth status of their dental site.Given the importance of an early detection and / or treatment of any existing dental condition inthe patient’s dental site, it is crucial to develop a diagnostic tool enabling dental practitionersto easily and efficiently assess the overall health status of the patient’s dental site at once. Adiagnostic tool providing dental practitioners with an effective guide to carry out theexamination of the patient’s dental site may also be time-saving both for practitioners and forpatients. In addition to this, there is a need to develop improved diagnostic tools to optimizethe communication between the dental practitioner and the patient. Summary It is one aspect of the present disclosure to provide a computer-implemented method thatovercomes the above-mentioned disadvantages. Accordingly, in one aspect there is disclosedherein a computer-implemented method comprising the steps of:^ obtaining a 3D model of a dental site, the 3D model comprising a plurality of data pointscollectively describing a surface of the dental site in 3D space; ^estimating a presence of one or more dental conditions defining one or more inspectionsites in the 3D model, wherein each inspection site comprises one or more of the plurality of data points; ^determining an aggregated severity level of the one or more dental conditions for eachof the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site; and ^determining an epicenter of the one or more aggregated severity levels of eachinspection site by determining a maximum aggregated severity level of the inspection site and associating the one or more data points of the maximum aggregated severity level with a location of the epicenter.Advantageously, the epicenter comprises quantified information relative to the aggregatedseverity of the one or more dental conditions, and it therefore localizes one or more specificareas of the dental site for which the overall health status is determined to be, locally orglobally, the lowest, i.e. for which the health status is the worst. In other words, the epicenterlocalizes one or more specific areas of the dental site which necessitate the highest attention inan examination process carried out by a dental practitioner. Consequently, one or morelocations of the dental site which necessitate careful inspection by the dental practitioner areaccurately determined at once. The method disclosed herein significantly optimizes the diagnostic efficiency, as it may provide the dental practitioner with a clear and effective guide to perform an examination of the dental site. Said examination may be started from theepicenter of each inspection site, thereby enabling to prevent the development of further dentalconditions in / around the epicenter and enabling to draw up a treatment plan for dentalconditions at an advanced stage of their development.The computer-implemented method may comprise obtaining a 3D model of a dental site. The3D model may be a digital representation of a 3D geometry of the dental site, wherein the 3Dgeometry may describe intraoral features of the patient, including at least teeth and gingiva,but it may also comprise hard palate, soft palate and so forth.The 3D model of the dental site may comprise a plurality of data points collectively describinga surface of the dental site in 3D space. The plurality of data points may be elements that theabove-mentioned 3D geometry consists of, namely the elements that can describe a surface of the dental site in 3D space. In one example, the plurality of data points may be the plurality of3D points of a point cloud, wherein each point in the point cloud is described by a set ofCartesian coordinates (x, y, z). In another example, the plurality of data points may be aplurality of facets of a mesh parametrizing the surface of the dental site in 3D space. Each ofsaid facets may be a triangle in the case where the mesh is a triangle mesh, or another polygonin the case where the mesh is a polygonal mesh, or any other geometrical shape suitable for representing a 3D model. In yet another example, the plurality of data points may be a pluralityof vertices comprised in a mesh, wherein the mesh may be a triangle mesh or any otherpolygonal mesh suitable for describing the surface of the dental site in 3D space. As the computer-implemented method disclosed herein may be based on using a trained learning model, it is an advantage that the 3D model comprises the plurality of data points, asthe plurality of data points enables the 3D model to be processed using the trained learningmodel. Therefore, the plurality of data points ensures that a discretized input which is suitablefor the trained learning model is provided to said training model. Advantageously, this allowsefficient processing of the obtained 3D model of the dental site. Further explanation of the usage of a trained learning model will be elaborated on throughout the description. The computer-implemented method may further comprise estimating a presence of one or more dental conditions defining one or more inspection sites in the 3D model. Each inspection site may comprise one or more of the plurality of data points. Estimating a presence of one or moredental conditions may comprise processing the plurality of data points using a trained learningmodel to estimate the presence of the one or more dental conditions. The trained learning modelmay be a pre-trained machine learning model for each of the one or more dental conditions, orit may be a pre-trained machined learning model for all of the one or more dental conditions.The pre-trained machine learning model may have been trained on training data comprisingdata points similar to the plurality of data points, the training data further comprising quantifiedlocal information about features of each data point in the training data.Advantageously, the processing of the plurality of data points provides local informationrelative to the one or more dental conditions for the 3D model, enabling to accurately localizeone or more regions of the 3D model for which the presence of at least one dental condition isestimated, i.e. the one or more inspection sites of the 3D model. A further advantage ofdetermining local information relative to the one or more dental conditions is that only the one or more data points comprised in each inspection site are further processed, thereby reducingthe computational time of the computer-implemented method.The computer-implemented method may further comprise determining an aggregated severitylevel of the one or more dental conditions for each of the one or more of the plurality of datapoints, thereby determining one or more aggregated severity levels for each inspection site.The aggregated severity level of the one or more dental conditions may be a sum or a weightedsum of severity levels, wherein each severity level may be determined for a specific dentalcondition. For each data point, the severity level of each dental condition may be a scorequantifying the severity of that specific dental condition at the location of that data point andsaid score may be defined in a suitable manner based on clinical aspects of the dental condition.Advantageously, determining the aggregated severity level for each data point in eachinspection site ensures that an overall dental health condition is locally quantified. Further,determining the aggregated severity level for each data point in each inspection site allows to determine one or more epicenters of the dental conditions in the 3D model, i.e. the one or more regions of the dental site which necessitate the highest attention by the dental practitioner. Thisis not possible with current diagnostic tools, which typically provide a quantification of theseverity of a dental condition at a time and do not enable to quantify the overall health statusof the dental site.The computer-implemented method may further comprise determining an epicenter of the oneor more dental conditions of each inspection site by determining a maximum aggregatedseverity level of each inspection site and associating the one or more data points of themaximum aggregated severity level with a location of the epicenter. Thus, one or moreepicenters of the dental conditions may be determined for the 3D model of the dental site. Inan alternative embodiment, the method may comprise determining only one epicenter of the dental conditions by determining a global maximum aggregated severity level among theaggregated severity levels determined for all of the inspection sites of the 3D model. In otherwords, the method disclosed herein may comprise determining at least one epicenter of the dental conditions. The one or more aggregated severity levels determined for each inspectionsite may vary over a broad range, for example in the case where the inspection site comprisesmany data points. This may result in the dispersion of the information relative to the overallhealth status of the dental site. It is therefore an advantage that one or more epicenters aredetermined, as this accurately localizes one or more areas of the 3D model of the dental site forwhich the overall health status is determined to be, locally or globally, the lowest / worst. Inother words, determining at least one epicenter enables to localize one or more areas of thedental site which necessitates the highest attention by the dental practitioner. Advantageously,the determined one or more epicenters provide the dental practitioner with an effective guideto perform an examination of the dental site of the patient, as the examination may be startedby the locations associated with the one or more data points of each epicenter where thechallenges associated with detected dental conditions are the highest.The computer-implemented method may further comprise associating the determined one or more aggregated severity levels with the plurality of data points of the 3D model and rendering the 3D model on a graphical user interface (GUI) with a graphical representation of the one or more aggregated severity levels. Each determined aggregated severity level maps to one datapoint of the 3D model, such that each data point of the 3D model is associated with quantifiedlocal information relative to the severity of the one or more dental condition. This severitymapping enables to generate a graphical representation of the one or more aggregated severitylevels. Advantageously, the graphical representation enables the dental practitioner to identifyat once the one or more locations of the dental site which necessitate careful inspection due totheir overall health condition.The graphical representation of the one or more aggregated severity levels may be displayedas an indication on the 3D model. The indication may be intended as a texture, a pointer, asymbol, or the like indicating the areas of the dental site affected by the one or more dentalconditions. Advantageously, the indication may be such that a contrast is generated betweenthe rendered 3D model and the one or more aggregated severity levels, thereby highlighting ina clear way the locations of the 3D model for which the presence of at least one dental conditionis estimated. This ensures that the dental practitioner is assisted and guided in performing theexamination process of the dental site of the patient, whereby the time needed to perform said examination is significantly reduced with respect to the case where the dental practitionerperforms the in-mouth examination without such a graphical representation. In practice, thisfast examination could have not been achieved if the dental practitioner was not provided with the indication of the areas of the dental site which are highly affected by the one or more dental conditions, i.e. with the indication of the inspection sites and the corresponding epicenter(s).The indication may be a color different than a rendering color of the 3D model, and an intensityof the color may decrease with a distance from the estimated epicenter of each inspection site.Therefore, the larger the intensity of the color in a region of the rendering of the 3D model, thelarger the severity of the one or more dental conditions determined in that region. This is an advantage, as the dental practitioner may start the examination of the patient’s dental site byinspecting those locations of the dental site which are displayed with the largest color intensityon the 3D model. Said locations may be the ones where a dental condition at an advanced stageof its development is present, or they may be the ones where more dental conditions are present.In both cases, the intensity of the color enables to draw the attention of the dental practitionerto the locations of the dental site which may necessitate a treatment plan and / or a plan toprevent the further development of the one or more dental conditions. Then, the dentalpractitioner may proceed with the in-mouth examination by moving to areas of the dental site which are displayed with progressively decreasing intensity. In other words, the intensity of thecolor generates a hierarchy of a clinical significance of one or more regions of the 3D model,thereby providing the dental practitioner with a guide to perform the examination processwhich makes the diagnostic process quick and effective.The indication may be a pattern, and a density of the pattern may decrease with a distance fromthe estimated epicenter of each inspection site. Therefore, the larger the density of the pattern in a region of the rendering of the 3D model, the more severe the determined overall health status in that region. This is an advantage, as the dental practitioner may start the examination of the patient’s dental site by inspecting the locations of the dental site which are displayed with the highest pattern density on the 3D model. Said locations may be the ones where a dental condition at an advanced stage of its development is present, or they may be the ones where more dental conditions are present. In both cases, the density of the pattern enables to draw the attention of the dental practitioner to the locations of the dental site which may necessitate a treatment plan and / or a plan to prevent the further development of the one or more dental conditions. Then, the dental practitioner may proceed with the in-mouth examination by moving to areas of the dental site which are displayed with progressively decreasing density. In other words, the density of the pattern generates a hierarchy of a clinical significance of one or more regions of the 3D model, thereby providing the dental practitioner with a guide which makes the diagnostic process quick and effective.The method may further comprise calculating the distance using a Euclidean distance measure.The Euclidean distance measure is advantageous when the plurality of data points comprises apoint cloud, as each point of the point cloud is described by Cartesian coordinates (x,y,z) in Euclidean space. The method may further comprise calculating the distance using a geodesic distance measure. The geodesic distance measure is advantageous when the plurality of data points comprises atriangle mesh comprising facets and vertices, as the facets and vertices of the mesh compriseinformation about a local curvature of the surface of the dental site. The method may further comprise calculating the distance using a facet distance measure. The facet distance measure may be intended as a measure of the minimum number of steps to gofrom a facet to another facet only moving through neighboring facets. Therefore, the facetdistance measure is advantageous when the plurality of data points comprises a plurality of facets comprised in a mesh describing the surface of the dental site in 3D space. The one or more dental conditions may comprise one or more of caries, gum recession, tooth wear, plaque, gum inflammation, and tooth crack. Those are among the most common dentalhealth conditions which can be detected at early stages of their development. However, the oneor more dental conditions may comprise any other dental condition which is not mentionedabove and which may be of interest in the examination process of the patient’s dental site.Estimating the presence of the one or more dental conditions may comprise processing the plurality of data points using a trained machine learning model for each of the one or more dental conditions. Advantageously, the trained machine learning model performs a localprocessing of the 3D model, and therefore outputs a local estimation of the presence of the oneor more dental conditions in the 3D model. This improves the accuracy of the localization ofone or more areas of the 3D model where one or more dental conditions are present. The trained machine learning model for each dental condition may output for each data point a severity level of each dental condition, and the severity level may be associated with a number on a predefined scale of severity of each dental condition. This ensures that the severity of eachdental condition is locally quantified in the 3D model. The predefined scale of severity of each dental condition may comprise a minimum numberwhich is indicative of an absent dental condition. Advantageously, one or more data points ofthe 3D model for which the minimum number on the scale of severity is determined are notfurther processed, reducing the computational complexity and computational execution timeof the method disclosed herein. In other words, the method comprises processing the datapoints which are associated with at least a number which is larger than the minimum number.The minimum number may be zero. Advantageously, as some steps of the method may involve mathematical manipulation of the severity levels, the number zero ensures that no contribution is given by the minimum severity level in said mathematical steps, e.g. in steps involving sumsor multiplications such as the calculation of the aggregated severity level for each data point.Furthermore, from an imaging perspective, the number zero corresponds to an absence of a color and / or pattern, e.g. in a red, green, blue (RGB) scale, or in a cyan, magenta, yellow,key / black (CMYK) scale, or in a dots per inch (DPI) scale and so forth. Hence, it is anadvantage that when the number zero is associated with a data point, that data point is displayedon the 3D model with the same color of the rendering color of the dental site. Determining the aggregated severity level of the one or more dental conditions for each data point comprises determining the severity level of each dental condition for each data point, and for each data point aggregating the determined severity level of each dental condition.Advantageously, the overall health condition is locally determined in the 3D model, i.e. foreach data point in each inspection site, based on a quantified severity of each dental conditiondetermined for that data point. This significantly improves the accuracy of the localization ofthe regions of the 3D model which are affected by the one or more dental conditions. Aggregating the determined severity level of each dental condition may comprise summing thedetermined severity level of each dental condition. Advantageously, summing the determinedseverity level of each dental condition may reduce the computational cost and time of themethod. Furthermore, as the sum assigns equal significance to each dental condition it preventsany dental condition from being overlooked, providing the dental practitioner with an objective and unbiased estimation of the presence of the one or more dental conditions in the dental site. Aggregating the determined severity level of each dental condition may comprise determining a weighted sum of the determined severity level of each dental condition. Advantageously, the weighted sum ensures that each dental condition properly contributes to the quantification of the overall health status of each data point, e.g. based on realistic clinical features of that dental condition. Determining the weighted sum may further comprise assigning a weight to the determined severity level of each dental condition based on a clinical significance of each dental condition.This ensures that dental conditions which are considered more important from a clinicalperspective, e.g. because their stage of development advances quickly and / or because they maycause a development of other dental conditions, contribute more to the aggregated severity level than dental conditions which are instead categorized as less important with respect to aclinical standard. Advantageously, the weighted sum enables to determine a realistic overallhealth status of the dental site of the patient.The plurality of data points may comprise a point cloud. The 3D model of the dental site maybe generated upon acquiring data relative to the dental site by means of a 3D intraoral scanner,which may capture and / or generate a point cloud representing the surface of the dental site. It is therefore an advantage that the computer-implemented method disclosed herein is configured to be performed upon obtaining a 3D model comprising a point cloud. The plurality of data points may comprise a plurality of facets of a triangle mesh. The 3D model of the dental site may be generated upon acquiring data relative to the dental site by means ofa 3D intraoral scanner which may process the acquired data to generate a triangle meshrepresenting the surface of the dental site in 3D. Said triangle mesh may comprise a plurality of facets, and it is therefore an advantage that the computer-implemented method disclosedherein is suitable to be performed upon obtaining a 3D model comprising a plurality of facets.The plurality of data points may comprise a plurality of vertices of a triangle mesh. The 3D model of the dental site may be generated upon acquiring data relative to the dental site bymeans of a 3D intraoral scanner, which may process the acquired data relative to generate atriangle mesh representing the surface of the dental site in 3D. Said triangle mesh may comprisea plurality of vertices, and it is therefore an advantage that the computer-implemented method disclosed herein is suitable to be performed by obtaining a 3D model comprising a plurality of vertices. The computer-implemented method may further comprise smoothing the one or more aggregated severity levels determined for each inspection site, whereby one or more smoothedaggregated severity levels may be determined for each inspection site. The aggregated severitylevel may change abruptly when moving from a data point to a neighboring data point in thesame inspection site or when moving between adjacent inspection sites. These inevitablediscontinuities in each inspection site may cause the generation of an unclear graphicalrepresentation of the one or more aggregated severity levels, e.g. the boundaries of saidlocations may appear as edgy. Advantageously, the smoothing step ensures that the aggregatedseverity level smoothly changes when moving between neighboring data points, improving thegraphical representation of the locations of the 3D model for which at least one dental conditionis determined and facilitating the diagnostic process for the dental practitioner.The one or more smoothed aggregated severity levels determined for each inspection site may linearly or exponentially decrease with the distance from the epicenter determined for eachinspection site. The linear behavior of the smoothed aggregated severity levels ensures that theaggregated severity levels of each inspection site decrease at a constant rate when moving away from the epicenter of that inspection site. This may be advantageous to provide the dental practitioner with an indication of all the data points for which at least one dental condition is determined, rather than enhancing the epicenter of each inspection site. On the other hand, the exponential behavior ensures that the one or more aggregated severity levels of each inspectionsite rapidly fall off when moving away from the epicenter of each inspection site. This may beadvantageous to provide the dental practitioner with an indication of the regions of the dental site which are affected by more severe dental conditions than others, i.e. it ensures that the epicenter of each inspection site stands out in the graphical representation of the one or more aggregated severity levels.The computer-implemented method may comprise determining a global epicenter of the oneor more dental conditions of the 3D model by determining a global maximum aggregatedseverity level among the one or more aggregated severity levels determined for the one or moreinspection sites, and associating the one or more data points of the global maximum aggregatedseverity level with a location of the global epicenter. Advantageously, a region of the 3D modelfor which the overall health status is determined to be, globally, the lowest, i.e. the worst, isaccurately determined. In other words, the global epicenter is the region of the dental site whichnecessitates, globally, the highest attention by the dental practitioner. Thus, the dentalpractitioner may promptly identify the location of the 3D model which necessitates particularattention and start the dental examination from the location of the global epicenter.The computer-implemented method may comprise normalizing the one or more aggregatedseverity levels of the one or more inspection sites with the determined global maximum aggregated severity level, whereby the normalized one or more aggregated severity levels of the one or more inspection sites may vary between 1 and 0, wherein 1 is the normalizedaggregated severity level of the global epicenter of the 3D model. Thus, the normalization maygenerate an order of clinical significance of the epicenters determined for the one or moreinspection sites of the 3D model, based on the aggregated severity of the epicenter of eachinspection site relative to the aggregated severity of the global epicenter. Advantageously, this provides a clear and effective diagnostic guide to the dental practitioner, who accordingly may start the examination of the dental site from the location of the global epicenter and proceed with locations for which a progressively decreasing aggregated severity level of the one ormore dental conditions is determined.The method may further comprise generating an ordered list of the one or more dentalconditions estimated for each inspection site and displaying the ordered list on the GUI uponreceiving an input signal. Therefore, the dental practitioner may have an accurate indication ofthe one or more dental conditions estimated for each inspection site and perform an in-mouth examination of the one or more inspection sites according to the ordered list. The method may further comprise determining an order of importance of the severity level of each of the one or more dental conditions estimated for each inspection site, wherein the ordermay be a descending order. Advantageously, the dental practitioner may perform the in-mouthexamination of the inspection sites of the patient’s dental site using as a guide the order ofimportance of the severity level determined for each dental conditions. For example, the dentalpractitioner may start inspecting the dental condition which is determined to be the most severein the generated ordered list and proceed with progressively less severe dental conditions. Thus,using the 3D model combined severity mapping guidance significantly facilitates theexamination of the patient’s dental site.The method may further comprise storing the ordered list generated for each inspection siteand loading the ordered list upon receiving an input signal. Advantageously, the stored orderedlist can be accessed at any time by the dental practitioner by inputting a signal, for instance to review the severity levels determined for each inspection site in a previous examination of thepatient’s dental site and compare them with the severity levels determined for the sameinspection site during the clinical visit of the patient.The ordered list may comprise the severity level of each of the one or more dental conditionsestimated for each inspection site. This provides the dental practitioner with a clearquantification of the severity of each dental condition locally determined for each inspectionsite. The input signal may comprise hovering or clicking on an inspection site on the displayed 3Dmodel with the graphical representation of the one or more aggregated severity levels. Thisprovides a simple and practical operation for the dental practitioner to load and visualize theordered list for each inspection site. In another aspect, there is disclosed herein a program product comprising instructions which,when executed by a computer, cause the computer to perform the method steps describedherein. Thus, in practice the method disclosed herein can be executed by any computer system configured to read the computer program product disclosed herein. This is an advantage, as the instructions comprised in said computer program ensure that the execution of one or more stepsof the method disclosed herein is automated.In yet another aspect, there is disclosed herein a non-volatile computer-readable medium comprising instructions which, when executed by a computer, cause the computer to perform the method according to any of the disclosed embodiments. Thus, in practice the instructions to perform the method disclosed herein can be performed on any computer system configured to read the non-volatile computer-readable medium disclosed herein. This is an advantage, as said instructions can be retained also when the computer system is turned off, enabling the user to access the instructions and / or the data stored on the non-volatile computer-readable medium at any time and from any computer system configured to read the non-volatile computer- readable medium disclosed herein. In yet another aspect, there is disclosed herein a system which may comprise an intraoral scanner. The system may further comprise a computer system. Said computer system may comprise a display, a communication interface, one or more processors, and one or more memories. The one or more memories may contain a program content executable by the one ormore processors. The program content may further comprise executable instructions to obtaina 3D model of a dental site, wherein the 3D model may comprise a plurality of data points collectively describing a surface of the dental site in 3D space. The program content may further comprise executable instructions to estimate a presence of one or more dental conditions defining one or more inspection sites in the 3D model, wherein each inspection site comprisesone or more of the plurality of data points. The program content may further compriseexecutable instructions to determine an aggregated severity level of the one or more dental conditions for each of the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site. The program content may further comprise executable instructions to determine an epicenter of the one or more aggregated severity levels of each inspection site by determining a maximum aggregated severity level ofthe inspection site and to associate the one or more data points of the maximum aggregatedseverity level with a location of the epicenter. Advantageously, the intraoral scanner enables the dental practitioner to perform a scan of the dental site of a patient, and therefore to acquire scan data relative to said dental site. The data acquired by the intraoral scanner may be used to generate the 3D model of the dental site. It isa further advantage that the system disclosed herein enables to determine the location of theepicenter of each inspection site, as the latter localizes a specific area of that inspection site forwhich the overall health status is determined to be the most severe. Consequently, one or morelocations of the dental site which necessitate careful inspection by a dental practitioner areaccurately determined at once. The system disclosed herein significantly optimizes thediagnostic efficiency, as it provides the dental practitioner with a clear and effective guide toperform an examination of the dental site. Said examination may be started from the epicenter of each inspection site, thereby enabling to prevent the development of further dentalconditions in / around the epicenter and enabling to draw up a treatment plan for dentalconditions at an advanced stage of their development. It is yet another advantage of the system disclosed herein that the computer system comprises the display, as the 3D model with agraphical representation of the one or more aggregated severity levels determined for eachinspection site may be displayed on the display. The dental practitioner may use the displayed graphical representation as a guide and / or support to perform the in-mouth examination of the dental site and assess its overall health status. Thus, the system disclosed herein significantly optimizes the diagnostic process. The communication interface may be configured to enable the computer system to exchangedata with one or more external devices and with one or more external networks. Thus, thecommunication interface enables the computer system to send and / or receive data from the intraoral scanner comprised in the system disclosed herein. This is an advantage, as thecomputer system may receive from the intraoral scanner scan data relative to the patient’sdental site acquired by the intraoral scanner. Furthermore, the communication interface enables the computer system to receive and / or send data to an external network which may be configured to perform at least a part of the processing of the 3D model, thereby facilitating andspeeding up the processing of the 3D model.The one or more external networks may comprise at least one cloud network. Thus, a part ofthe processing of the 3D model may be performed by the at least one cloud network.Advantageously, the cloud network may be configured to handle heavy workloads efficiently,thereby reducing the computational load of the computer system and optimizing the processingof the 3D model. The at least one cloud network may comprise executable instructions to process the plurality of data points using a trained machined learning model for each of the one or more dentalconditions. Thus, the computer system may send the obtained 3D model of the dental site tothe cloud network, and receive the output of the trained machine learning model of the cloud network by means of the communication interface, without the need of having said machine learning model stored on the one or more memories comprised in the computer system. Thissignificantly saves memory space of the computer system, and it also speeds up the processingof the plurality of data points.In yet another aspect, there is disclosed herein is a computer-implemented method comprisingthe steps of: -obtaining a first three-dimensional (3D) model of a dental site comprising a first set ofdata points, wherein the first set of data points describes a surface of the dental site at a first time; -obtaining a second 3D model of the dental site comprising a second set of data points,wherein the second set of data points describes the surface of the dental site at a second time; -comparing the first 3D model with the second 3D model to determine one or moredifferences therebetween; -determining one or more heat scores quantifying one or more clinical changes in thedental site between the first time and the second time by determining a heat score for each of the one or more differences; and -determining an epicenter of the one or more heat scores for each of the one or moredifferences by determining a maximum heat score for the corresponding difference. Therefore, two 3D models of the dental sites obtained at different times are compared to determine any clinical change which occurred in the dental site between the first time and the second time. For example, the first time may be representative of a time when the patient went to the dental clinic for a first clinical examination, and the second time may be representative of a time when the patient went to the dental clinic for a second clinical examination. Clinical changes may comprise changes in the geometry of one or more parts of the dental site changesin the color of one or more parts of the dental site, the appearance / disappearance of one or moredental conditions which were not detected at the first time, changes in the severity of one or more dental conditions with respect to the severity of the corresponding dental condition determined at the first time, and so forth. Each of the one or more differences may correspond to a particular clinical change among the clinical changes listed above, and among any other clinical change which may be determined upon comparing the 3D models of the dental site obtained at two different times. Advantageously, the method disclosed herein enables thedetermination of quantified information about the one or more determined differences betweenthe first and second 3D models by determining the one or more heat scores. A further advantage of the present disclosure is that the epicenter of the one or more heat scores for each of the oneor more differences is determined. The epicenter of each difference is indicative of the part orparts of the dental site in which the largest difference between the first time and the secondtime is determined, i.e. the part(s) of the dental site which may benefit from a careful in-mouthexamination by the dental practitioner. Accordingly, the dental practitioner is provided with anobjective and quantitative indication of the area or areas of the dental site in which clinicallyrelevant changes have occurred since the previous clinical examination of the patient. The output of the method of the present disclosure may therefore guide the dental practitioner toperform the in-mouth examination of the dental site. In other words, the dental practitioner maystart to inspect the areas corresponding to the epicenter determined for each of the one or moredifferences and proceed by examining areas with decreasingly significant clinical changes.The method may further comprise obtaining first augmentation data indicating an estimatedpresence of one or more dental conditions at a first set of inspection sites in the first 3D model.Advantageously, the first augmentation data provides localized information about the one ormore dental conditions which are detected in the dental site at the first time. Said localizedinformation may be used to track the progress and / or regression of each of the one or more dental conditions. The method may further comprise obtaining second augmentation data indicating an estimated presence of the one or more dental conditions at a second set of inspection sites in the second3D model. Advantageously, the second augmentation data provides localized information about the one or more dental conditions which are detected in the dental site at the second time. Said localized information may be used to track the progress and / or regression of each of the one or more dental conditions. The second time may be subsequent to the first time. Therefore, the method disclosed herein enables dental practitioners to compare the 3D models of a patient’s dental site obtained at twodifferent clinical examinations of the patient.The first augmentation data and the second augmentation data may further comprise segmented tooth surface data and segmented gingiva surface data. Advantageously, the segmented datamay be used to determine a spatial relationship in 3D space between one or more parts of thefirst 3D model and one or more parts of the second 3D model, e.g. parts in the first 3D modeldescribing a given tooth in 3D space and parts of the second 3D model describing the sametooth in 3D space.The comparing step may comprise determining a spatial relationship between at least a part ofthe first 3D model and at least a part of the second 3D model. Determining the spatial relationship ensures that only corresponding parts of the 3D models are compared with each other, therefore ensuring that an accurate comparison between the 3D models obtained at two different times is performed according to the method disclosed herein. Determining the spatial relationship may comprise using the first augmentation data and using the second augmentation data to determine that the at least a part of the first 3D model and theat least a part of the second 3D model represent a same part of the dental site. The firstaugmentation data and the second augmentation data may comprise segmented tooth surfacedata and segmented gingiva data obtained at different times. Furthermore, a tooth number orteeth numbers according to the Universal Tooth Numbering System may be comprised both inthe segmented first data and in the segmented second data. Thus, the segmented surface data may be used to determine part(s) of the first data and of the second data representing the sameportion of the dental site, i.e. describing a same tooth, a same group of teeth, a same quadrantand so forth. This ensures that the changes between the first 3D model and the second 3D modelare determined by comparing spatially corresponding parts of the 3D models.At least a part of the first augmentation data and / or at least a part of the second augmentation data may be obtained upon processing the first set of data points and / or the second set of data points by using a trained machine learning model for each of the one or more dental conditions. The trained learning model may be a pre-trained machine learning model for each of the one or more dental conditions, or it may be a pre-trained machined learning model for all of the one or more dental conditions. The pre-trained machine learning model may have been trained on training data comprising data points similar to the first set of data points and the second set of data points, the training data further comprising quantified local information about features of each data point in the training data. For example, the features may comprise a presence of one or more dental conditions at the corresponding data point, and / or a severity level of the one or more dental conditions at the data point.Therefore, the first and / or second augmentation data provide local information relative to theone or more dental conditions for the first and / or second 3D model, respectively, enabling toaccurately localize one or more regions of the first and / or second 3D model for which thepresence of at least one dental condition is estimated, i.e. one or more inspection sites of the first / second 3D model. Advantageously, this local information may be used to determine an improvement or a worsening in the one or more dental condition between the first time and the second time, i.e. by comparing the first augmentation data with the second augmentation data.The trained machine learning model for each dental condition may output for each data pointin the first set of data points and / or for each data point in the second set of data points a severitylevel of each dental condition. The severity level provides quantified information about theclinical severity of each dental condition determined for each data point in the first set of datapoints and / or for each data point in the second set of data points. Advantageously, the quantifiedinformation is particularly suitable to compare the severity level determined for each dentalcondition at two different times, e.g. by calculating the difference between the severity leveldetermined at the first time and at the second time for the same dental condition. In this way,the dental practitioner may be provided with a quantified difference between the severity levels determined at two different times, which may in turn enable the dental practitioner to objectively determine whether a dental condition improved or worsened since the last clinical examination.Each inspection site in the first set of inspection sites may comprise one or more data pointswithin the first set of data points and each inspection site in the second set of inspection sitesmay comprise one or more data points within the second set of data points. Therefore, comparing the first augmentation data with the second augmentation data may comprise comparing each of the one or more data points within the first set of data points with thecorresponding data point within the second set of data points. Advantageously, this ensures that locally quantified information about a given data point in an inspection site of the first set of inspection sites may be compared with locally quantified information about a corresponding data point in an inspection site of the second set of inspection sites. The first augmentation data may further comprise one or more first severity levels of one or more dental conditions determined for each inspection site in the first set of inspection sites,and the second augmentation data may further comprise one or more second severity levels ofthe one or more dental conditions determined for each inspection site in the second set ofinspection sites. Each of the one or more first severity levels may be a severity level determinedfor a data point comprised in the first set of inspection sites. Similarly, each of the one or more second severity levels may be a severity level determined for a data point comprised in the second set of inspection sites. Advantageously, this ensures that the difference between the severity levels determined for corresponding data points in the first and second set of inspectionsites may be locally quantified. In turn, this provides the dental practitioner with unambiguousand objective information about the improvement or the worsening of the one or more dental conditions in specific regions of the dental site. Comparing the first 3D model with the second 3D model may comprise determining a presence of at least a part of the dental site in the second 3D model which is not present in the first 3Dmodel. This effectively highlights new parts of the dental site which were not present in theprevious clinical examination, such as one or more dental crowns, onlays, inlays, and so forth. The at least a part of the dental site may comprise a dental prosthesis. This enables the dental practitioner to quickly identify the region of the dental site where a dental prosthesis wasapplied between the first time and the second time, e.g. to check in-mouth the condition of thedental prosthesis. In some embodiments, the method may also enable to determine a difference between the appearance of the dental prosthesis at the first time and at the second time, forexample to determine whether the dental prosthesis got damaged.The at least a part of the dental site may comprise a dental filling. This enables the dentalpractitioner to quickly identify the region of the dental site where a filling was applied betweenthe first time and the second time, e.g. to perform an in-mouth check of the condition of thefilling. In some embodiments, the method may also enable to determine a difference betweenthe appearance of the filling at the first time and at the second time, e.g. to determine whetherthe filling got damaged.Comparing the first 3D model with the second 3D model may comprise determining one ormore differences between second color data comprised in the second set of data points and firstcolor data comprised in the first set of data points. The color of the teeth may change for severalreasons, including tooth decay, poor oral hygiene, patient’s habits such as tobacco use and so forth. Furthermore, the color of the gingiva may change due to dental pathologies such as gingivitis, or due to bacterial infections. It is therefore an advantage that the one or more differences between the first color data and the second color data are determined, as this may enable the dental practitioner to perform a more detailed in-mouth examination to determinethe cause of such changes and to define a treatment plan for the patient, and / or to provide thepatient with some recommendations such as changes in diet or oral hygiene.Determining the one or more differences between the second color data and the first color data may comprise: -computing one or more difference values, each difference value quantifying adifference between a color coordinate of a data point in the second det of data points and a color coordinate of a corresponding data point in the first set of data points; -comparing each of the one or more difference values with a threshold value, whereinresponsive to determining that a difference value among the one or more difference values is equal or greater that a threshold value the difference value is classified as clinically relevant, and wherein responsive to determining that a difference value among the one or more difference values is smaller than the threshold value the difference value is classified as not clinically relevant. Therefore, only the difference values among the one or more difference values which fulfill the threshold criterion are classified as being clinically significant, i.e. of interest from a clinicalperspective. Clinically significant differences may be associated, for example, with a rate ofchange in the color of the patient’s tooth or teeth which is above a threshold defined by thestandards of the American Dental Association, or by the standards of the European Centre forDisease Prevention and Control, or by any other global or national health organization. Advantageously, this provides the dental practitioner with an automated method to identify regions of the dental site of the patient which need special attention as they do not satisfy dental health standards. The method may further comprise comparing the first augmentation data with the second augmentation data to determine the one or more differences therebetween. Advantageously, comparing the localized information obtained at the second time with the localized information obtained at the first time may enable the dental practitioner to track a regression of each of theone or more dental conditions, and / or to detect a worsening of the one or more dentalconditions. In other words, the comparison between the first augmentation data and the second augmentation data may assist the dental practitioner in tracking the progress of an ongoing treatment plan for the patient, and consequently in determining whether the treatment plan needs to be updated, and / or whether a new treatment plan is needed.Comparing the first augmentation data with the second augmentation data may comprisedetermining one or more preexisting inspection sites by determining one or more inspectionsites in the second set of inspection sites for which at least one corresponding inspection sitein the first set of inspection sites is identified. In other words, a preexisting inspection site maybe an inspection site in the second set of inspection sites for which a spatial relationship with an inspection site in the first set of inspection sites can be determined, i.e. an inspection site in the second set of inspection sites which at least partially overlap in 3D space with an inspection site in the second set of inspection sites. Therefore, the one or more preexisting inspection sites correspond to areas of the dental site in which the presence of at least one dental condition was determined at the first time, and in which the presence of at least one dental condition isdetermined also at the second time. Advantageously, this enables the dental practitioner tomonitor the progress of the preexisting inspection sites, e.g. by determining whether anychanges occurred between the first and the second time in terms of size, color, estimated dentalconditions, and any other property of the one or more preexisting inspection sites. The method may further comprise: -computing a difference value for each of the one or more preexisting inspection sites,the difference value quantifying a difference between a second size of the corresponding preexisting inspection site in the second set of inspection sites and a first size of the corresponding preexisting inspection site in the first set of inspection sites;- comparing the difference value with a threshold value, wherein responsive todetermining that the difference value is equal or larger than the threshold value thedifference value is classified as clinically relevant, and wherein responsive to determining that the difference value is smaller than the threshold value the difference value is classified as not clinically relevant. The difference value may be a positive number if the size of the inspection site has increasedbetween the first time and the second time, and the difference value may be a negative value ifthe size of the inspection site has decreased between the first time and the second time. Anegative difference value may be indicative of an improvement in the patient’s oral health, asit may indicate that a region of the dental site affected by dental conditions is getting smallerover time. This may indicate, for example, that a treatment plan is properly working. Therefore,it may be relevant for the dental practitioner to be provided with an indication of both an improvement and a worsening in the patient’s oral health. Furthermore, only the difference values among the one or more difference values which fulfill the threshold criterion are classified as being clinically significant, i.e. of interest from a clinical perspective. Clinically significant differences may be associated, for example, with a rate of change in the size of aninspection site which is above a threshold defined by the standards of the American DentalAssociation, or by the standards of the European Centre for Disease Prevention and Control, or by any other global or national health organization. Advantageously, this provides the dental practitioner with an automated method to identify regions of the dental site of the patient which need special attention as they do not satisfy dental health standards. The method may further comprise: -computing for each preexisting inspection site one or more difference values, eachdifference value quantifying a difference between a second severity level of each of theone or more dental conditions determined for the corresponding preexisting inspectionsite in the second set of inspection sites and a first severity level of the dental condition determined for the corresponding preexisting inspection site in the first set of inspection sites; -comparing the one or more difference values with a threshold value, wherein responsiveto determining that a difference value is equal or greater than the threshold value the difference value is classified as clinically relevant, and responsive to determining that a difference value is smaller than the threshold value the difference value is classified as not clinically relevant.The difference value may be a positive number if the severity level of a dental condition hasincreased between the first time and the second time, and the difference value may be a negativevalue if the severity level of the dental condition has decreased between the first time and thesecond time. A negative difference value may be indicative of an improvement in the patient’soral health, as it may indicate that the severity of a dental condition is decreasing over time.This may indicate, for example, that a treatment plan is properly working. In some cases, it may be relevant for the dental practitioner to be provided with an indication of both an improvement and a worsening in the patient’s oral health, and the absolute value of eachdifference value may be compared with the threshold value. Furthermore, only the differencevalues among the one or more difference values which fulfill the threshold criterion are classified as being clinically significant, i.e. of interest from a clinical perspective. Clinicallysignificant differences may be associated, for example, with a rate of change in the severity ofa dental condition which is above a threshold defined by the standards of the American DentalAssociation, or by the standards of the European Centre for Disease Prevention and Control, or by any other global or national health organization. Advantageously, this provides the dental practitioner with an automated method to identify regions of the dental site of the patient which need special attention as they do not satisfy dental health standards. Comparing the first augmentation data with the second augmentation data may further comprise determining one or more inspection sites in the second set of inspection sites forwhich no corresponding inspection site in the first set of inspection sites is identified and / ordetermining one or more inspection sites in the first set of inspection sites for which nocorresponding inspection site in the second set of inspection sites is identified. The one or moreinspection sites in the second set of inspection sites for which no corresponding inspection site in the first set of inspection sites can be identified may be representative of inspection sites which emerged in the dental site at the second time, and which were not present in the dentalsite at the first time. Advantageously, said inspection sites may be indicative of the emergenceof one or more dental conditions in new areas of the dental site at the second time compared to the first time, i.e. the areas of the dental site corresponding to the new inspection sites. Similarly, the one or more inspection sites in the first set of inspection sites for which no corresponding inspection site in the second set of inspection sites can be identified may be representative of inspection sites which disappeared between the first and second times. Advantageously, said inspection sites may be indicative of the regression of one or more dentalconditions in the areas of the dental site corresponding to the inspection sites which disappearedbetween the first and second times. The method may further comprise associating the determined one or more heat scores for the one or more differences with one or more data points of the second set of data points and rendering a view of the second 3D model on a GUI with a graphical representation of the one or more heat scores. Each determined heat score maps to one data point of the second 3D model, such that each data point of the second 3D model is associated with quantified local information relative to the one or more differences determined between the first time and thesecond time at the corresponding data point. This mapping enables the generation of a graphicalrepresentation of the one or more heat scores. Advantageously, the graphical representation enables the dental practitioner to identify at once the one or more locations of the dental site in which a clinical change has occurred since the previous clinical examination, and to directly localize those changes on the second 3D model.The method may further comprise associating the determined one or more heat scores for theone or more differences with one or more data points of the first set of data points and rendering a view of the first 3D model on a GUI with a graphical representation of the one or more heatscores. Each determined heat score maps to one data point of the first 3D model, such that eachdata point of the first 3D model is associated with quantified local information relative to theone or more differences determined between the first time and the second time at thecorresponding data point. This mapping enables the generation of a graphical representation ofthe one or more heat scores. Advantageously, the graphical representation enables the dental practitioner to identify at once the one or more locations of the dental site in which a clinical change has occurred with respect to the first time, and to directly localize those changes on the first 3D model. The graphical representation of the one or more heat scores for the one or more differences may be displayed as an indication on the view of the second 3D model and / or as an indication on the view of the first 3D model. The indication may be intended as a texture, a pointer, a symbol, or the like indicating the areas of the dental site in which one or more clinical changes have occurred since a previous clinical examination. Advantageously, the indication may be such that a contrast is generated between the rendered first / second 3D model and the one ormore heat scores, thereby highlighting in a clear way the locations of the first / second 3D modelfor which the presence of at least one clinical change is determined. This ensures that the dental practitioner is assisted and guided in performing the examination process of the dental site of the patient, whereby the time needed to perform said examination is significantly reduced with respect to the case where the dental practitioner performs the in-mouth examination withoutsuch a graphical representation. In practice, this fast examination could not have been achievedif the dental practitioner was not provided with the indication of the areas of the dental site which are highly affected by the one or more differences, i.e. with the indication of the one ormore heat scores and the corresponding epicenter(s).The indication may be a color different than a rendering color of the view of the second 3D model and different than a rendering color of the view of the first 3D model. This ensures that a contrast is generated between the one or more heat scores and the view of the first / second 3D model. Advantageously, the contrast draws the attention of the dental practitioner to the areas in which the one or more differences are determined.A different color may be used for the one or more heat scores determined for each of the oneor more differences, and an intensity of the color may decrease with a distance from the determined epicenter of the one or more heat scores for the corresponding difference. In practice, each of the one or more differences corresponds to a particular clinical change, e.g. a change in the size of an inspection site, a change in the severity level of a dental condition, achange in color and so forth. Therefore, using a different color for the heat scores for each ofthe one or more differences ensures that a contrast is generated between the visual indications corresponding to different categories of clinical change. For example, the one or more heat scores determined for a change in the color data between the first and second times may be displayed with blue, the one or more heat scores determined for a change in the severity of caries may be displayed with red, the one or more heat scores determined for a change in the size of a preexisting inspection site may be displayed with green and so forth. This ensures that the dental practitioner may quickly and intuitively distinguish between visual indications corresponding to different clinical changes. Advantageously, this reduces the ambiguity of the visual indication of the one or more heat scores, thereby facilitating and accelerating the clinical examination. Furthermore, the larger the intensity of the color in a region of the rendering of the first / second 3D model, the larger the difference between the first time and second time determined in that region. This is an advantage, as the dental practitioner may start the examination of the patient’s dental site by inspecting those locations of the dental site which are displayed with the largest color intensity on the first / second 3D model. Said locations may be, for example, the oneswhere the severity of a dental condition got worse, or they may be the ones where new dentalconditions developed, and so forth. Therefore, the intensity of the color ensures that theattention of the dental practitioner is drawn to the locations of the dental site which maynecessitate a treatment plan and / or a plan to prevent further worsening of the patient’s oral health. Then, the dental practitioner may proceed with the in-mouth examination by moving to areas of the dental site which are displayed with progressively decreasing intensity. In other words, the intensity of the color generates a hierarchy of a clinical significance of one or moreregions of the first / second 3D model in which the one or more clinical changes are determined,thereby providing the dental practitioner with a guide to perform the examination process which makes the diagnostic process quick and effective. The indication may be a pattern different than a rendering color of the view of the second 3D model and different than a rendering color of the view of the first 3D model. This ensures that a contrast is generated between the one or more heat scores and the view of the first / second 3D model. Advantageously, the contrast draws the attention of the dental practitioner to the areas in which the one or more differences are determined.A different pattern may be used for the one or more heat scores determined for each of the oneor more differences, and a density of the pattern may decrease with a distance from theepicenter of the one or more heat scores of the corresponding difference. Each of the one or more differences corresponds to a particular clinical change, e.g. a change in the size of an inspection site, a change in the severity level of a dental condition, a change in color and so forth. Therefore, using a different pattern for the heat scores determined for each of the one or more differences ensures that a contrast is generated between the visual indications corresponding to different categories of clinical change. This ensures that the dental practitioner may quickly and intuitively distinguish between visual indications correspondingto different clinical changes. Advantageously, this may reduce the ambiguity of the visualindication of the one or more heat scores, thereby facilitating and accelerating the clinical examination.Furthermore, the larger the density of the pattern in a region of the rendering of the first / second3D model, the larger the difference between the first time and second time determined in that region. This is an advantage, as the dental practitioner may start the examination of the patient’s dental site by inspecting those locations of the dental site which are displayed with the largestpattern density on the first / second 3D model. Said locations may be the ones where the severityof a dental condition got worst, or they may be the ones where new dental conditions developed.Therefore, the density of the pattern ensures that the attention of the dental practitioner is drawnto the locations of the dental site which may necessitate a treatment plan and / or a plan to prevent further worsening of the patient’s oral health. Then, the dental practitioner may proceed with the in-mouth examination by moving to areas of the dental site which are displayed withprogressively decreasing density. In other words, the density of the pattern generates ahierarchy of a clinical significance of one or more regions of the first / second 3D model in which the one or more clinical changes are determined, thereby providing the dental practitioner with a guide to perform the examination process which makes the diagnostic process quick and effective. The one or more differences may define one or more regions of clinical change in the first 3D model and / or in the second 3D model, and each of the one or more differences may be associated with a type of clinical change. Therefore, the one or more differences are accurately localized in the first / second 3D model, enabling the dental practitioner to quickly and effectively identify the areas of the dental site which may necessitate an in-depth examination. The type of clinical change is associated with one of: a change in a size of a preexisting inspection site, a change in a severity level of a dental condition among one or more dental conditions, a color change in a region of dental site, or a change in the 3D geometry of a region of the dental site. These are among the most common clinical changes which may occur in the dental site of the patient between clinical examinations. However, the present disclosure may be suitable for any type of clinical change not listed above. The method may further comprise determining one or more aggregated heat scores for the one or more differences for each of the one or more regions of clinical change. Therefore, when more than one difference is determined in a region of clinical change, the aggregated heat score may quantify the total change occurred in that region. Advantageously, this may provide thedental practitioner with quantified information relative to the overall change occurred in oneor more regions of the dental site. Determining the one or more aggregated heat scores may comprise summing the one or moreheat scores for each of the one or more differences of the corresponding region of clinicalchange. Advantageously, the summing operation may enable to save computational time andtherefore to accelerate the method of the present disclosure.Determining the one or more aggregated heat scores may comprise computing a weighted sumof the one or more heat scores for each of the one of more differences of the corresponding region of clinical change. Advantageously, the weighted sum ensures that the aggregated heat score of each given region of clinical change is determined based on realistic properties of each of the one or more differences determined in that region of clinical change. The realistic properties may, for example, comprise a clinical significance of each of the one or more differences. Computing the weighted sum may comprise assigning a weight to each of the one or more differences, wherein the weight may be based on a clinical significance of the corresponding difference. This ensures that the differences which are considered more important from aclinical perspective, e.g. because their stage of development is advancing quickly relative to ahealth standard and / or because they may be correlated with the insurgence of other dentalconditions and / or because they do not fulfill global or national health standards, contributemore to the aggregated heat score than differences which are instead categorized as lessimportant with respect to a clinical standard. Advantageously, the weighted sum enables todetermine a realistic change of the overall health status of the dental site of the patient betweenthe first and second times. The first augmentation data may further comprise a first rate of change of a severity of each of one or more dental conditions determined at each inspection site in the first set of inspectionsites, and the second augmentation data may further comprise a second rate of change of theseverity of each of the one or more dental conditions determined at each inspection site in the second set of inspection sites. The first rate of change of the severity of each of the one or moredental conditions may provide the dental practitioner with an indication of how quickly thecorresponding dental condition was progressing or regressing at a previous clinical examination of the patient, i.e. at the first time. Similarly, the second rate of change of the severity of each of the one or more dental conditions may provide the dental practitioner withan indication of how quickly the corresponding dental condition is progressing or regressing ata subsequent clinical examination of the patient, i.e. at the second time. Advantageously, therates of change can provide the dental practitioner with a quantitative tool to determine theeffectiveness of a treatment, and / or to take preventive action before one or more conditionsbecomes severe, and so forth.The first rate of change of each dental condition may be determined based on a differencebetween a first severity level of the corresponding dental condition determined at a first time and a second severity level of the corresponding dental condition determined at a second time, wherein the second time may be subsequent to the first time. The first time may be the time when a first clinical examination of the patient was performed, and the second time may be the time when a second clinical examination of the patient was performed. Advantageously, the first rate of change of the severity of each dental condition may provide the dental practitioner with a quantitative indication of how quickly the severity of the corresponding dental conditionimproved or worsen between the first time and the second time. This quantitative indicationmay be used by the dental practitioner to track the progress of a treatment between the two clinical examinations. The second rate of change of the severity of each dental condition may be determined based on a difference between a third severity level of the corresponding dental condition determined at a third time and a fourth severity level of the corresponding dental condition determined at afourth time, wherein the fourth time may be subsequent to the third time. The third time maybe the time when a third clinical examination of the patient was performed, and the fourth time may be the time when a fourth clinical examination of the patient is performed. Advantageously, the second rate of change may provide the dental practitioner with a quantitative indication of how quickly the severity of the corresponding dental conditionimproved or worsen between the third time and the fourth time. This quantitative indicationmay be used by the dental practitioner to track the progress of a treatment between the two clinical examinations. The third time may coincide with the second time, whereby the second severity level may coincide with the third severity level. Therefore, three clinical examinations of the patient maybe performed: the first one at the first time A, the second one at the time second (the same asthe third time) time B, and the fourth one at the fourth time C. Accordingly, the first rate ofchange may quantify how quickly the dental condition changed between A and B, and similarly the second rate of change may quantify how quickly the dental condition changed between B and C. Advantageously, the comparison between the first rate of change and the second rate of change may provide the dental practitioner with an objective and quantitative indication ofwhether the rate of change increased or decreased in between clinical examinations. This maybe helpful to determine the progress of a treatment over consecutive time intervals, for caseswhen it is important to slow down the worsening of the dental condition.Alternatively, the third time may be subsequent to the second time. In this case, four clinicalexaminations of the patient may be performed: the first one at the first time A, the second oneat the second time B, the third one at the third time C and the fourth one at the fourth time D. Accordingly, the first rate of change quantifies how quickly the severity of the dental condition changed between A and B, and the second rate of change quantifies how quickly the severityof the dental condition changed between C and D. Advantageously, the comparison betweenthe first rate of change and the second rate of change may provide the dental practitioner with an objective and quantitative indication of a trend of the rate of change of the dental condition over time. This may be helpful to determine the difference between the rate of change of the severity of a dental condition in two different conditions, for example when a first treatment is interrupted at time B and a second treatment is started at time C. Comparing the first augmentation data with the second augmentation data may comprisedetermining a difference between a second rate of change of the severity of each dentalcondition and a first rate of change of the severity of the corresponding dental condition.Advantageously, comparing the rates of change provides the dental practitioner with anobjective and quantitative indication of the efficiency of a treatment. In some cases, this mayassist the dental practitioner in estimating an ending time of the treatment, e.g. by estimating,based on the difference between two rates of change, when the change will be between 90%and 100%, meaning that the dental condition almost or completely disappeared. Brief description of the drawings Aspects of the present disclosure may be best understood from the following detailed description alongside the following drawings. The drawings are schematic and simplified for the sake of clarity, and they are intended to show details to improve the understanding of the claims, whereas other details may be left out. Throughout the description and the drawings, the same reference numbers are used for identical or corresponding parts. The individual features of each aspect may each be combined with any or all features of the other aspects. These and other aspects, features and / or technical effects will be apparent from and elucidated with reference to the illustrations described hereinafter in which:Figure 1 illustrates a flowchart according to an embodiment of the computer-implementedmethod disclosed herein; Figure 2 illustrates an example of dental scanning system according to the present disclosure;Figures 3a and 3b illustrate example reference diagrams for describing an embodiment of aworkflow of the computer-implemented method disclosed herein;Figure 4 illustrates an example reference diagram for describing a trained machine learningmodel for a dental condition on which the computer-implemented method disclosed herein may be based;Figures 5a and 5b illustrate example reference diagrams for describing an embodiment of aworkflow of the computer-implemented method disclosed herein;Figures 6a, 6b and 6c illustrate a flowchart according to an embodiment of the computerimplemented method disclosed herein;Figures 7a, 7b, 7c, 7d, 7e, 7f and 7g illustrate an example of computer-implemented smoothingmethod according to the present disclosure;Figures 8a and 8b illustrate flowcharts according to embodiments of the computer-implemented method disclosed herein;Figures 9a and 9b illustrate examples of a graphical representation on a graphical-user interface(GUI) of an output of the computer-implemented method disclosed herein;Figure 10 illustrates an example of a visualization on a graphical-user interface (GUI) of anoutput of the computer-implemented method disclosed herein.Figure 11 illustrates an embodiment of the computer system disclosed herein.Detailed description The detailed description set forth below in connection with the appended drawings is intended as a description of various examples according to the disclosure. The detailed descriptionincludes details for the purpose of providing a thorough understanding of various concepts andexamples covered throughout the description. However, it will be apparent to those skilled in the art that these concepts and examples may be practiced without the specific details mentioned or in combination with one or more examples described herein. Several examples of the devices, systems, mediums, programs and methods are described by various modules, components, steps, processes, algorithms, etc. Depending upon particular application, design constraints or other reasons, these elements may be implemented using electronic hardware, computer program, or any combination thereof. In the following several examples of the method and system described herein will be disclosed in more detail. The solutions presented herein generally aim at improving the automatic determination of anoverall health status of a dental site by determining a location of one or more epicenters of oneor more dental conditions, including caries, cracks, recession, wear, plaque, or any other relevant restorative elements forming part of the dental site. The solutions presented hereinfurther aim at guiding or improving the dental health assessment and treatment process byproviding a diagnostic tool that ensures an efficient and objective process. Digital information relative to a patient’s dental site may be obtained during a scanning session,when a dental practitioner scans, by means of an intraoral scanner, the dental site within an oralcavity of the patient. The intraoral scanner may be configured to acquire surface information of the dental site. The intraoral scanner may employ a scanning principle such as triangulation- based scanning, confocal scanning, focus scanning, ultrasound scanning, stereo vision, structure from motion, optical coherent tomography OCT, or any other scanning principle.Detailed information relative to examples of intraoral scanners will be given in the followingdetailed description.Figure 1 illustrates a flowchart 100 of an embodiment of the computer-implemented methoddisclosed herein. A three-dimensional (3D) model of a dental site is obtained in step 101. Thedental site is part of an oral cavity of a patient, and it comprises at least the patient’s teeth andgingiva, but it may also comprise soft palate, hard palate and any other part of the patient’s oralcavity. The obtained 3D model may be generated during a clinical visit of the patient, uponscanning the dental site of the patient with an intraoral scanner. Alternatively, the 3D modelmay have been generated and stored during a previous clinical visit of the patient, and thus said3D model of the dental site may be obtained upon loading the stored 3D model, e.g. from amemory where it has been stored during the previous scanning session. The 3D model maycomprise a plurality of data points which collectively describe a surface of the dental site in 3D space. For example, the plurality of data points may comprise a point cloud, a polygon mesh, a voxel model, a triangulated point cloud, or any other method of storing information about asurface of the scanned dental site in 3D space, i.e. a 3D geometry of the dental site. In anotherexample, the plurality of data points may comprise a triangulated mesh, wherein the triangulated mesh comprises a set of triangles which are connected by their common edges orvertices. In general, the plurality of data points comprised in the 3D model describes at leastthe 3D geometry of the surface of the patient’s dental site.In step 102, a presence of one or more dental conditions is estimated, thereby defining one ormore inspection sites in the 3D model, wherein each inspection site comprises one or more ofthe plurality of data points. The one or more dental conditions may comprise one or more of caries, cracks, wear, plaque, recession, and any other dental condition which may be of interest for assessing the overall dental health of the patient. Estimating the presence of the one or more dental conditions may comprise processing the plurality of data points comprised in the 3D model using a trained machine learning model for each dental condition. For example, the presence of caries may be estimated by inputting each data point in a trained machine learningmodel for caries detection, the presence of cracks may be estimated by inputting each datapoint in a trained machine learning model for cracks detection, the presence of wear may beestimated by inputting each data point in a trained machine learning model for wear detection,and so forth for each dental condition. The trained machine learning model for each dental condition may be a pre-trained machine learning model that has been trained using a training data set comprising data points similar to the plurality of data points of the 3D model, whereinthe training data set also comprises information about features of each data point of the trainingdata set. For example, the features may be the presence or absence of caries for each data pointin the training data set for the trained machine learning model for caries, and the features maybe the presence or absence of cracks for each data point in the training data set for the trainedmachine learning model for tooth cracks, and so forth for each dental condition. The featuresmay further or alternatively represent a severity label, and / or size, and / or area and so forth of each dental condition. The trained machine learning model for a specific dental condition may output for each datapoint a severity level of that dental condition, wherein the severity level may be a scoreindicating the severity of that condition. For example, the severity level of the dental conditionmay be chosen from the group of no condition (N), initial condition (I), moderate condition(M) and / or severe condition (S). Alternatively, the severity level of the dental condition maybe chosen between presence (P) of the condition and no presence (NP) of the condition. Further,each severity level may be associated with a number on a predefined scale of severity of each dental condition, such that each severity level quantifies the severity of that condition. The scale of severity of each dental condition may be defined in a suitable manner describingclinical aspects of that specific condition, i.e. different numbers may be comprised in the scaleof severity of different dental conditions based on clinical features of each dental condition.The scale of severity may comprise a minimum number, e.g. the number zero, which isindicative of an absent dental condition. For example, the severity level N may be associatedwith the number 0, the severity level I may be associated with the number 1, the severity levelM may be associated with the number 3, and the severity level S may be associated with thenumber 9, i.e. the mapping: (^, ^, ^, ^) ↔ (0, 1, 3, 9) may hold for a specific dental condition. In another example, the severity level NP may be associated with the number 0, and the severity level P may be associated with the number 1, i.e. the mapping: (^^, ^) ↔ (0, 1)may hold for a specific dental condition. The above-mentioned mappings are intended asexamples, and any severity scale suitable to quantify the severity of a dental condition may beused. The one or more data points for which the presence of at least one dental condition isestimated by the trained machine learning model define one or more inspection sites of the 3Dmodel. In particular, the one or more data points comprised in each inspection site areassociated with at least one severity level of a dental condition which is different from theminimum severity level of that dental condition. In step 103, an aggregated severity level of the one or more dental conditions for each of thedata point in each inspection site is determined, thereby determining one or more aggregatedseverity levels for each inspection site. The aggregated severity level for each data point may be determined by summing the severity level of each dental condition determined for that data point. For example, referring to a same data point, a severity level of 3 may have been determined for caries, a severity level of 0 may have been determined for cracks, a severitylevel of 9 may have been determined for wear, a severity level of 3 may have been determinedfor recession, and a severity level of 1 may have been determined for plaque. In this example,the aggregated severity level of (3+0+9+3+1)=16 is determined for the data point.The aggregated severity level for each data point may alternatively be determined bydetermining a weighted sum of the severity level determined for each dental condition. In thiscase, each dental condition contributes to the aggregated severity level with a weight which may depend on the clinical significance of the dental condition, and the clinical significance of a dental condition may be established by a dental practitioner or by any other dental healthexpert. For example, the clinical significance of a dental condition may be chosen based on therate at which the stage of the dental condition advances or based on a possible insurgence ofother conditions due to the presence of that dental condition. For example, the caries may beweighted with a weight w1=3, the cracks may be weighted with a weight of w2=1, the wear may be weighted with a weight of w3=2, the recession may be weighted with a weight of w4=2 and the plaque may be weighted with a weight of w5=1. Thus, also with reference to theexample above, the aggregated severity level (^^ × 3 + ^^ × 0 + ^^ × 9 + ^^ × 3 + ^^ × 1) = 34 isdetermined for the data point. It will be appreciated by those skilled in the art that the sum ofthe severity level of each dental condition has the advantage of reducing the processing timeof the data points. On the other hand, the weighted sum has the advantage of providing anaccurate and realistic indication of an overall health condition of a data point, as it accounts forrealistic clinical features of each dental condition. In step 104, an epicenter of the one or more aggregated severity levels of each inspection site is determined by determining a maximum aggregated severity level of the inspection site. The epicenter of each inspection site may comprise more than one data point if more data points ofthe inspection site are determined to have the same maximum aggregated severity level. Forexample, an inspection site may contain 6 data points ^^, ^^, ^^, ^^, ^^, ^^ for which theaggregated severity levels 10, 15, 15, 15, 5, 4 have been determined, respectively. In thisexample, the maximum severity level of the inspection site is determined to be 15, which isassociated with the data points Therefore, the epicenter of the inspection sitecomprises the data point ^^, ^^, ^^ and is associated with the aggregated severity level 15.In step number 105, the one or more data points of the maximum aggregated severity level determined for each inspection site are associated with a location of the epicenter. Thus, one ormore locations of the dental site for which the highest aggregated severity level has been locallydetermined are accurately determined. These locations may be reported to the dentalpractitioner performing the scanning, e.g. as a graphical representation displayed on a renderingof the 3D model of the dental site on a display of the computer system on which the method isperformed. In this way, the dental practitioner may be provided with a clear indication of theoverall health status of the patient’s dental site at once, and with an indication of the one ormore regions of the dental site which necessitate a careful inspection, either to treat existingdental conditions and / or to prevent the development of other conditions. Figure 2 illustrates an example of intraoral scanning system 200 according to the present disclosure. The intraoral scanning system 200 comprises an intraoral scanner 201 and a computer system 205. The intraoral scanner 201 may be configured to acquire intra-oral scan data relative to a dental site 210 of a patient, for example during a live scanning session whichmay be carried out by a dental practitioner when the patient visits a dental studio. As alsomentioned above, the intraoral scanning device 201 may employ any scanning principle such as triangulation-based scanning, confocal scanning, focus scanning, ultrasound scanning, stereo vision, structure from motion, optical coherent tomography OCT, or any other scanning principle suitable for acquiring intraoral scan data. In one example, the intraoral scanner 201 may be configured to acquire surface information of the dental site 210 by operated by projecting a pattern and translating a focus plane along an optical axis of the intraoral scanner 201, and capturing a plurality of two-dimensional (2D) images at different focus plane positions such that each series of captured 2D images corresponding to each focus plane forms a stack of 2D images. The stack of 2D images is also referred to herein as sub-scan. During the scanning, a number of sub-scans is acquired for a number of given angle view of the dental site210, i.e. for a given arrangement of the intraoral scanner 201 relative to the dental site 210. Theintraoral scanner 201 is generally moved and angled relative to the dental site 210, such thatleast sets of sub-scans overlap at least partially, in order to generate a reconstruction of the 3D model of the dental site 210 by stitching overlapping sub-scans together in real-time. The stitching process, also known as registration and fusion, may be implemented by means of an Iterative Closest Point (ICP) algorithm. As the generation of the reconstruction of the 3D model progresses, the progress of the reconstruction of the 3D model may be displayed on a display206 of the computer system 205. In another example, intraoral scanner 201 is a triangulationscanner, where a time varying pattern is projected onto the dental object and a sequence of images of the different pattern configurations are acquired by one or more cameras located at an angle relative to the projector unit. Color texture of the dental site 210 may be acquired by illuminating the dental site 210 using different monochromatic colors such as individual red, green and blue colors or by illuminating the object using multichromatic light such as white light. A 2D image may be acquired during a flash of white light. More generally, the intraoral scanner 201 comprises one or more light projectors 204 configured to generate an illumination pattern to be projected on at least a part of the dental site 210 during a scanning session. The light projector(s) preferably comprises a light source, a mask having a spatial pattern, and one or more lenses such as collimation lenses or projection lenses. The light source may be configured to generate light of a single wavelength or acombination of wavelengths (mono- or polychromatic). The combination of wavelengths maybe produced by using a light source configured to produce light (such as white light) comprising different wavelengths. Alternatively, the light projectors may comprise multiple light sources such as Light Emitting Diodes (LEDs) individually producing light of different wavelengths (such as red, green and blue) that may be combined to form light comprising the different wavelengths. Thus, the light produced by the light source may be defined by a wavelength defining a specific color, or a range of different wavelengths defining a combination of colors such as white light. In an embodiment, the scanning device comprises a light source configured for exciting fluorescent material of the teeth to obtain fluorescence data from the dental object. Such a light source may be configured to produce a narrow range of wavelengths. In another embodiment, the light from the light source is infrared (IR) light, which is capable of penetrating dental tissue. The light projector(s) may be Digital Light Processing (DLP)projectors using a micro mirror array to generate a time varying pattern, or a diffractive opticalelement (DOF), or back-lit mask projectors, wherein the light source is placed behind a maskhaving a spatial pattern, whereby the light projected on the surface of at least a part of the dentalsite 210 is patterned. The back-lit mask projector may comprise a collimation lens forcollimating the light source, said collimation lens being placed between the light source and the mask. The mask may have a checkboard pattern, such that the generated illumination pattern is a checkboard pattern. Alternatively, the mask may feature other patterns such as lines or dots, etc.The intraoral scanner 201 may further comprise optical components for directing the light fromthe light source to the surface of the dental object. The specific arrangement of the opticalcomponents depends on whether the intraoral scanner 201 is a focus scanning device, a scannerusing triangulation, or any other type of intraoral scanning device. The light reflected from the dental site 210 in response to the illumination of the dental site isdirected, using the optical components of the intraoral scanner 201, towards the image sensor(s)202. The image sensor(s) 202 are configured to generate a plurality of images based on the incoming light received from the illuminated dental object. The image sensor(s) 202 may be a high-speed image sensor such as an image sensor configured for acquiring images with exposure of less than 1 / 1000 second or frame rates in excess of 250 frames pr. second (fps). Asan example, the image sensor may be a rolling shutter (CCD) or global shutter sensor (CMOS).The image sensor(s) 202 may be a monochrome sensor including a color filter array such as a Bayer filter and / or additional filters that may be configured to substantially remove one or more color components from the reflected light and retain only the other non-removed components prior to conversion of the reflected light into an electrical signal. For example, such additional filters may be used to remove a certain part of a white light spectrum, such as a blue component, and retain only red and green components from a signal generated in response to exciting fluorescent material of the teeth.The intraoral scanning system 200 may further comprise one or more processors 203configured to generate scan data (such as intraoral scan data) by processing the data acquired by the intraoral scanner 201, e.g. by processing the 2D images acquired by the intraoral scanner 201. The one or more processors 203 may be a part of the intraoral scanner 201. For example,the processor(s) 203 may comprise a Field-programmable gate array (FPGA) and / or anAdvanced RISC Machine (ARM) processor located in the intraoral scanner 201. The scan datamay comprise any of: 2D images, 3D point cloud, depth data, texture data, intensity data, color data, and / or combinations thereof. As an example, the scan data may comprise one or more point clouds, wherein each point cloud comprises a set of 3D points describing at least a part of the 3D dental site 210. As another example, the scan data may comprise images, each image comprising image data e.g. described by image coordinates and a timestamp (x,y,t), whereindepth information can be inferred from the timestamp. The intraoral scanner 201 may beconfigured for acquiring a number of sub-scans (i.e. a stack of raw 2D images, wherein rawmeans that the 2D images are not processed) by means of the image sensor(s) 202. In this case,the sub-scans are provided as input to the processor(s) 203. The processing of the raw 2Dimages may comprise the step of determining which part of each of the 2D images are in focus in order to generate / deduce depth information from the images. Said depth information may be used to generate 3D point clouds comprising a set of 3D points in space described by a set ofCartesian coordinates (x,y,z). The 3D point clouds may be generated by the processor(s) 203or by any other processing unit. Each 2D / 3D point may further comprise a timestamp indicating when the 2D / 3D point was recorded, i.e. from which image in the stack of 2D images the point originates. The timestamp is correlated with the z-coordinate of the 3D points, i.e. the z-coordinate may be inferred from the timestamp. Accordingly, the output of the processor(s)203 is the scan data, and the scan data may comprise image data and / or depth data, e.g.described by image coordinates and a timestamp (x, y, t) or alternatively described as (x, y, z).The intraoral scanner 201 may be configured to transmit, either through wired connection orwireless connection 209, scan data and other types of data to the computer system 205.Examples of data include any of 3D information, texture information such as IR images, fluorescence images, reflectance color images, x-ray images, and / or combinations thereof. Thecomputer system 205 may comprise a display 206 to display a rendering 207 of the 3D modelof the dental site 210 generated during the scanning session, and / or a rendering 207 of the 3Dmodel of the dental site 210 generated in a scanning session of the dental site 210 performed at a previous time. The system 200 may further be configured to display a live view of the intraoral scanner 201. The scanner system 200 may be further configured to detect one or more input signal(s). As an example, the input signal may be detected through one or more sensors (not shown here) provided in the intraoral scanner 201. In this example, the intraoral scanner may be further configured to transmit the input signal to the computer system 205. In another example, the input signal may be further or alternatively detected through one or more external devicesconnected to the computer system 205, such as a keyboard 208 and / or a mouse 211 or any otherinput device. Figures 3a and 3b illustrate example reference diagrams 300 for describing an embodiment ofa workflow of the computer-implemented method disclosed herein. Referring to Figure 3a, a3D model 301 of a dental site is obtained. Said 3D model may be generated during a scanningsession of the dental site of a patient performed during a clinical visit of the patient, or it mayhave been generated in a scanning session performed at a previous time and stored, e.g. in a memory of a computer system on which the method disclosed herein is performed, such as the computer system indicated as reference number 205 in Figure 2. The obtained 3D model comprises a plurality of data points 302. In one example, the plurality of data points 302 may be a point cloud / a plurality of point clouds in 3D space, wherein each point cloud comprises a set of points in 3D space described by a set of Cartesian coordinates (x, y, z). In another example, the plurality of data points 302 may comprise a plurality of facets comprised in a mesh describing the surface of the dental site in 3D space. Each facet may be a triangle in the case of a triangulated mesh, or any other polygon in the case of a polygonal mesh. In yet another example, the plurality of data points 302 may comprise a plurality ofvertices comprised in a mesh describing the surface of the dental site in 3D space. In general,the plurality of data points may comprise any other data which is suitable to describe the surfaceof the dental object in 3D space.Once the 3D model 301 has been obtained, it is processed in step 303 to estimate a presence ofone or more dental conditions of the dental site. The one or more dental conditions maycomprise one or more of caries, cracks, recession, wear, plaque, and any other dental condition which may be of interest to the dental practitioner to assess the health status of the dental site.The processing 303 may output a number of inspection sites, wherein each inspection sitecomprises one or more of the plurality of data points 302 for which at least one dental conditionhas been estimated in 303. The processing 303 may comprise processing each of the plurality of data points, and therefore the output of 303 comprises an output for each of the plurality of data points 302. For example, with reference to Figure 3a, the processing 303 may define threeinspection sites of the 3D model 301: 304a comprising three data points (^^^ , ^^^ , ^^^), 304bcomprising three data points (^^^ , ^^^ , ^^^), and 304c comprising two data points (^^^ , ^^^). Ingeneral, the inspection sites defined by the output of 303 may be intended as regions of the 3D model 301 which are affected by at least one dental condition. The example given in Figure 3aillustrates two inspection sites comprising three data points and one inspection site comprisingtwo data points, but it will be appreciated by those skilled in the art that an inspection site maycomprise more or less than three / two data points. In step 305 an aggregated severity level is determined for each data point comprised in each inspection site, whereby one or more aggregated severity levels are determined for eachinspection site. The aggregated severity level of a data point may also be referred to herein as“heat score” of the data point. The heat score of a data point quantifies the overall health statusof that data point, and it therefore indicates the aggregated / overall severity of the one or moredental conditions estimated for the data point. For example, if for the data point ^^^ thepresence of caries, cracks and plaque is estimated by the processing 303, the heat score ^^^determined for ^^^indicates the aggregated severity determined for caries, cracks andplaque for ^^^ . With reference to Figure 3a, the heat scores 305a, 305b, 305c are determinedfor the inspection site 304a, the heat scores 305d, 305e, 305f are determined for the inspection site 304b and the heat scores 305g, 305h are determined for the inspection site 304c, namely the heat scores: are determined in 305 for each data point comprised in the inspection site 304a, 304b and 304c,respectively. In step 306, an epicenter of the one or more dental conditions is determined for each inspection site by determining a maximum heat score of each inspection site, namely: ^^ = max[^^^ , ^^^ , ^^^] = ^^^ ^^ = max[^^^ , ^^^] = ^^^ wherein ^^is the epicenter 306a determined for the inspection site 304a, ^^is the epicenter 306b determined for the inspection site 304b, and ^^is the epicenter 306c determined for the inspection site 304c. In the example shown in Figure 3a, the maximum heat score of theinspection site 304a is determined to be ^^^ , i.e. the epicenter 306a is ^^ = ^^^ , and themaximum heat score of the inspection site 304b is determined to be ^^^, i.e. the epicenter 306bis ^^ = ^^^ , and the maximum heat score of the inspection site 304c is determined to be ^^^ ,i.e. the epicenter 306c is ^^ = ^^^. It will be appreciated by those skilled in the art that two ormore data points of an inspection site may be assigned with the same heat score and that saidsame heat score may be determined to be the maximum heat score of the inspection site. In thiscase, the epicenter of the inspection site comprises the two or more data points with the samemaximum heat score.Once the epicenter of each inspection site is determined, the one or more data points of themaximum heat score of each inspection site are associated with a location of the epicenter .With reference to Figure 3a, the data point ^^^ may be associated with the location 307a of theepicenter ^^ 306a, the data point ^^^ may be associated with the location 307b of the epicenter^^ 306b, and the data point ^^^ may be associated with the location 307c of the epicenter^^ 306c. Therefore, the regions of the dental site for which the overall health status isdetermined to be locally the most severe, i.e. the regions of the dental site which necessitatethe highest attention, are accurately determined. The locations of the epicenters of the one ormore inspection sites are the regions of the dental site which may benefit from a carefulexamination by the dental practitioner, and potentially demand a treatment plan. Thus, thedental practitioner may use the location of the epicenter of each inspection site as a guide toperform the manual examination of the dental site of the patient. For example, the examinationmay be started from the location of the epicenter of each inspection site, and it may proceed bychecking the health condition of the region around said epicenter. Such a guide significantly improves and fasten the diagnostic efficiency, allowing the dental practitioner to quicklyidentify and inspect the problematic locations of the dental site, potentially preventing thedevelopment of additional dental conditions in the location of the epicenter.Even though in the example illustrated in Figure 3a the epicenter is determined for eachinspection site of the 3D model, in an alternative embodiment (not illustrated here) the methoddisclosed herein may comprise determining only one epicenter of the dental conditions, namelya global epicenter of the dental conditions, by determining a global maximum heat score amongthe one or more heat scores determined for all of the inspection sites of the 3D model. In thiscase, the global epicenter may be intended as the region of the 3D model for which the healthstatus is, globally, the worst and which therefore necessitates the highest attention by the dental practitioner. In other words, the method described in the example illustrated in Figure 3a may comprise determining at least one epicenter of the dental conditions in the 3D model of the dental site, and said epicenter may be intended as a local and / or as a global epicenter.Figure 3b is a reference diagram of an embodiment of the processing for estimating a presenceof one or more dental conditions in a dental site as the one illustrated as reference number 303in Figure 3a. Estimating the presence of the one or more dental condition in the dental sitecomprises processing each of the plurality of data points 302 comprised in the 3D model 301of the dental site using a trained machine learning model for each dental condition 309.Accordingly, if the 3D model 301 comprises N data points 302^^(308^), ^2 (308^), ^^(308^), … , ^^(308^ ), each of these data points is input in a trainedmachine learning model for each dental condition. Referring to Figure 3b, the processing 303 comprises estimating a presence of a number of K dental conditions. Accordingly, each of the N data points 302 are inputted in a trained machine learning model 309a for the dental condition number 1, in a trained machine learning model 309b for dental condition number 2, and so forth, up to the trained machine learning model309c for dental condition number K. For each data point, each of the trained machine learningmodels for a specific dental condition outputs a severity level for that dental condition. Inparticular, the trained machine learning model 309a outputs for each data point in 302 a severity level 310a for the dental condition number 1, the trained machine learning model 309b outputsfor each data point in 302 a severity level 310b for dental condition number 2, and so forth, upto the trained machine learning model 309c which outputs for each data point in 302 a severitylevel 310c for dental condition number K. Therefore, referring to Figure 3a the output of thetrained machine learning models 309, i.e. the output of the processing 303, is: wherein ^^^, ^^^, … , are the severity levels 310a of dental condition 1 determined by themachine learning model 309a for the data point^ (308^), ^ (308^), ^ (308^), … , ^ (308^), respectively, ^^, ^^^ ^ ^ ^ ^ ^, … , ^^^are the severitylevels 310b of dental condition 2 determined by the machine learning model 309b for the data( ) ( ) ( ) point ^ 308^ , ^ 308^ , ^ 308^ , … , ^ (308d), respectively, and so forth, up to^ ^ ^ ^^ ^^, ^ , … , ^^ which are the severity levels 310c determined by the machine learning model^ ^ ^( ) ( ) 309c of dental condition K for the data point ^ 308^ , ^ 308^ , ^ (308^), … , ^ (308^),^ ^ ^ ^respectively. The example illustrated in Figure 3b refers to a trained machine learning modelto detect the presence of the one or more dental conditions, however the presence of each dental condition in the dental site may be estimated by using other methods, e.g. mathematical and / or statistical methods using color information comprised in the scan data used to generate the 3D model of the dental site.Referring again to Figure 3a, a severity level for each of the K dental conditions may bedetermined for each data point in the defined inspection sites, i.e. the inspection sites 304a, 304b and 304c, wherein the severity level of each of the K dental conditions is determined accordingly to the workflow of the processing 303 illustrated in Figure 3b. For example, for ^^^, ^ , … , ^the data point ^ the severity levels ^ may be determined in 303 for the K dental^^ ^^ ^^ ^^conditions. Accordingly, the aggregated severity level ^ 305a determined for the data point^^^ ^^, ^ , … , ^^ is determined by aggregating the severity levels ^ . Similarly, for the data^^ ^^ ^^ ^^^ ^^, ^ , … , ^point ^ the severity levels ^ may be determined in 303 for the K dental^^ ^^ ^^ ^^conditions. Accordingly, the aggregated severity level ^ 305d determined for the data point^^^ ^^, ^ , … , ^^ is determined by aggregating the severity levels ^ . Similar discussion holds^^ ^^ ^^ ^^for the aggregated severity levels 305c, 305d, 305e, 305f, 305g, 305h determined for the datapoints ^ , ^ , ^ , ^ , ^ , ^ , respectively. Aggregating the severity level of each of the K^^ ^^ ^^ ^^ ^^ ^^dental conditions may comprise summing the severity level determined for each of the K dental conditions, or it may comprise determining a weighted sum of the severity level determined for each of the K dental conditions, wherein a weight of each dental condition may be determined based on a clinical significance of that dental conditions or based on clinical features of that dental condition.Figure 4 illustrates an example reference diagram 400 of an embodiment of a trained machinedlearning model for a dental condition according to the present disclosure. The trained machinelearning model for the dental condition 401 receives as input a number N of data points302 (^ , ^ , … , ^ ). The N data points 302 may be N points of a 3D point cloud, or the data^ ^ ^points may be N facets of a triangulated mesh, or the data points may be N vertices of a triangulated mesh, or the data points may be any other data suitable to describe a surface of thedental site in 3D space. For each data point ^^ 406, the trained machine learning model 401outputs a probability ^^ 408, where here the index ^ = [1, … , ^ ]. Said probability ^^ 408 maybe a vector, and each element of the vector may be a probability that the dental condition ispresent in that data point with a specific severity level, wherein the latter is a number on apredefined scale of severity 402 of the dental condition. Therefore, the probability ^^408comprises as many elements as the number of severity levels on the predefined scale of severityof the dental condition 402. In general, the scale of severity of a specific dental conditioncomprises numbers, and each of said numbers quantifies the severity of the specific dentalcondition. For example, referring to Figure 4, the scale of severity of the dental condition maycomprises the numbers 0 (407a), 1 (407b), 3 (407c), 9 (407d), corresponding to an absence ofthe dental condition, an initial dental condition, a moderate dental condition and a severe dentalcondition, respectively. Thus, the probability ^^ 408 which is output for each data point ^^ 406by the trained machine learning model 401 may be a vector ^^ = [^^^ ^^^ ^^^ ^^^], wherein ^^^is the probability 403a that the dental condition is present in the data point ^^ with the severitylevel 0407a, ^^^ is the probability 403b that the dental condition is present in the data point ^^with the severity level 1407b, ^^^ is the probability 403c that the dental condition is present inthe data point ^^ with the severity level 3 407c, and ^^^ is the probability 403d that the dentalcondition is present in the data point ^^with the severity level 9407d. Here, the data point ^^406 is any of the N data points 302, i.e. ^ = [1, .. , ^].Accordingly, with reference to Figure 4, the output of the trained machine learning model 401for the data point ^^ is a vector of probabilities 408 ^^ = [0.10.20.30.4], wherein 0.1 is theprobability 403a for ^^ to be associated with the severity level 0 407a of the dental condition,0.2 is the probability 403b for ^^ to be associated with the severity level 1 407b of the dentalcondition, 0.3 is the probability 403c for ^^ to be associated with the severity level 3 407c ofthe dental condition, and 0.4 is the probability 403d for ^^to be associated with the severitylevel 9407d of the dental condition. Similarly, the output of the trained machine learning model401 for the data point ^^ is a vector of probabilities 408 ^^ = [0.10.40.30.2], wherein 0.1 isthe probability 403a for the data point ^^ to be associated with the severity level 0407a of thedental condition, 0.4 is the probability 403b for ^^to be associated with the severity level 1407b of the dental condition, 0.3 is the probability 403c for the data point ^^ to be associatedwith the severity level 3407c of the dental condition, and 0.2 is the probability 403d for the data point ^^to be associated with the severity level 9407d of the dental condition. Similararguments are held for all the remaining N-2 data points in 302. Therefore, the output of thetrained machine learning model 401 for the input N data points is the set of vectors 403: . . . ^^ → ^^ = [0.10.20.30.4]The maximum probability in the vector of probabilities 408 estimated for each data point ^^ 406is then determined in 404 for each data point 406, namely:^^^^^ = max P^ = max[0.10.20.30.4] = 0.4^^^^^ = max ^^ = max[0.10.40.30.2]= 0.4^^^^^ = max ^^ = max[0.10.20.40.3]= 0.4. . . ^^^^^ = max ^^ = max[0.10.40.30.2]= 0.4Accordingly, each data point ^^ 406 is assigned with a severity level 405 of the dental condition,wherein for each data point the severity level 405 is the severity level for which the maximumprobability 404 is determined for that data point. For example, the maximum probability 404^^^^^ = 0.4 determined for ^^ is the probability that the dental condition is present in the datapoint ^^ with a severity level 9. Accordingly, the severity level 405 = 9 is determined forthe data point ^^. Similarly, the maximum probability 404 ^^^^^ = 0.4 determined for ^^is theprobability that the dental condition is present in the data point ^^ with the severity level 1.Accordingly, the severity level 405 ^^ = 1 is determined for the data point ^^ in 305. Similararguments hold for all the N data points in 302, such that the severity levels 405:^^ = 9^^ = 1^^ = 3. . . ^^ = 9 are determined for the N data points 302 input in the trained machine learning model 401.The numbers 0, 1, 3, 9 are used in the example illustrated in Figure 9 just for illustrativepurposes of an embodiment of the trained machine learning model according to the present disclosure. In practice, the numbers defining the scale of severity of the dental condition may be chosen by a dental practitioner, or by any other dental expert, based on a clinical significanceof the specific dental condition and / or based on the clinical features of that condition. Forexample, the scale of severity for each dental condition may comprise four numbers corresponding to an absence of the dental condition, an initial stage of development of the dental condition, a moderate stage of development of the dental condition and a severe stage of development of the dental condition, respectively. In this example, the dental expert may infer that caries are more clinically significant than plaque, e.g. because caries may cause thedevelopment of other dental conditions. Thus, the dental expert may decide to define a scale ofseverity for caries containing larger numbers than the scale of severity for plaque, e.g. the scale of severity for caries may be chosen to comprise the numbers (0, 4, 9, 16) and the scale of severity for plaque may be chosen to comprise the numbers (0, 1, 2, 3), corresponding to absent,initial, moderate and severe levels, respectively. Accordingly, if a presence of initial caries isdetermined for a first data point and a presence of a severe plaque is determined for a second data point, the first data point will be assigned with the severity level 4 and the second datapoint with the severity level 3, such that a higher clinical significance will be assigned to thefirst data point than to the second data point. The scale of severity defined by the dental practitioner may be input to one or more processors, e.g. the processor(s) of a computer systemperforming the method described herein. The method described herein may further oralternatively comprise a predefined scale of severity for each dental condition. In another example, the numbers defining the scale of severity of a dental condition may befixed numbers raised to a variable exponent. Said exponent may be varied, e.g. by the dentalpractitioner or the dental expert, depending on the clinical significance of a specific dental condition, and / or depending on a desired gap between the severity levels of a dental condition. For example, the predefined numbers may be 0, 1, 2, 3, corresponding to an absent dental condition, an initial dental condition, a moderate dental condition, and a severe dentalcondition, respectively, and each of the predefined number may be raised by an exponent e,namely the severity scale may be defined as:0^1^2^3^In one example, the dental practitioner may want to have an indication of a specific dentalcondition at all its stages of development, i.e. they may want to minimize the gap between theabsent, initial, moderate and severe levels. Accordingly, they may choose the exponent to be^ = 1. In another example, the dental practitioner may be interested only in the severe stage ofa dental condition, i.e. they want to increase the gap between the absent, initial, moderate and severe levels of the scale of severity of the dental condition. Accordingly, they may choose theexponent to be ^ = 3, such that the severe condition will correspond to a severity of 27,whereas a moderate condition will be associated with a severity of 8, i.e. the severe conditionwill be quantified as much more relevant than the moderate one. In yet another example, theexponent e may be chosen to be different for different dental conditions, for example based on a clinical significance of each dental condition. For example, the dental practitioner may chooseto assign a higher clinical significance to caries than to plaque, and accordingly fix the exponent^ = 3 for caries and ^ = 1 for plaque. Thus, if a first data point is estimated to have severecaries and a second data point is estimated to have severe plaque, the former will have a largerseverity than the latter, i.e. the data points with severe caries will be given more significancefor inferring the health status of the dental site than the data points with severe plaque. It willbe appreciated by those skilled in the art that the above-mentioned examples are intended to beillustrative, and that the exponent may be chosen in an appropriate way depending on theclinical features of the inspected dental condition. Alternatively, the exponent ^ may be a setfeature in the method described herein, or it may be received as an input from a user of a system,e.g. a computer system, configured to perform the steps of the method disclosed herein.Figures 5a and 5b illustrate example reference diagrams 500 for describing an embodiment ofa workflow of the computer-implemented method disclosed herein. As a first step of 500, a 3Dmodel 301 of the dental site is obtained. The 3D model 301 comprises a triangulated mesh,wherein the triangulated mesh comprises ^ facets 501, and in particular it comprises a facet^^(502), a facet ^^ (503), a facet ^^ (504), and so forth, up to a facet ^^(505). The N facets501 collectively describe the surface of the dental site in 3D space, and therefore they compriseinformation about at least the 3D geometry of the dental site. In step 506, the presence of one or more dental conditions in the 3D model 301 of the dental site is estimated. In particular, thepresence of the dental conditions estimated in step 506 by inputting the N facets 501in a trained machine learning model 506a for dental condition number 1, and in a trained machine learning model 506b for dental condition number 2, and so forth, up to a trained machine learning model 506c for dental condition number K. The trained machine learning model for each dental condition outputs for each data point a severity level of that dental condition. For example, the trained machine learning model 506a for dental condition number 1 outputs the severity level ^^^507a of the dental condition 1 for the facet 502 ^^, the severity level ^^^ 507a of the dental condition 1 for the facet 503 ^^, theseverity level ^^^ 507a of the dental condition 1 for the facet 504 ^^, and so forth, up to theseverity level 507a of the dental condition 1 for the facet 505 ^^. Similarly, the trainedmachine learning model 506b for dental condition number 2 outputs the severity level ^^^507bof the dental condition 2 for the facet 502 ^^, and the severity level ^^^507b of dental condition2 for the facet 503 ^^, and the severity level ^^^507b of dental condition 2 for the facet 504 ^^, and so forth, up to the severity level 507b of dental condition 2 for the facet 505 ^^. Similararguments hold for the trained machine learning model for condition 3, the trained machinelearning model for condition 4, and so forth, up to the trained machine learning model 506c fordental condition K, which outputs the severity level ^^^ 507c of the dental condition K for thefacet 502 ^^, and the severity level ^^^ 507c of dental condition K for the facet 503 ^^, and theseverity level ^^^ 507c of dental condition K for the facet 504 ^^, and so forth, up to the severitylevel ^^^ 507c of dental condition K for the facet 505 ^^. In other words, the output of 506 is: The above-mentioned severity levels are numbers on a scale of severity of each dentalcondition, and said scale of severity comprises a minimum number which indicates an absence of the dental condition. The minimum number may be, for example, zero, such that a facet for which all the severity levels ^ ^ ^= 0, where ^ ∈ [1, ^] and ^ = [1, … , ^], is a facet for which nodental condition is present. If the minimum number on the scale of severity is different fromzero ^^^^ ≠ 0, then a facet for which all the severity levels ^^ ^= ^^^^, with ^ ∈ [1, ^] and^ = [1, … , ^], is a facet for which no dental condition is present.The one or more facets among the N facets 501 for which the presence of at least one dentalcondition is estimated in 506 define one or more inspection sites of the 3D model 301. For example, with reference to Figure 5a, an inspection site 508 of the 3D model 301 may comprisefour facets ^^(502), ^^(509), ^^(505), ^^(510) for which the presence of at least one dentalcondition was estimated in 506. Thus, at least one of the severity levels determined for the facet 502 is different than the minimum severity level of the scale of severityof a dental condition, and at least one of the severity levels determined for thefacet 509 is different than the minimum severity level of the scale of severity of a dental condition, and at least one of the severity levels[^^ ^ ^^, ^^, … , ^^] determined for the facet 505 isdifferent than the minimum severity level of the scale of severity of a dental condition, and atleast one of the severity levels determined for the facet 510 is different thanthe minimum severity level of the scale of severity of a dental condition.Figure 5b illustrates an example reference diagram of a method for determining an epicenterof the one or more dental conditions. First, for each facet in each inspection site, an aggregatedseverity level / heat score of the K dental conditions is determined in 511. In particular, the heatscore is determined for each of the facets ^^(502), ^^(509), ^^(505), ^^(510) which arecomprised in the inspection site 508 defined accordingly to Figure 5a. With reference to theexample illustrated in Figure 5b, the heat score of each facet in the inspection site 508 is determined by summing the severity levels of each dental condition 1,…,K determined for that ^facet. Therefore, the heat score 511a ^^ = + ⋯ + ^^is determined for the facet ^^,the heat score 511b ^ = ^^ + ^^ + ⋯ + ^^ is determined for the facet ^ , the heat score 511c^ ^ ^ ^ ^^ = ^^ + ^^ + ⋯ + ^^^ ^ ^ ^ is determined for the facet ^^ and the heat score 511d ^^ = + +⋯ + ^^ is determined for the facet ^ . In other words, four heat scores^( ) ( ) ( ) ^ 511^ , ^ 511^ , ^ 511^ , ^ (511^) are determined in 511 for the inspection site 508^ ^ ^ ^by aggregating, for each facet comprised in the inspection site 508, the severity level( )507^ , where ^ = 1,4,8, ^, i.e. the severity level determined for the dental conditions 1, 2,…,K for the facets 502, 509, 505 and 510 of the inspection site 508.In another example (not shown here), the heat score for each facet in the inspection site 508 may be determined by determining a weighted sum of the severity levels of each dentalcondition determined for that facet. In this example, the weights ^ , … , ^ may be determined^ ^^for the dental condition 1,…, K, respectively, such that the heat score ^ = ^ × ^+^ ^ ^^ ^ × + ⋯ + ^ × ^ is determined for the facet ^ , the heat score ^ = ^ ×^ ^ ^^ ^^^^ × ^ + ⋯ + ^ × ^ is determined for the facet ^ , the heat score ^ = ^ × ^+^ ^ ^ ^ ^^ ^ ^^ ^^^ × ^ + ⋯ + ^ × ^ is determined for the facet ^ and the heat score ^ = ^ × ^+^ ^ ^ ^ ^^ ^ ^^^ × + ⋯ + ^ × ^is determined for the facet ^ . The weights of each dental^ ^ may be defined by the dental practitioner, or by any dental expert performing the examination of the dental site, based on the clinical significance of each dental condition, such that the higher the clinical significance attributed to the dental condition, the larger the weight. Further or alternatively, the weights assigned to each dental condition may be predefined in the method described herein, e.g. they may be predefined and stored in a memory of the system performingthe method, and said predefined weights may be set in accordance with clinical standards. Inthis way, the heat score of each facet may quantify an overall health status of the inspectionsite, while giving more weight to those dental conditions which demand particular attention bythe dental practitioner and / or the dental expert. After determining the heat score of each facet comprised in the inspection site 508, theepicenter 512a of the dental conditions 1,…,K is determined in step 512 for the inspection site508 by determining a maximum heat score of the inspection site 508. For example, the heatscores = 10, ^^ = 25, ^^ = 9, ^^ = 25 may have been determined in 511 for the inspectionsite 508. Accordingly, the epicenter 512a of the inspection site 508 is determined to be:^ = max[^^, ^^, ^^, ^^] = ^^ = ^^The epicenter 512a of the dental conditions 1,…, K of the inspection site 508 comprises, in theexample illustrated in Figure 5b, two facets for which a same maximum heat score isdetermined, i.e. the facet 505 ^^ and the facet 509 ^^ . It will appear clear to those skilled inthe art that the epicenter of an inspection site may comprise more than two facets, or even a single facet, depending on the number of facets comprised in the inspection site for which thedetermined heat score is the maximum heat score of the inspection site. Furthermore, theepicenter 512a determined in 512 is a local epicenter, i.e. the epicenter 512a is locallydetermined for the inspection site 508 in the example illustrated in Figure 5b. However, inanother example (not illustrated here) the epicenter 512a may be a global epicenter, i.e. theepicenter 512a may be determined upon determining the global maximum heat score amongthe heat scores determined for all of the inspection sites defined for the 3D model 301. In otherwords, even though the example illustrated in Figure 5b refers to the epicenter of the inspectionsite 508, the method of the present disclosure may comprise determining a global epicenter, i.e.the global maximum heat score among all the heat scores determined for the data points of the3D model.Once the epicenter 512a is determined, in 513 the facets associated with the maximum heatscore 511b and 511d are associated with a location 514 of the epicenter 512a. The latter may be used by the dental practitioner and / or by any dental expert as an accurate guide to perform the examination of the dental site. For example, the examination may be started by looking at the location of the dental site associated with the epicenter of the dental conditions, as this is determined to be the location for which the overall health status is locally the most severe, and which may therefore need careful investigation to draw up a treatment plan and / or a prevention plan for the patient.Figure 6a illustrates a flowchart 600 of an embodiment of the computer implemented methodaccording to the present disclosure. During a scanning session 601, a dental site of a patient isscanned in step 602. The scanning may be performed by means of an intraoral scanner as the one illustrated as reference number 201 in Figure 2. Based on the data acquired during thescanning session 601, a 3D model of the scanned dental site is generated in step 603. Forexample, the intraoral scanner may acquire a number of 2D sub-scans for a number of viewingangles of the dental site, and thus the 3D model may be generated in step 603 by stitchingtogether sub-scans which at least partially overlap. In one example, the scanning session 601is carried out in a clinical visit of the patient, and therefore the 3D model is generated in step603 during the clinical visit. In another example, the scanning session was performed in aprevious clinical visit of the patient, and thus the 3D model was generated in said previous clinical visit and stored, e.g. on a memory of the computer system shown as reference number 205 in Figure 2. The 3D model of the dental site is obtained in step 604, e.g. by loading the 3D model from thememory on which it was stored during the scanning session 601. The 3D model comprises aplurality of data points 605 collectively describing a surface of the dental site in 3D space. Asalso mentioned above, said plurality of data points 605 may comprise a point cloud comprisinga plurality of points in 3D space, each point described by Cartesian coordinates (x, y, z), or theplurality of data points 605 may comprise a plurality of facets comprised in a triangulated mesh,and / or a plurality of vertices comprised in a triangulated mesh, or any other data suitable to describe the surface of the dental site in 3D space. In step 606, a presence of one or more dental conditions in the 3D model obtained in step 604is estimated. The one or more dental conditions may comprise one or more of caries, toothwear, gum recession, tooth crack, plaque, and / or gum inflammation. However, the method disclosed herein is not limited to the above mentioned dental conditions, and any other dental condition of interest for inferring a health condition of the dental site may be comprised in the one or more dental conditions. Step 606 comprises processing the plurality of data points 605 using a trained machine learning model for each of the one or more dental conditions. Accordingly, the plurality of data points 605 is input in a trained machine learning model for caries 607, and in a trained machine learning model for tooth wear 608, and in a trained machine learning model for gum recession609, and in a trained machine learning model for tooth crack 610, and in a trained machinelearning model for plaque 611, and in a trained machine learning model for gum inflammation612. Each of these trained machine learning models may be a pre-trained machine learningmodel which has been trained with a training data set comprising data points similar to theplurality of data point 605, wherein each data point in the training data set may comprisefeatures relative to that data point in the training data set. For example, the features may be thepresence or absence of a specific dental condition in a specific data point of the training dataset, such that each data point in the training data set of the trained machine learning model forthe specific dental condition comprises information about the presence or absence of the dentalcondition in that data point. Further, the features may be a severity label, and / or size, and / orarea and so forth of each dental condition. For example, each point in the training machinelearning model for caries 607 may comprise information about a presence or an absence ofcaries and / or a severity level of caries in that point, and each data point in the training data setof the trained machine learning model for tooth wear 608 may comprise information about apresence or absence of tooth wear and / or about a severity level of tooth wear in that point, andeach data point in the training data set for the machine learning model for gum recession 609may comprise information about a presence or absence of gum recession and / or about aseverity level of gum recession in that point, and so forth for each of the one or more dentalconditions. The presence of each dental condition may be determined using methods differentthan the one illustrated in the example of Figure 6a (not illustrated here), e.g. mathematicaland / or statistical methods using color information comprised in the data relative to the dental site acquired during the scanning session 601. The trained machine learning model for each dental condition used in step 606 to process theplurality of data points 605 outputs in step 613, for each of the plurality of data points 605, aseverity level of each dental condition. Therefore, for each of the plurality of data points 605,a severity level of caries 614 is determined, and a severity level of tooth wear 615 is determined,and a severity level of gum recession 616 is determined, and a severity level of tooth crack 617 is determined, and a severity level of plaque 618 is determined, and a severity level of gum inflammation 619 is determined. The severity level of each dental condition may be a numberon a predefined scale of severity of that dental condition, the number quantifying the severityof that dental condition. For example, the scale of severity of caries may comprise the numbers0, 1, 3 and 9, wherein 0 indicates the absence of caries, 1 indicates initial caries, 3 indicates moderate caries and 9 indicates severe caries. The scale of severity of each dental conditionmay be chosen by a dental practitioner performing an examination of the dental site based onclinical features of that dental condition and / or based on a clinical significance of that dentalcondition. For example, the dental practitioner may decide and / or need to assign more clinicalsignificance to caries than plaque, and accordingly the numbers on the scale of severity of caries corresponding to initial, moderate and severe caries may be chosen to be larger than thenumbers on the scale of severity of plaque corresponding to initial, moderate and severe plaque,respectively. In general, the scale of severity of each dental condition comprises a minimumseverity level which is indicative of an absence of the dental condition. The one or more datapoints in 605 for which at least one among the determined severity levels 614, 615, 616, 617,618 and 619 is different than the minimum severity level of the scale of severity of the respective dental condition define one or more inspection sites of the 3D model. Therefore, each inspection site of the 3D model may be understood as a region of the 3D model for which the presence of at least one dental condition is determined.For each data point comprised in each inspection site defined in step 613, an aggregatedseverity level / heat score is determined in step 620, thereby determining one or more heat scoresfor each inspection site. The heat score of each data point in step 620 may comprise the sum621 of the severity levels 614, 615, 616, 617, 618 and 619 determined for that data point, orthe heat score of each data point may comprise a weighted sum 622 of the severity levels 614,615, 616, 617, and 618 and 619 determined for that data point. Determining the weighted sum622 of the severity levels of each dental conditions determined for each data point maycomprise determining a weight for each of the dental conditions, wherein the weight may be, for example, chosen by a dental practitioner based on a clinical significance of each dentalcondition. For example, the dental practitioner may decide that caries have the largest clinicalsignificance among the one or more dental conditions, and accordingly they may assign tocaries the largest weight, and the dental practitioner may decide that gum recession has thesecond largest clinical significance after caries, and accordingly they may assign to gum recession the second to largest weight, and so forth. As in step 620 one heat score is determined for each data point in each defined inspection site, one or more heat scores are determined for each defined inspection site, and the number of heat scores determined for each inspection site depends on the number of data points comprised in that inspection site. In step 623, an epicenter of the one or more dental conditions is determined for each inspection site by determining a maximum heat score of each inspection site. The epicenter may be understood as a local epicenter, i.e. it may be determined by determining a local maximum foreach inspection site. The epicenter of each inspection site may comprise one or more data pointsof the inspection site, and it indicates a region of the inspection site for which the healthcondition is determined to be locally the most severe, and which may therefore necessitate acareful inspection by the dental practitioner. In another example (not shown here), oneepicenter of the dental conditions is determined by determining a global maximum heat score among the one or more heat scores of all of the inspection sites of the 3D model, i.e. a global epicenter is determined. The global epicenter may comprise one or more data points of the 3D model, and it indicates a region of the inspection site for which the health condition is determined to be globally the most severe, and which may therefore necessitate the highest attention by the dental practitioner.Figure 6b illustrates a flowchart of an example of additional steps of the method illustrated inFigure 6a. In step 624, the one or more data points of the maximum aggregated severity levelare associated with a location of the epicenter on the dental site. Similarly, the one or more heatscores determined for each inspection site in step 620 are associated with the plurality of datapoint 605 of the 3D model in step 625. Associating the one or more heat scores with the plurality of data points 605 may comprise mapping each heat score to one data point of the 3Dmodel, such that a graphical representation of the one or more heat scores may be generated.Therefore, in step 626 the 3D model of the dental site is rendered with a graphicalrepresentation of the one or more heat scores, for example on a display of the computer systemas the one shown as reference number 206 in Figure 2.The one or more heat scores may be displayed as an indication, e.g. a color or a pattern, on therendering of the 3D model. Accordingly, the epicenter of each inspection site may be identifiedby the dental practitioner as the region in the graphical representation of the inspection site displayed with the most intense color or the densest pattern, and the intensity of the color orthe density of the pattern may decrease in each inspection site with a distance from the epicenterof each inspection site. Here, the distance may be intended as a Euclidean distance, or ageodesic distance, or a facet distance, or any other way suitable to measure a distance betweenthe data points comprised in the 3D model. The intensity of the color or the density of the pattern may be intended as the graphical representation of the heat score of each data point in each inspection site, such that the heat score of each inspection site is maximum at the epicenterof the inspection site, and it falls off when moving away from the epicenter, reaching aminimum value when reaching a border of the epicenter. Therefore, by looking at the rendering of the 3D model on the display of the computer system, the dental practitioner may have an accurate indication of the regions of the dental site for which the largest heat score was locally determined, and which need an accurate examination due to their determined health status. Figure 6c illustrates a flowchart of an example of additional steps to the method illustrated inFigure 6a, alternative to the additional steps illustrated in Figure 6b. In step 628, a smoothingof the one or more heat scores determined for each inspection site is performed, therebyremoving potential discontinuities between the one or more heat scores determined forneighboring points in each inspection site. The smoothing step 628 generates one or moresmoothed heat scores 629 for each inspection site.In one embodiment, the smoothing step 628 may comprise any global transformation of theheat scores determined for all of the one or more inspection sites of the 3D model. In otherwords, the smoothing step 628 may comprise a function depending on the epicenters of all ofthe inspection sites of the 3D model. For example, given a 3D model in which a number n ofinspection sites has been estimated, the smoothing step 628 may comprise a function oftransforming the heat scores in such a way that, given a data point ^^ of any of the n inspectionsite its smoothed heat score 629 is: wherein ^^ is the heat score of the epicenter of the j-th (wherein ^ = [1, … , ^]) inspection siteof the 3D model and is the distance between the data point ^^and the epicenter of the j-thinspection site. In the present example, the smoothed heat scores 629 depend on a globalepicenter of the 3D model, i.e. on a global maximum heat score. In general, any function ofglobally transforming all the heat scores determined for the 3D model may be comprised in the smoothing step 628. In another embodiment, the smoothing step 628 may be performed locally for each inspectionsite. In one example, the smoothing step 628 may comprise any function of transforming the heat scores of each inspection site such that the heat scores within the inspection siteexponentially fall off with a distance from the epicenter of that inspection site. In this example,given an inspection site for which the maximum heat score ^^^^is determined, the smoothing step may comprise a function of transforming the heat scores of the inspection site such thatthe heat score of a data point in the inspection site at a distance ^ from the epicenter has a heatscore of ^(^) = ^^^^^^^×^ , where ^ is a constant controlling the rate at which the heat scoresdecrease with the distance. In another example, the smoothing step 628 may comprise anyfunction of transforming the heat scores of each inspection site such that the heat scores of the inspection site linearly fall off with a distance from the epicenter of the inspection site. In this example, given an inspection site for which the maximum heat score ^^^^is determined, thesmoothing step 628 may comprise a function of transforming the heat scores of the inspectionsite such that the heat score of a data point in the inspection site at a distance ^ from theepicenter has a heat score of ^(^) = ^^^^ − ^ × ^ , wherein ^ is a constant which controls therate at which the heat scores decrease with the distance. Here, the distance may be a Euclidean distance, a geodesic distance, a facet distance, or any other suitable way to measure a distancebetween two data points. In yet another example, the smoothing step 628 may comprise anyfunction of transforming the heat scores of each inspection site such that the heat scores of theinspection site are distributed according to a Gaussian distribution centered at the value of theheat score of the epicenter of that inspection site. In this example, given an inspection site forwhich the maximum heat score ^^^^ is determined, the smoothing step 628 may comprise afunction of transforming the heat scores of the inspection site such that the heat score of a datapoint ^^ in the inspection site is: wherein ^^^^^^^^^ is the heat score initially assigned to the data point ^^, i.e. the heat scoredetermined for the data point before the smoothing method 628 is performed, ^ =^∑^^^ ^^ ^ − ^^^^^^^^^^ ^^^^^^^ is the variance, and ^ is the total number of data point in theinspection site. An example of a smoothing method will be provided in the detailed descriptionof Figure 7.The smoothed heat scores 629 are normalized in step 630, such that one or more normalizedand smoothed heat scores 631 are generated for each inspection site. Normalizing the one ormore heat scores may comprise determining a global maximum heat score of the 3D model, namely determining the heat score which is the largest among all the one or more heat scoresdetermined for all the one or more inspection sites of the 3D model. Said global maximum heatscore is the global epicenter of the one or more dental conditions of the 3D model, i.e. theregion of the 3D model for which the aggregated severity of the one or more dental conditionsis, globally, the largest. Accordingly, the global epicenter of the 3D model corresponds to theregion of the 3D model for which the overall health status is determined to be, globally, theworst. In other words, the global epicenter of the 3D model corresponds to the region of the3D model which necessitates the highest attention. Once the global maximum heat score is determined, the one or more heat scores determined for each inspection site may be normalized to said global maximum heat score, whereby the normalized global maximum heat score isequal to 1. The normalization step also determines an order of clinical significance of theepicenters of the one or more inspection sites of the 3D model. For example, three epicentersmay have been determined for three different inspection sites of the 3D model, with a smoothed heat score of 5, 10, and 3, respectively. In this example, 10 is the global maximum heat score of the 3D model, and, accordingly, the normalized smoothed heat scores of the epicenters of the 3D model are 0.5, 1, and 0.3, respectively. It appears clear that the global epicenter of the 3D model, i.e. the epicenter with a normalized smoothed heat score equal to 1, is the region with the largest clinical significance, followed by the epicenter with a normalized smoothed heat score of 0.5, and finally by the epicenter with a normalized smoothed heat score of 0.3. In step 632, the one or more data points of the global maximum aggregated severity level are associated with a location of the epicenter on the 3D model, and, similarly, the determined one or more normalized and smoothed heat scores 631 are associated with the plurality of data points 605 of the 3D model. The mapping performed in step 632 enables to generate a graphical representation of the one or more normalized and smoothed heat scores, such that a rendering of the 3D model with said graphical representation is displayed on a GUI in step 633. The graphical representation of the one or more normalized and smoothed heat scores may be an indication, such that each inspection site is displayed with said indication on the rendering ofthe 3D model. The indication may be a color or a pattern, such that for each inspection site theintensity of the color or the density of the pattern decreases when moving away from theepicenter determined for that inspection site. For example, the intensity of the color or density of the pattern may decrease with a distance from the epicenter. As also mentioned above, the distance may be intended as a Euclidean distance, or as a geodesic distance, or as a facet distance, or as any other suitable way to measure a distance between the data points comprisedin the 3D model. Importantly, due to the smoothing step 628 and the normalization step 629,the intensity of the color or the density of the pattern smoothly changes within each inspectionsite, and possible discontinuities in the color or pattern of each inspection site due todiscontinuous changes of the one or more heat scores between neighboring points of theinspection site are removed by the smoothing. Therefore, the regions of the 3D model whichnecessitate careful examination can be clearly visualized in the rendering of the 3D model bythe dental practitioner. The intensity of the color or density of the pattern is maximum at the determined location ofthe global epicenter of the 3D model and progressively decreases based on the heat score ofthe epicenter of each inspection site relative to the global epicenter determined for the 3Dmodel. Therefore, the determined order of clinical significance of the epicenters of the one ormore inspection sites is visualized as an order of the color intensity or of the pattern density inthe graphical representation of the one or more heat scores in the rendering of the 3D model. Accordingly, the dental practitioner may easily and quickly identify the region of the dental site for which the health status is determined to be globally the most severe, as this region isdisplayed as the most intense or as the densest on the rendering of the 3D model. Based on theintensity of the color or by the density of the pattern, the dental practitioner may perform theexamination of the dental site by starting from the region with the largest intensity or densestpattern, and then proceeding to regions with progressively lower intensity or density.Figures 7a, 7b, 7c, 7d, 7e, 7f and7g illustrate a reference diagram 700 of an embodiment of acomputer-implemented smoothing method according to the present disclosure. Illustrated inFigure 7a is an example of a processing outcome of the steps of the method according to thepresent disclosure. A plurality of facets is provided, wherein the plurality of facets is comprisedin a triangulated mesh which may describe a surface of a dental object in 3D space. As a resultof the processing, one or more of the plurality of facets have been assigned with a heat scorewhich may be indicative of an aggregated severity level of one or more dental conditions, andwhich may have been determined according to the method disclosed herein. Accordingly, theoutcome of the processing as illustrated in Figure 7a may have been obtained by inputting theplurality of facets in a trained machine learning model for each of the one or more dentalconditions and aggregating for each facet the severity level output by the trained machinedlearning model for each dental condition for that facet. The one or more facets for which theheat score is different from zero, i.e. for which the presence of at least one dental condition is estimated, define one or more inspection sites of the 3D model of the dental site. With referenceto Figure 7a, four inspection sites 701, 701’, 701’’, 701’’’ have been defined for the trianglemesh. Each of the inspection sites comprises a local maximum heat score which is a local epicenter of the one or more dental conditions of that inspection site. For example, the epicenter of the inspection site 701 comprises the facets 702 with the heat score 20, the epicenter of theinspection site 701’ comprises the facets with the heat score 15, the inspection site 701’’comprises the facets with the heat score 4 and the inspection site 701’’’ comprises the facetswith the heat score 11Figure 7a further illustrates a first step of the smoothing method 700, in which a globalmaximum heat score is determined among the heat scores of the inspection sites defined forthe plurality of facets. With reference to Figure 7a, the global maximum heat score among theheat scores of the inspection sites 701, 701’, 701’’ and 701’’’ is 20. Accordingly, a globalmaximum epicenter is determined for the plurality of facets, wherein the global maximum epicenter comprises the facets 702 assigned with the global maximum heat score 20.Figure 7b illustrates a second step of the method 700, following the step illustrated in Figure7a. The second step comprises subtracting the global maximum heat score determined in thefirst step of the method from the heat score of each facet of each inspection site. Accordingly,the heat score of each facet belonging to any of the inspection sites 701, 701’, 701’’, 701’’’ isreduced by 20. Thus, the facets 702 originally assigned with the global maximum heat score20 are assigned, in this second step, with a heat score of 0, the facets 703 originally assignedwith a heat score equal to 17 are assigned in this second step with a heat score equal to -3, and so forth. Figure 7c illustrates a third step of the method 700 following the step illustrated in Figure 7b. The third step comprises determining the neighboring facets of the facets with a heat scoreequal to 0 which do not belong to any inspection site, i.e. which originally were not assignedwith any heat score in Figure 7a. Said neighboring facets are assigned with a score of -1. Forexample, given the facet 702’ in Figure 7c, its neighboring facets which do not belong to anyinspection site are the facets 704. Accordingly, after the third step of the method 700 isperformed the inspection site 701 comprises more facets than it originally did, i.e. than it didin Figure 7a. Furthermore, the third step illustrated in Figure 7b comprises determining theneighboring facets of the facets with a heat score equal to 0 which instead belong to at leastone inspection site, and assigning to these facets a score of -1 only if, after the second step ofthe method (i.e., in Figure 7b), their score is smaller than -1. For example, given the facet 702’’,the neighboring facet 703 in the inspection site 701 was assigned in step two of the methodwith a heat score of -3, as illustrated in Figure 7b. Therefore, with reference to Figure 7c, inthe third step of the method the score -1 is assigned to the facet 703.Figure 7d illustrates the fourth step of the method 700 following the step illustrated in Figure7c. The fourth step comprises adding a score of +1 to the score of each facet, wherein now thescore is the score determined in the third step of the method (i.e., in the step illustrated in Figure7c). Thus, referring to Figure 7d, the facets 702 are assigned with a heat score of +1, the facets704 are assigned with a heat score of 0, the facet 703 is assigned with a heat score of zero, andso forth.Steps three and four of the method (i.e., the steps illustrated in Figure 7c and 7d, respectively)are iteratively repeated, such that an increasing number of facets is progressively assigned witha score and a size of the inspection sites 701, 701’, 701’’ and 701’’’ progressively increases. Inone example, the method may be iteratively repeated until all of the plurality of facets in the3D model are assigned with a score. In another example, the iterations of the method may berepeated until a predefined heat score is assigned to at least one facet An example would be apredefined number of 255, in which case the smoothing step is iteratively repeated until at leastone facet is assigned with a heat score of 255, i.e. it is repeated until the at least one facet ofthe global epicenter is assigned with the heat score of 255, and one or more facets are assignedwith the minimum heat score of 0. The numbers 0 and 255 are the minimum and maximum values, respectively, of a color intensity in an 8-bit red, green, blue (RGB) system, thus in this example each facet assigned with a heat score after the smoothing step maps to a specific colorintensity in the RGB system. This choice may facilitate the generation of a graphicalrepresentation of the one or more heat scores. Alternatively, the predefined number may be chosen to be a predefined percentage of the total number of facets in the plurality of facets, e.g. 30% of the total number of facets, or 60% of the total number of facets, and so on.The iterative results of the smoothing method 700 are illustrated in Figures 7e, 7f and 7gwherein a value of the heat score is illustrated with a dot, such that the larger the heat score,the larger the size of the dot. In particular, Figure 7e illustrates a result of a first iteration of the method, i.e. the iteration illustrated in Figure 7d, wherein the facets 702 corresponding to the global maximum heat score determined in Figure 7a are assigned with a heat score equal to 1. Accordingly, after the first iteration only the facets 702 comprised in the global epicenter ofthe 3D model are assigned with a score and thus shown in Figure 7e. Figure 7f illustrates a result of a second iteration of the method, i.e. the iteration following the one illustrated in Figure 7d. This iteration assigns to the facets 702 a score equal to +2, and tothe facets 704 a score equal to +1. Accordingly, the facets 702 are illustrated in Figure 7f withdots larger than the ones in Figure 7d, and the neighboring facets 704 of 702 appear in Figure7f, as they are assigned with a positive score.The iterative method may be performed up to the iteration shown in Figure 7g wherein thescore of the global maximum heat score of the inspection sites is illustrated with the largestdots among the heat scores of the remaining facets. After a number of iterations, the inspectionsite 701’ is also visible in Figure 7g. It appears clear from Figure 7e, 7f and 7g that, overall, the smoothing method spreads the size of the inspection site 701 and ensures that the heat score smoothly changes when moving through neighboring facets. After the smoothing method 700 is performed, the heat scores of each facet in each inspection site facets falls off with a distance from the epicenter of the corresponding inspection site, i.e. it falls off with a distance from the local maximum heat score. In one example, said distance may be a Euclidean distance. In another example, the distance from the epicenter of each inspection site may be a geodesic distance. In yet another example, the distance from theepicenter may be a facet distance. The last example may be particularly relevant in the case thecomputer-implemented method disclosed herein is performed upon obtaining a 3D model of the dental site which comprises a plurality of facets, such as in the case of method 700. Thefacet distance between a facet of an inspection site and the epicenter of that inspection site maybe obtained by calculating a minimum number of steps needed to move from the epicenter to the facet, while only moving through neighboring facets. In general, the measure of the distancemay be appropriately chosen based on the plurality of data points comprised in the 3D model,and the choice of this measure will appear clear to those skilled in the art, e.g. a Euclidean distance measure is more appropriate to measure the distance between points in 3D space of a point cloud.In the example illustrated in Figure 7g, given a facet at the distance ^ from the epicenter 702the heat score ^ of that facet after the smoothing method has been performed may bedetermined as: ^(^) = ^^^^ − ^,wherein ^^^^ is the maximum heat score of the inspection site, i.e. the heat score of theepicenter of the inspection site. It will be appreciated that in the illustrated example theepicenter is actually the global epicenter of the 3D model, however the above mentioneddiscussion may hold for each inspection site of the 3D model. For example, if after thesmoothing the facets 702 comprised in the epicenter of the inspection site are assigned with theheat score ^^^^ = 10, the smoothed heat score of the immediate neighboring facets 704 whichare at a facet distance of ^ = 1 from the epicenter 702 is ^(1) = 9, and the smoothed heatscore of the immediate neighboring facets 705 of the facets 704 which are at a facet distanceof ^ = 2 from 702 is ^(2) = 8 and so forth.A gap between the smoothed heat scores of facets belonging to the inspection site may beincreased by elevating the heat score of each facet of the inspection site to a same predefinedexponent ^, e.g. ^ = 2 , such that ^^^^ = 100, ^(1) = 81, ^(2) = 64 and so forth. The effectof this exponent is to control the rate at which the heat scores fall off when moving away fromthe epicenter of the inspection site, i.e. the larger the exponent, the larger the gap between theheat scores of neighboring facets and the larger the rate at which the heat score decreases whenmoving from the epicenter of the epicenter of the inspection site to a boundary of the inspectionsite. Controlling the rate at which the heat score falls off in an inspection site may beadvantageous, for example, for generating a graphical representation of the heat scores andvisualizing a rendering of the 3D model with said graphical representation. In this case, theexponent may be used to control the intensity of the color or the density of the pattern of thegraphical representation of the heat scores, such that the intensity of the color or the density ofthe pattern of each inspection site falls off quickly / slowly when moving away from theepicenter of each inspection site. Advantageously, increasing the exponent ^ enables toemphasize the graphical representation of the epicenter of each inspection site in the renderingof the 3D model, such that a dental practitioner may easily and quickly identify the regions ofthe 3D model which necessitate careful examination due to their health conditions.Figures 8a and 8b illustrate a flowchart 800 of embodiments of the computer-implementedmethod according to the present disclosure. In step 801, a first 3D model of a dental site comprising a first set of data points is obtained, wherein the first set of data points describes asurface of the dental site at a first time. The first 3D model may be obtained by acquiring firstintraoral scan data of the dental site by means of an intraoral scanner, for example during a firstclinical examination of the patient carried out when the patient visits the dental clinic at thefirst time. Therefore, the first time may be the time when said first clinical examination is carried out. The first intraoral scan data may be used to generate the first 3D model of the dentalsite, and the generated first 3D model may be stored on a storage unit, e.g. the storage unit of a computer system on which the computer-implemented method disclosed herein is carried out and / or a remote storage unit such as a cloud storage. The first set of data points may comprisea plurality of points in 3D space forming a point cloud, wherein each point is described byCartesian coordinates (x,y,z). The first set of data points may alternatively comprise a plurality of facets of a 3D mesh, wherein each facet may be a triangular facet in the case of a triangular mesh, or each facet may be any other polygon in the case of a polygonal mesh suitable todescribe the surface of the dental site in 3D space at the first time. The first set of data pointsmay alternatively comprise a plurality of vertices of a 3D mesh, wherein each vertex may be a vertex of a triangle in the case of a triangular mesh, or each vertex may be a vertex of any otherpolygon in the case of a polygonal mesh suitable to describe the 3D surface of the dental siteat the first time. In general, the first set of data points may comprise any other data format which is suitable to describe the surface of the dental site in 3D space. In step 802, a second 3D model of the dental site comprising a second set of data points isobtained at a second time, wherein the second set of data points describes the surface of thedental site at the second time. The second 3D model of the dental site may be obtained by acquiring intraoral scan data of the dental site by means of an intraoral scanner, for exampleduring a second clinical examination of the patient carried out when the patient visits the dentalclinic at the second time. The second time may be subsequent to the first time. For example,the second time may be a most recent time when the patient visits the dental clinic, and the first time may be the latest time when the patient visited the dental clinic before the most recent time. Alternatively, the first time may be any time preceding the most recent clinical examination of the patient. Similarly to the first set of data points, the second set of data points may comprise a plurality of points in 3D space forming a point cloud, wherein each point is described by Cartesian coordinates (x,y,z). The second set of data points may alternatively comprise a plurality of facets of a 3D mesh, wherein each facet may be a triangular facet in the case of a triangular mesh, or each facet may be any other polygon in the case of a polygonalmesh suitable to describe the surface of the dental site in 3D space at the second time. Thesecond set of data points may alternatively comprise a plurality of vertices of a 3D mesh,wherein each vertex may be a vertex of a triangle in the case of a triangular mesh, or each vertex may be a vertex of any other polygon in the case of a polygonal mesh suitable to describethe 3D surface of the dental site at the second time. In general, the second set of data pointsmay comprise any other data format which is suitable to describe the surface of the dental site in 3D space. Preferably, the first set of data points and the second set of data points comprise a same type of data points, i.e. they both comprise a plurality of facets, or a plurality of points in 3D space and so forth. In step 803, the first 3D model is compared with the second 3D model to determine one or more differences therebetween. Comparing the first 3D model with the second 3D model maycomprise determining a spatial relationship between at least a part of the first 3D model and atleast a part of the second 3D model. Several methods may be used to determine the spatialrelationship between the at least a part of the first 3D model and the at least a part of the second3D model. For example, the first 3D model and the second 3D model may be aligned in 3Dspace, such that the two 3D models are represented in a common 3D frame. Then, for each data point in the first set of data point, a closest data point in the second set of data points may be determined, such that spatially corresponding data points of the first 3D model and of the second 3D model may be determined. In another example, which may be used when the first set of data points comprises a first mesh and the second set of data points comprises a second mesh, for each vertex in the first mesh a ray may be shot in a direction normal to the vertex. The shot ray may intersect the second mesh in an intersection point, and the vertex of the secondmesh closest to the intersection point may be considered to be the vertex of the second meshcorresponding to the vertex of the first mesh. Therefore, the spatial relationship between data points of the first 3D model and of the second 3D model may be determined. In yet another example, first augmentation data comprising segmented tooth surface data and segmented gingiva surface data relative to the first 3D model may be obtained, and second augmentation data comprising segmented tooth surface data and segmented gingiva surface data relative to the second 3D model may be obtained. The first segmentation data and the second segmentation data may be obtained upon processing the first 3D model and the second 3D model by using a trained machine learning model which is configured to receive as input data points similar to the first set of data points and second set of data points, and to assign each data point to a segment, wherein the segment is one of tooth or gingiva. Furthermore, thetrained machine learning model may further output a tooth number for each data point whichis determined to belong to a tooth surface, wherein the tooth number may be assigned according to the International Tooth Numbering System or any other tooth numbering system. Therefore,determining the spatial relationship between the at least a part of the first 3D model and the atleast a part of the second 3D model may further comprise using the first augmentation data and the second augmentation data to determine the one or more data points in the first set of data points and the one or more data points in the second set of data points corresponding to a same tooth or a same part of the gingiva. In yet another example, a user of the computer-implemented method disclosed herein may manually select corresponding regions of the first 3D model andof the second 3D model. In yet another example, a combination of the methods discussed abovemay be used to determine the spatial relationship between the at least a part of the first 3D model and the at least a part of the second 3D model. In general, any method suitable todetermine a spatial correspondence in 3D space of at least a part of the first set of data pointswith at least a part of the second set of data points may be used according to the method of the present disclosure. Once the spatial relationship between the at least a part of the first 3D model and the at least a part of the second 3D model is determined, the one or more differences between the first 3Dmodel and the second 3D model are determined in step 803. In one example, determining theone or more differences between the first 3D model and the second 3D model may comprise determining one or more differences between second color data comprised in the second set of data points and first color data comprised in the first set of data points. More particularly, determining the one or more differences between the second color data and the first color data may comprise computing one or more difference values, each difference value quantifying a difference between a color coordinate of a data point in the second set of data points and a color coordinate of a corresponding data point in the first set of data points. For example, let ^^be a data point in the first set of data points, wherein comprises a color coordinate ^^^=(^^, ^^, ^^) describing the color of the surface of the dental site at a location corresponding tothe point ^^in the Red, Green, Blue (RGB) space. Similarly, let ^^be a data point in the secondset of data points which corresponds to the point ^^ in 3D space, wherein ^^ comprises a colorcoordinate ^^^ = (^^, ^^, ^^) describing the color of the surface of the dental site at a locationcorresponding to the point ^^in the RGB space. Accordingly, the difference between the first color coordinate and the second color coordinate may be determined as: ^− ^ ^ − ^ ^ − ^^ = ^^ − ^ = ( ^ ^,^ ^^^,^ ^^ ^^) ^^^^^^In other words, the difference value determined for a point in 3D space is indicative of apercentage of change of a given feature such as geometry and / or color at that point betweenthe first time and the second time. The above mentioned steps may be repeated for each datapoint in the second set of data points and for each corresponding data point in the first set ofdata points, therefore determining one or more difference values similar to ^^^for each coupleof corresponding data points. Each difference value may then be compared with a thresholdvalue, wherein responsive to determining that a difference value among the one or more difference values is equal or greater than the threshold value the difference value is classified as clinically relevant, and wherein responsive to determining that a difference value among the one or more difference value is smaller than the threshold value the difference value isclassified as not clinically relevant. In one example, the absolute value of the difference value^^^ may be compared with a threshold ^, and if the condition |^^^| ≥ ^ is satisfied then thedifference value is classified as clinically relevant. Using the absolute value of each differencevalue ensures that both positive and negative changes may be classified as clinically relevant, i.e. both an increase and a decrease in the color coordinate may be classified as clinically relevant. In another example, each difference value may be directly compared with thethreshold value, and if the condition ^^^ ≥ ^ is fulfilled then the difference value is classifiedas clinically relevant. This ensures that only positive difference values may be classified as clinically relevant, i.e. only an increase in color coordinates may be classified as clinicallyrelevant. In general, if the difference value quantifying a difference in the color data at a givenpoint in 3D space of the dental site between the first time and second time satisfies the threshold criterion, then the difference in color at the given point is classified as clinically relevant. The same steps are repeated for each couple of corresponding data points between the first and second sets of data points. The criterion for which a change in color is classified as clinically relevant may be based on health standards defined by the American Dental Association, or by any other global / international / national health association.In a further or alternative example, determining the one or more differences in step 803 maycomprise determining a presence of at least a part of the dental site in the second 3D model which is not present in the first 3D model. For example, a dental crown, an inlay, an onlay, adental bridge, and / or any other dental restoration may have been placed in the patient’s dentalsite at a time between the first time and the second time. Therefore, the at least a part of the dental site in the second 3D model which is not present in the first 3D model may comprise a dental prosthesis. In a further or alternative example, the patient may have had a dental filling in between the first time and second time. Therefore, the at least a part of the dental site in thesecond 3D model which is not present in the first 3D model may comprise the dental filling.Determining the presence of the at least a part of the dental site in the second 3D model whichis not present in the first 3D model may comprise comparing the 3D geometry described by thefirst set of data points with the 3D geometry described by the second set of data points. The comparison may be performed by using mathematical algorithms or by using a trained machine learning model configured to obtain as input the first 3D model and the second 3D model and configured to output one or more parts of the second 3D model which are not present in the first 3D model. The trained machine learning model may be, for example, a ConvolutionalNeural Network (CNN) which has been trained by using a training data set comprising couplesof 3D models of a same dental object, wherein for each couple each of the two 3D models hasbeen obtained at different times. The training data set may further comprise, for each couple of3D models, an indication of one or more parts which are present in one of the two 3D modelsbut which are absent in the corresponding 3D model. The mathematical algorithms maycomprise, for example, determining that a spatial relationship cannot be determined between one or more data points of the second 3D model and one or more data points of the first 3D model, i.e. that a region in 3D space of the second 3D model does not correspond to any regionin 3D space of the first 3D model. In yet another further or alternative example, step 803 maycomprise determining other differences than the one discussed above between the first 3D model and the second 3D model, and further examples will be provided with reference to Figure 8b. In step 804, one or more heat scores quantifying one or more clinical changes in the dental site between the first time and the second time are determined by determining a heat score for eachof the one or more differences. Each of the one or more heat scores may be a number on anumerical scale comprising a minimum heat score and a maximum heat score. For example, each number on the numerical scale may be associated with a range of difference values. For example, the numerical scale may be [0, 1, 2, 3], wherein 0 is the minimum heat score and 3 is the maximum heat score. For example, the minimum heat score 0 may be associated withdifference values smaller than 10%, and it may be indicative of a not clinically relevant change.The heat score 1 may be associated with difference values within the range [10%, 40%], and it may be indicative of an initial stage of clinical relevance of the change. The heat score 2 may be associated with difference values within the range [40%, 70%], and it may be indicative of a medium stage of clinical relevance of the change. Finally, the heat score 3 may be associatedwith difference values larger than 70%, which may be indicative of a critical or severerelevance of the change. It will be appreciated that the example above has illustrative purpose only, and that the ranges may be chosen according to the American Dental Association oral health standards, or according to any other global or international association. For example, a change in the color between two clinical visits performed at a distance of less than 6 months which is larger than 30% may be considered of medium clinical relevance by one or more dental health standards, e.g. because such a change is often observed with a decay of the tooth. Therefore, the ranges may be adapted to clinical considerations based on realistic data collected by one or more global / international dental organizations for a population of patients. Furthermore, the ranges may vary depending on one or more of an age of the patient, theethnicity of the patient, the gender of the patient, one or more dental conditions estimated inthe patient’s dental site and so forth. Furthermore, the heat score may be chosen among a rangecomprising more than 4 heat scores, such as 5 or 6 heat scores, or less heat scores, such as 3 or 2 heat scores. For example, the score 0 may be assigned for difference values below 10% and the score 1 may be assigned for difference values above 10%. This may be advantageous, for example, when the dental practitioner is not interested in the severity of the clinical change, but when they are interested in any change which has occurred in the patient’s dental site. The heat score 0 may be determined for those difference values which may be attributed todiscrepancies between the accuracy of the first 3D model and the accuracy of the second 3Dmodel. For example, a region of the first 3D model may be more sampled than thecorresponding region in the second 3D model, therefore the density of data points in the region of the first 3D model may be larger than the density of data points in the corresponding region of the second 3D model. This may result in differences between the first and second 3D models which are not associated with real clinical changes.The gap between the heat scores may be increased by elevating the heat scores by an exponent^, for example if the dental practitioner is interested in the severity of the clinical changesbetween the first time and the second time. With reference to the example discussed above, theheat scores may be chosen between [0^, 1^, 2^, 3^], such that a contrast between different heatscores may be magnified.Finally, in step 805, an epicenter of the one or more heat scores for each of the one or moredifferences is determined by determining a maximum heat score for the corresponding difference. The epicenter may be a local epicenter, i.e. it may be determined for each of one or more regions of clinical change. For example, a difference in the color data between the first time and the second time may be determined at a point ^^in 3D space, and a heat score may be determined for said difference. At the point ^^ a difference ^^ in the color databetween the first time and the second time may be determined, and a heat score ^^may be determined for said difference. At the point ^^, a difference ^^in the color data between the first time and the second time may be determined, and a heat score ^^may be determined forsaid difference. The points ^^, ^^, ^^ may be comprised in the same region of clinical change.Accordingly, the epicenter ^ = max[^^, ^^ , ^^] may be determined for the difference in colordata in the dental site for the corresponding region of clinical change. In other words, for eachregion of clinical change, the epicenter ^ may correspond to the one or more data points in saidregion wherein the largest change in color has occurred between the first time and second time. In another example, the epicenter is a global epicenter, such that a global maximum heat score is determined for each of the one or more differences. In this case, the maximum heat score for each of the one or more differences may be determined globally among the one or more heat scores for the corresponding difference determined for all of the one or more regions of clinicalchange. It will be appreciated that a similar discussion and results hold for any other differencedetermined between the first and second 3D models, such as geometrical differences, diagnostic differences and so forth, as it will be further discussed in the following. Figure 8b illustrates a flowchart of an embodiment of the computer-implemented method further to the embodiment illustrated in Figure 8a. Steps 806a and 807a are the same as steps 801 and 802 of Figure 8a, respectively. Therefore, the same considerations discussed above inrelation to the first 3D model of the dental site obtained in step 806a and the second 3D modelof the dental site obtained in step 807a hold. Besides obtaining the first 3D model of the dentalsite in step 807a, in step 807b first augmentation data indicating an estimated presence of oneor more dental conditions at a first set of inspection sites in the first 3D model is obtained. Eachinspection site in the first set of inspection sites comprises one or more data points of the firstset of data points, and a presence of at least one of the one or more dental conditions is estimatedfor each of the one or more data points. In other words, the first set of inspection sites may be understood as a set of regions of the first 3D model in which the presence of the one or moredental conditions has been estimated at the first time. Similarly, in step 807b secondaugmentation data indicating an estimated presence of the one or more dental conditions at a second set of inspection sites in the second 3D model is obtained. Each inspection site in the second set of inspection sites comprises one or more data points of the second set of data points, and a presence of at least one of the one or more dental conditions is estimated for each of the one or more data points. In other words, the second set of inspection sites may be understoodas a set of regions of the second 3D model in which the presence of the one or more dentalconditions has been estimated at the second time. The one or more dental conditions may comprise any of one or more of caries, gum recession, tooth wear, plaque, gum inflammation, and tooth crack. Those are among the most common dental health conditions which can be detected at early stages of their development. However,the one or more dental conditions may comprise any other dental condition which is notmentioned above, and which may be of interest in the examination process of the patient’sdental site. At least a part of the first augmentation data and / or at least a part of the secondaugmentation data may be obtained upon processing the first set of data points and / or thesecond set of data points by using a trained machine learning model for each of the one or moredental conditions. The trained machine learning model for each dental condition outputs for each data point in the first set of data points and / or for each data point in the second set of data points a severity level of the corresponding dental condition. The trained learning model may be a pre-trained machine learning model for each of the one or more dental conditions, or it may be a pre-trained machined learning model for all the one or more dental conditions. The pre-trained machine learning model may have been trained on training data comprising data points similar to the first set of data points and to the second set of data points. The trainingdata may further comprise quantified local information about features of each data point in thetraining data. The trained machine learning model for each dental condition may output foreach data point a probability that the dental condition is present in the data point. The trained machine learning model for each dental condition may further output, for each data point inwhich a presence of the dental condition is estimated, a severity level of the dental condition,wherein the severity level quantifies the clinical severity of the dental condition at the data point. Accordingly, the first augmentation data may further comprise one or more first severitylevels of the one or more dental conditions determined for each inspection site in the first setof inspection sites. Each of the one or more first severity levels determined for each inspectionsite in the first set of inspection sites is a first severity level determined for a data point in thecorresponding inspection site at the first time. Similarly, the second augmentation data mayfurther comprise one or more second severity levels of the one or more dental conditions determined for each inspection site in the second set of inspection sites. Each of the one or more second severity levels determined for each inspection site in the second set of inspection sites is a second severity level determined for a data point in the corresponding inspection site at the second time. In one example, the first augmentation data may have been generated at the first time, e.g. by processing the first 3D model in a first clinical examination carried out at the first time. Accordingly, the first augmentation data may have been stored on a storage unit, e.g. a local storage unit of the computer system on which the computer-implemented method is performed or a remote local storage unit such as a cloud storage. In this case, the first augmentation data may be obtained in step 806b upon downloading the first augmentation data from the storageunit. In another example, the first 3D model may be processed at the same time as the second3D model, e.g. at the second time which may be the time when a second clinical examination of the patient is performed. In this case, the first augmentation data and the second augmentation data may be obtained in step 806b and 807b upon processing the first 3D model and the second 3D model by using the trained machine learning model for each of the one or more dental conditions. In yet another example, the second augmentation data may have been generated at the second time, e.g. by processing the second 3D model in a second clinical examination carried out at the second time. Accordingly, the second augmentation data may have been stored on a storage unit, e.g. the local storage unit on which the computer- implemented method is performed or a remote local storage unit such as a cloud storage. In this case, the first augmentation data and the second augmentation data may be obtained in steps 806b and 807b, respectively, upon downloading the first and second augmentation data from the storage unit(s) whereon they have been stored at the first and second time. For example, the downloading may be performed at a third time, which may be the time when athird clinical examination is performed or it may be any time subsequent to the second timewhen the dental practitioner may remotely inspect the oral health of the patient, e.g. after thelatest clinical examination of the patient. The first augmentation data and the second augmentation data may further comprise segmented tooth surface data and segmented gingiva surface data, as discussed in the detailed description of Figure 8a. In step 808, the first 3D model is compared with the second 3D model, and the firstaugmentation data is compared with the second augmentation data to determine one or moredifferences therebetween. The comparison between the first and second 3D model is performedas discussed in the detailed description of Figure 8a, therefore in the following a detaileddescription of the comparison between the first and second augmentation data is provided. Inone example, comparing the first augmentation data with the second augmentation datacomprises determining one or more preexisting inspection sites by determining one or moreinspection sites in the second set of inspection sites for which at least one correspondinginspection site in the first set of inspection sites is identified. Therefore, as a first step, a spatialrelationship is determined between one or more inspection sites in the second set of inspection sites and one or more inspection sites in the first set of inspection sites, i.e. spatially corresponding inspection sites are identified between the first set of inspection sites and thefirst set of inspection sites. It may be determined that a second inspection site in the second setof inspection sites spatially corresponds to a first inspection site in the first set of inspectionsites if a spatial relationship can be identified between at least one data point in the secondinspection site and at least one data point in the first inspection site. Therefore, if a secondinspection site in the second set of inspection sites and a first inspection site in the first set ofinspection sites overlap at least partially in 3D space, then the first and second inspection sitesare determined to be spatially corresponding inspection sites, i.e. the first and second inspectionsites correspond to a same preexisting inspection site. In other words, preexisting inspectionsites are determined by determining inspection sites which are identified at the second time in the second 3D model, and which were already identified at the first time in the first 3D model. In a further or alternative example, comparing the first augmentation data with the secondaugmentation data may comprise determining one or more inspection sites in the second set ofinspection sites which do not spatially correspond to any inspection site in the first set of inspection sites. In other words, newly formed inspection sites are determined, i.e. inspection sites which emerged between the first time and the second time. This may be indicative, forexample, of the development of one or more dental conditions in new areas of the dental siteat the second time with respect to the first time. Similarly, comparing the first augmentationdata with the second augmentation data may further comprise determining one or moreinspection sites in the first set of inspection sites which are no longer present in the second setof inspection sites. In other words, inspection sites which disappeared between the first timeand the second time, e.g. because of the regression of one or more dental conditions at thelocations of the dental site corresponding to the disappeared inspection sites, are determined.When the one or more preexisting inspection sites are determined, a difference value isdetermined for each of the one or more preexisting inspection sites. In one example, thedifference value of each preexisting inspection site quantifies a difference between a secondsize of the corresponding preexisting inspection site in the second set of inspection sites and a first size of the corresponding preexisting inspection site in the first set of inspection sites. Thefirst / second size of the preexisting inspection site may be determined upon determining the sizeof the corresponding inspection site in the first / second set of inspection sites. For example, the size may be determined by counting the number of facets within the corresponding inspectionsite in the first / second set of inspection sites, and / or by determining an area of each facet in theinspection site and by summing over the areas of the facets, and / or by counting the number ofpixels within the inspection site and so forth. The area of each preexisting inspection site may be expressed in mm2, cm2or any other unit of measurement suitable for measuring an area. Forexample, a preexisting inspection site ^ may be determined, corresponding to inspection site ^^with size in the first set of inspection sites and corresponding to inspection site ^^with size ^^^^ ^^in the second set of inspection site. Accordingly, the difference value ^ =^^^^^may bedetermined for the preexisting inspection site ^, which quantifies the percentage of change inthe size of the preexisting inspection site ^ at the second time relative to the first time. Such adifference value is therefore determined for each preexisting inspection site, by considering corresponding couples of inspection sites in the first set of inspection sites and second set of inspection sites. Then, the difference value determined for each preexisting inspection site is compared with a threshold value, wherein responsive to determining that the difference value is equal of larger than the threshold value the difference value is classified as clinically relevant, and wherein responsive to determining that the difference value is smaller than the threshold value the difference value is classified as not clinically relevant. With reference to the examplediscussed above, the difference value ^^^ may be compared with a threshold value ^ todetermine whether ^^^ ≥ ^. The threshold value ^ may be a positive number, such that onlypreexisting inspection sites whose size expanded with respect to the first time will be classified as clinically relevant. In some cases, the absolute value of the difference value of eachinspection site may be compared with the threshold value, e.g. in the example mentioned aboveit may be determined whether the condition |^^^| ≥ ^ is fulfilled. Using the absolute value ofthe difference value may be relevant to identify any change in the size of preexisting inspection sites, both when the change is positive, i.e. when the preexisting inspection site expanded in size, and when the change is negative, i.e. when the preexisting inspection site decreased in size. The latter case may be indicative of a regression of the one or more dental conditionsestimated at the preexisting inspection site, which may for example indicate to the dentalpractitioner that a treatment plan is being effective. Comparing the first augmentation data with the second augmentation data may further comprise determining, for each preexisting inspection sites, one or more difference values quantifying a difference between a second severity level of each of the one or more dental conditions determined for the corresponding preexisting inspection site in the second set ofinspection sites and a first severity level of the corresponding dental condition determined forthe corresponding preexisting inspection site in the first set of inspection sites. Each preexistinginspection site in the first set of inspection sites comprises one or more data points of the first set of data points. Each of these one or more data points is associated with a first severity level of each of the one or more dental conditions, i.e. a severity level of each of the one or more dental conditions that has been determined at the first time. Similarly, each preexisting inspection in the second set of inspection sites comprises one or more data points of the second set of data points. Each of these one or more data points is associated with a second severity level of each of the one or more dental conditions, i.e. a severity level of each of the one or more dental conditions that is determined at the second time. For example, the preexistinginspection site ^ may correspond to the inspection sites ^^ and ^^ in the first and second sets ofinspection sites, respectively. The inspection site may comprise the data points(^^, ^^^ ^, ^^^) of the first set of inspection sites, wherein the data points are associated with the severity levels (^^ ^ ^^, ^^, ^^) for caries, respectively. The inspection site ^^ may comprise the data points^^^) of the second set of data points, wherein the data points are associated with theseverity levels for caries, respectively. Furthermore, the spatial correspondence^ → ^^, ^^ → ^^, ^^^^^^^^^^ ^^ ^ ^ ^ ^ → ^^may be determined. Accordingly, the differences:^^= ^^^ ^^^ − ^^ ^^^= ^^ ^ may be determined for the couples of data points (^^, ^^), (^^, ^^), (^^^ ^ ^ ^ ^, ^^^), respectively. In other words, for each preexisting inspection site, determining the difference value between severity levels comprises determining the one or more data points comprised in the preexisting inspection site at the first time, determining the corresponding one or more data points comprised in the preexisting inspection site at the second time and determining a differencebetween the severity level of each dental condition determined for each data point at the firsttime and for the corresponding data point at the second time. In this way, for each preexistinginspection site, one or more difference values are determined. In some cases, the preexisting inspection site in the second set of inspection sites may comprise more data points than the corresponding preexisting inspection site in the first set of inspection sites, or vice versa, i.e. the density of data points in the preexisting inspection sites may be larger at the second timethan at the first time, or vice versa. Therefore, more than one data point in the preexistinginspection site at the second time may spatially correspond to a same data point in thecorresponding preexisting inspection site at the first time, or vice versa. For example, with reference to the example above, the inspection site ^^may comprise the data points(^^, ^^^ ^, ^^^) and the inspection site ^^may comprise the data points(^^, ^^^, ^^^ ^ ^, ^^^), wherein both the data points ^^, ^^^may be spatially associated with the data point ^^^ ^ ^. If the severity levels for cariesassociated with ^^^, ^^^^are determined to be ^^, ^^^^ ^ , respectively, then the average severitylevel ^ ^^^^^ ^ =^^^^^^^ may be determined for the data points corresponding to ^^^. Accordingly,^^ ^^^^^ ^ ^ ^^^^ ^ ^^ ^) ^, ^ , ^ which are^ the difference value ^ = is determined for the data points ^(^ ^ ^ ^ ^^^spatially corresponding data points. It will be appreciated that the above mentioned discussion can be extended to any number of corresponding data points. Once the one or more differences values for each preexisting inspection site are determined, each of these differences values is compared with a threshold value, wherein responsive to determining that a difference value is equal or greater than the threshold value the difference value is classified as clinically relevant, and responsive to determining that a difference valueis smaller than the threshold value the difference value is classified as not clinically relevant.With reference to the example above, the three difference values may be compared against a threshold ^, wherein the threshold may be set based on health standards defined by healthorganizations such as the American Dental Association or any other health association, namelyit may be verified whether one or more of the three conditions ≥ ^, ^ ≥ ^, ^ ≥ ^ ^ ^fulfilled. For example, it may be verified that ≥ ^, then the change of the severity level ^^, ^ ) may be classified as clinicallycaries at the location corresponding to the data points (^^ ^relevant. Furthermore, it may be determined that ^ and ^ do not fulfill the threshold criterion.^ ^Accordingly, the change of the severity level of caries at the locations corresponding to the^ ^ ^ ^) ( ), ^ , ^ , ^ is classified as not clinically relevant. As also mentioned above, thepoints ^(^ ^ ^ ^absolute value of the difference values may be compared against the threshold criterion in the case where the dental practitioner is interested in any kind of change in severity, i.e. bothpositive and negative, for example to get insights about the regression of the one or more dentalconditions and to track the effectiveness of a dental treatment.In steps 809, one or more heat scores quantifying one or more clinical changes in the dental site between the first time and the second time are determined by determining a heat score for each of the one or more differences. The detailed description of step 805 of Figure 8a also holds for step 810, with the only difference that in step 810 the one or more differences also encompass the differences determined between the first augmentation data and the second augmentation data. In general, the one or more differences may comprise any type of difference which can be determined in the dental site between the first time and the second time, based on 3D models obtained at the first and second times, and based on any indications related to the 3D models, e.g. diagnostic information, geometric information, color information and so forth. Similarly, in steps 810 an epicenter of the one or more heat scores is determined for each of the one or more differences by determining a maximum heat score for the corresponding difference. This step is equivalent to step 805 of Figure 8a, and therefore the detailed description of step 805 also holds for step 810.Once steps 801-805 of Figure 8a and / or the steps 806a-810 of Figure 8b are completed, asmoothing algorithm similar to the one illustrated in Figures 7a-7g may be performed to smoothout the one or more heat scores determined for neighboring data points, such as neighboringfacets of a triangular mesh. The post-processing of the one or more heat scores by using thesmoothing method according to the present disclosure may be performed upon associating each of the one or more heat scores with a data point in the first set of data points and / or associating each of the one or more heat scores with a data point in the second set of data points. In other words, each of the one or more heat scores may be mapped to a specific location in 3D spacecorresponding to one or more data points of the first 3D model and to one or more data pointsof the second 3D model.Figures 9a and 9b illustrate examples 900 of a graphical representation on a graphical-userinterface (GUI) of an output of the computer-implemented method disclosed herein. In one example, the output comprises one or more aggregated severity levels determined for a 3Dmodel of a dental site, and said 3D model comprises a plurality of data points. In anotherexample, the output comprises one or more heat scores quantifying one or more clinicalchanges in the dental site between a first time and a second time, wherein the one or moredifferences are determined by comparing a first 3D model of the dental site with a second 3Dmodel of the dental site. In yet another example, the output comprises one or more aggregatedheat scores quantifying a total clinical change in one or more regions of the dental site between the first time and second time. The graphical representation of the aggregated severity levelsmay be generated upon associating the determined aggregated severity levels with the pluralityof data points comprised in the 3D model. Similarly, the graphical representation of the one ormore heat scores and / or one or more aggregated heat scores may be generated upon associatingthe determined heat scores / aggregated heat scores with one or more data points of a first set ofdata points comprised in the first 3D model and / or associating the determined heatscores / aggregated heat scores with one or more data points of a second set of data pointscomprised in the second 3D model. Hereafter, for the sake of simplicity and unless otherwisespecified, the first 3D model / second 3D model will be referred to as “3D model”, but it will beappreciated that the rendering on the GUI may comprise both or either of the first and / orsecond 3D models. Furthermore, the discussion will refer to the one or more heat scores,however it will be appreciated that the discussion holds for the one or more aggregated heatscores. In the examples illustrated in Figure 9a and 9b, the graphical representation of the aggregated severity levels and / or of the heat scores is a color different than a rendering color of the 3D model. In the case of the graphical representation of the one or more heat scores, the heat scoresfor each individual difference may be graphically represented, or the heat scores for all of theone or more differences may be graphically represented, or the aggregated heat scores for theone or more differences may be graphically represented. The heat scores for each differencemay be graphically represented using different colors and a color bar may be displayed in theGUI to associate each color with a specific difference, such that the dental practitioner mayeasily interpret the displayed graphical representation. The examples illustrated in Figure 9aand 9b are representative of all the above-mentioned cases. The color may be, for example, red,or blue, or green, or any other color suitable to generate a color contrast with the renderingcolor of the 3D model. The intensity of the color of a point in the 3D model is determined basedon the aggregated severity level associated with said point, such that the larger the intensity,the larger the aggregated severity level of the one or more dental conditions determined in thatpoint. In the case of the graphical representation of the one or more heat scores, the intensityof the color of a point in the 3D model is determined based on the heat score for each differenceassociated with said point, such that the larger the intensity, the larger the heat score for the corresponding difference between the first and second 3D models. In another example (notshown here), the graphical representation of the aggregated severity levels / heat scores may bea pattern, and the density of the pattern in a point of the 3D model may be determined basedon the aggregated severity level / heat score(s) associated with said point.With reference to Figure 9a, regions 901 and 902 may be two inspection sites of the 3D model,i.e. a presence of at least one dental condition is estimated for each data point in 901 and 902. Accordingly, each data point in 901 and in 902 is associated with an aggregated severity level. The data points in 901 may all be associated with the same aggregated severity level, and accordingly an epicenter of the inspection site 901 comprises all its data point. An epicenter 903 of the inspection site 902 is determined to comprise only a part of the data points comprised in 902, i.e. said part of data points includes the data points of 902 associated with a maximum aggregated severity level determined for the inspection site 902. Accordingly, in the graphical representation of the inspection sites 901 and 902the intensity of the color decreases when moving away from the epicenter of each inspection site. For example, Figure 9a illustrates as the intensity of the color decreases when moving from the epicenter 903 of the inspection site 902 to a boundary of the inspection site 902. As for the inspection site 901, a single colorintensity is displayed in the graphical representation of said inspection site to indicate that allthe data points comprised in 901 are comprised in the epicenter of 901. It can be noted in Figure9a that the graphical representation of the inspection site 901 appears edgy rather than smooth,and that the color intensity discontinuously changes when moving away from the epicenter 903of the inspection site 902. Discontinuities in the color intensity in the graphical representationof the aggregated severity levels may appear when the aggregated severity levels are notsmoothed out through a smoothing method, e.g. as the one illustrated as reference number 700in Figures 7a-7g, before generating the graphical representation.In another example, regions 901 and 902 may be two regions of clinical change of the 3Dmodel in which at least one difference between the first time and second time is determined,i.e. two regions where the one or more differences between the first 3D model and the second 3D model are determined. Accordingly, each data point in 901 and 902 is associated with one or more heat scores, wherein each heat score is determined for a specific difference among theone or more differences and therefore quantifies a specific clinical change at that data pointbetween two times, e.g. a change in the geometry at the data point, a change in the color dataat the data point, a change in the severity level of a dental condition at the data point and soforth. In some examples, region 901 and / or region 902 may be a preexisting inspection site,namely an inspection site which was identified in the first 3D model at the first time and whichis also identified in the second 3D model at the second time, whose size changed between thefirst and second times. If the size of the preexisting inspection site increased over time, a largerheat score is associated with the one or more data points which were not comprised in thepreexisting inspection site at the first time, but which are comprised in the preexistinginspection site at the second time, i.e. the new data points in the preexisting inspection site. Ifinstead the size of the preexisting inspection site decreased over time, a larger heat score may be associated with the points which are no longer comprised in the preexisting inspection site. In another example, region 901 and / or region 902 may be a preexisting inspection site where a difference in the severity level of one or more dental conditions is determined between the first and second time. In this case, larger heat scores are associated with the data points comprised in 901 and / or 902 where a worsening of the one or more dental conditions is determined with respect to data points where an improvement of the one or more dental conditions is determined. In particular, the larger the worsening of the one or more dental conditions at a data point in 901 and / or 902, the larger the heat score associated with the data point. Anepicenter 903 of the region of clinical change 902 may be determined to comprise only a partof the data points comprised in 902, i.e. the data points of 902 associated with a maximum heatscore determined for the region of clinical change 902. The maximum heat score, i.e. theepicenter 903, may be determined for each individual difference determined between the firsttime and second time. For example, the graphical representation of the regions 901 and 902may represent the difference in the severity of caries between the first time and second time,and the intensity of the color decreases when moving away from the epicenter of the change inthe severity of caries of each region. In general, Figure 9a illustrates that the intensity of thecolor decreases when moving from the epicenter 903 of the region 902 to a boundary of theregion 902. As for the region 901, a single color intensity is displayed in the graphicalrepresentation of said region to indicate that all the data points comprised in 901 are comprisedin the epicenter of 901. It can be noted in Figure 9a that the graphical representation of theregion 901 appears edgy rather than smooth, and that the color intensity discontinuouslychanges when moving away from the epicenter 903 of the region 902. Discontinuities in thecolor intensity in the graphical representation of the heat scores may appear when the heatscores are not smoothed out through a smoothing method, e.g. as the one illustrated as referencenumber 700 in Figures 7a-7g, before generating the graphical representation.The effect of the smoothing method on the graphical representation of the aggregated severitylevels / heat scores is illustrated in Figure 9b. Due to the smoothing of the aggregated severitylevels / heat scores, the intensity of the color of an inspection site / region of clinical change 905of the 3D model smoothly decreases when moving from an epicenter 904 of the inspectionsite / region of clinical change 905 to a boundary of the inspection site / region of clinical change905. Compared to the graphical representation illustrated in Figure 9a, the graphical representation illustrated in Figure 9b clearly indicates the region(s) of the dental site whichhave the highest aggregated severity of the one or more dental conditions or, similarly, theregion(s) of the dental site which have the highest heat score for a specific difference or theregion(s) of the dental site which have the highest aggregated heat score for the one or moredifferences. Therefore, the graphical representation generated with the smoothed aggregatedseverity levels / heat scores provides the dental practitioner with a more realistic and accurateindication of the regions of the dental site which necessitate careful inspection. However, itwill appear clear to those skilled in the art that the graphical representation generated withoutsmoothing the aggregated severity levels / heat scores is still an effective guide for the dentalpractitioner to perform the examination of the dental site.Figure 10a illustrates a rendering 1000 of a 3D model of a part of a dental site with a graphicalrepresentation of aggregated severity levels of one or more dental conditions determined for a3D model according to the present disclosure. The rendering 1000 may be displayed on adisplay 206 of a computer system 205 on which the computer-implemented method disclosed herein is executed. A user of the computer system 205, e.g. a dental practitioner performing anexamination of the dental site, may want to examine the one or more dental conditions detectedin an inspection site 905 of the 3D model, and get information about the severity level of eachdental condition determined for that inspection site. Accordingly, the user may input a signalwhich is received by the computer system 205. The input signal may comprise clicking orhovering a pointer 1002 on an epicenter 904 of the inspection site 905 of the dental site, whereinthe inspection site 905 is displayed on 206 as a color different than a rendering color of the 3Dmodel. For example, the inspection site 905 may be displayed with a red color, and an intensityof the red color may be maximum at the epicenter 904 of the inspection site 905 and decreasewhen moving from the epicenter 904 to a boundary of the inspection site 904. The clicking orhovering may be performed by means of a mouse 211, or a keyboard 208 of the computersystem, or it may be performed by means of a touchpad (not shown here) of the computersystem 205. Even though the example illustrated in Figure 10 refers to a computer system 205,similar discussion holds for any system configured to perform the steps of the method describedherein, and / or configured to display the rendering 1000 of the 3D model according to thepresent disclosure. For example, a tablet (not shown here) may be configured to perform thesteps of the method disclosed herein, and it may be configured to display on a display the rendering 1000 of the 3D model with the graphical representation of the heat scores.Alternatively, the tablet may be configured just to display the rendering 1000 of the 3D modelwith the graphical representation of the heat scores on the display the tablet. In this case, the clicking or hovering may be performed by means of a finger touch of the tablet.The computer system 205 may be configured to generate an ordered list of the one or moredental conditions estimated in the inspection site 904, and to store said ordered list in a memoryof the computer system 205. The computer system may be further configured to load from the memory the ordered list of the inspection site 904 upon clicking or hovering on a region of the3D model corresponding to that inspection site. With reference to Figure 10a, the user mayclick or hover on the epicenter 904 of the inspection site 905. Accordingly, an ordered list 1003 of the one or more dental conditions determined in the epicenter 904 are loaded and displayedon the display 206. The ordered list 1003 may comprise the name of the detected one or moredental conditions, together with the severity level determined for each of the dental conditions.The order may be decreasing in a clinical significance of the dental conditions, such that thedental condition for which a maximum severity level is determined is displayed on the top of the ordered list 1003. For example, with refence to Figure 10a, in the inspection site 1004 the presence of caries, tooth wear, gum recession, gum inflammation, tooth crack and plaque is determined, with a corresponding severity level of 15, 10, 4, 3, 1 and 1, respectively. Accordingly, caries are displayed on the top of the ordered list 1003, followed by tooth wear,gum recession and so forth. The ordered list 1003 may provide the dental practitioner with anindication of the severity of each dental condition detected in the inspection site 804, such that the dental practitioner may use the ordered list 1003 as a guide to perform the examination of the dental site. For example, based on the ordered list 1003, the dental practitioner may startthe examination of the dental site by inspecting caries in the location associated with theinspection site 904, and then they may move to examine tooth wear, and so forth. Even thoughthe example illustrated in Figure 10 shows only one inspection site, the above mentioneddiscussion holds for all of the inspection sites determined for the 3D model. This substantially enhances the diagnostic efficiency.Figure 10b illustrates a rendering 1000’ of a 3D model of a part of the dental site with agraphical representation of aggregated heat scores for one or more differences determinedbetween a first 3D model of the dental site and a second 3D model of the dental site, whereinthe rendering 1000’ is further or alternative to the rendering 1000 illustrated in Figure 10a. Therendering 1000’ may be displayed on the display 206 of the computer system 205 on which thecomputer-implemented method disclosed herein is executed. The user of the computer system205, e.g. the dental practitioner performing an examination of the dental site, may want toexamine the one or more differences determined in a region of clinical change 905 of the 3Dmodel, e.g. the first 3D model and / or the second 3D model, and get information about theseverity of each clinical change determined at the region of clinical change 905. Accordingly,the user may input a signal which is received by the computer system 205. The input signalmay comprise clicking or hovering the pointer 1002 on an epicenter 904 of the region of clinicalchange 905 of the dental site, wherein the region of clinical change 905 is displayed on 206 asa color different than a rendering color of the 3D model. In this example, the epicenter 904 isdetermined upon determining a maximum aggregated heat score in the region of clinical change905. For example, the region of clinical change 905 may be displayed with a red color, and anintensity of the red color may be maximum at the epicenter 904 of the region of clinical change 905 and decrease when moving from the epicenter 904 to a boundary of the region of clinical change 905. The clicking or hovering may be performed by means of a mouse 211, or a keyboard 208 of the computer system, or it may be performed by means of a touchpad (not shown here) of the computer system 205. Even though the example illustrated in Figure 10brefers to the computer system 205, similar discussion holds for any system configured toperform the steps of the method described herein, and / or configured to display the rendering1000’ of the 3D model according to the present disclosure. For example, a tablet (not shownhere) may be configured to perform the steps of the method disclosed herein, and it may beconfigured to display on a display the rendering 1000’ of the 3D model with the graphicalrepresentation of the aggregated heat scores. Alternatively, the tablet may be configured just todisplay the rendering 1000’ of the 3D model with the graphical representation of the heat scoreson the display of the tablet. In this case, the clicking or hovering may be performed by meansof a finger touch of the tablet. The computer system 205 may be configured to generate an ordered list of the one or moredifferences, i.e. clinical changes, determined in the region of clinical change 905, and to storesaid ordered list in a memory of the computer system 205. The computer system may be furtherconfigured to load from the memory the ordered list of the region of clinical change 905 uponclicking or hovering on a region of the 3D model corresponding to that region of clinical change. With reference to Figure 10b, the user may click or hover on the epicenter 904 of theregion 905. Accordingly, an ordered list 1003 of the one or more differences determined in theepicenter 904 are loaded and displayed on the display 206. The ordered list 1003 may comprisethe name of the detected one or more differences, i.e. the name of each clinical changedetermined at the selected location of the dental site, together with the heat score determinedfor each of the differences. The order may be decreasing in a clinical significance of thedifferences, such that the difference for which a maximum heat score is determined is displayedon the top of the ordered list 1003. For example, with refence to Figure 10b, in the epicenter904 a change of 50% in the severity of caries, a change of 30% in the severity of plaque, a colorchange of 15% and a change in the size of a preexisting inspection site of 10% are determined. Accordingly, the change in the severity of caries is displayed on the top of the ordered list 1003,followed by the change in the severity of plaque, color change, and change in the size of thelesion. This means that, since the past clinical examination carried out at the first time, theseverity of caries at the second time has changed of 50%, the severity of plaque has changedof 30% and so forth. The change may be both positive or negative, meaning that any kind of change may be visualized in the rendering 1000’, or the change may be only positive, meaningthat only conditions which got worse or inspection sites which expanded in size may bevisualized in the rendering 1000’. The ordered list 1003 may provide the dental practitioner with an indication of the severity of each difference determined at the region of clinical change 905, such that the dental practitioner may use the ordered list 1003 as a guide to perform the examination of the dental site. For example, based on the ordered list 1003, the dental practitioner may start the examination of the dental site by inspecting caries in the locationassociated with the epicenter 904, and then they may move to examine plaque, and so forth.Even though the example illustrated in Figure 10b only shows one region of clinical change,the above mentioned discussion holds for all of the regions of clinical change determined forthe 3D model. This substantially enhances the diagnostic efficiency.Figure 11 illustrates a system 1100 according to the present disclosure. The system 1100 maybe the computer part of the dental scanning system as illustrated as reference number 200 inFigure 2. The system 11000 comprises a computer system 205, comprising a communicationinterface 1111 which enables the computer system 205 to exchange data with other devices. For example, the computer system 205 may be configured to receive and send data to an intraoral scanner as the one illustrated as reference number 201 in Figure 2. The computer system 205 may comprise one or more processors 1101 configured to process executable instructions, e.g. provided by a computer program. The processors 1101 may comprise a graphic processing unit 1102 and / or be configured as a central processing unit (CPU). The one or more processors 1101 may be further configured to process, partially or completely, the datareceived by the intraoral scanner to generate a 3D model representation(s) 301 of the dentalsite. The computer system 205 further comprises a display device 206, a keyboard, touchpad, a mouse or touchscreen for entering data and activating virtual buttons (user interactionelements) visualized on the display 206. The display device 206 may be a computer screen, atouchpad screen or e.g. a smart phone screen comprising a graphical user interface 1015 andhaving a visual display, wherein the 3D model representation(s) 301 and e.g. a detected healthcondition of the dental site is displayed, and / or a detected difference between two or more 3Dmodels obtained at different times is displayed. The computer system 205 may comprise amemory 1103 comprising a program content executable by the processor(s) 1101, the programcontent comprising executable instructions to perform the computer-implemented methodaccording to the present disclosure. Further, the computer system 205 may comprise one ormore storage units 1104 configured to store data such as the image data 1105 acquired fromthe intraoral scanning device during a scan session. Image data 1105 from a plurality of different intraoral scanners may be stored in storage 1104. Furthermore, image data 1105 that has been processed for using e.g. the method disclosed herein may be stored in storage 1104. Patient specific identification data, diagnostic data acquired from other scanning modalities that an intraoral scanner, and other patient relevant information may be stored in storage 1104. The storage media / medium 1104 may be configured as cloud storage or for example storage on multiple computer services which are configured to communicate with each other over a network 1114. Processing and storage of data relevant for analysis by e.g. a diagnostic module may be performed in a cloud setup and loaded into a computer therefrom and / or performedlocally. The storage 1104 may also store 3D model representation(s) 301 generated historicallyfor a patient, as well as 3D model representation(s) 301 generated during a clinical visit.Furthermore, the system 1100 comprises one or more dental condition detection programs1106, each of them configured as detection program 1106 for each of one or more dentalcondition. Each of the diagnostic module or programs 1106 comprises executable instructionsfor detecting a dental condition in the 3D model representation(s) 301. The dental conditionsdetection programs 1106 may be configured with a 3D representation receipt module 1107 forobtaining 3D model representation(s) 301 from image data 1113 of an imaging device 1112,such as the intraoral scanner illustrated as reference number 201 in Figure 2. Machine learningmodules 401, 506, 607-612 are configured to receive a plurality of data points comprised inthe 3D model 301, wherein each trained model forming a part of each machine learningmodules 401, 506, 607-612 is configured to assign severity level probabilities to each of thedata points input thereto. A post-processing module 1109 is configured to perform one or morepost-processing steps to the output from each of the trained learning models configured to formpart of each machine learning module 401, 506, 607-612. The post-processing module may beconfigured to map the assigned severity level probabilities and / or heat scores for the one ormore differences to data points of the 3D model 301. The post-processing module may befurther configured to aggregate the severity level of each dental condition determined for each data point, and / or to aggregate the heat score of each determined difference and map the aggregated severity levels / heat scores to the data points of the 3D model 301. The display module 1110 may be configured to represent the 3D model 301 with a graphical representation of the determined aggregated severity levels and / or heat scores and / or aggregated heat scores in the graphical user interface 1115 of the display device 206. Items1. A computer-implemented method comprising the steps of:^ obtaining a three-dimensional (3D) model of a dental site, the 3D modelcomprising a plurality of data points describing a surface of the dental site in 3D space; ^estimating a presence of one or more dental conditions defining one or moreinspection sites in the 3D model, wherein each inspection site comprises one or more of the plurality of data points; ^determining an aggregated severity level of the one or more dental conditions foreach of the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site; and ^determining an epicenter of the one or more aggregated severity levels of eachinspection site by determining a maximum aggregated severity level of each inspection site and associating the one or more data points of the maximum aggregated severity level with a location of the epicenter.2. The computer-implemented method according to item 1, further comprising associatingthe determined one or more aggregated severity levels with the plurality of data points of the 3D model and rendering the 3D model on a graphical user interface (GUI) with a graphical representation of the one or more aggregated severity levels.3. The computer-implemented method according to item 2, wherein the graphicalrepresentation of the one or more aggregated severity levels is displayed as an indication on the 3D model.4. The computer-implemented method according to item 3, wherein the indication is a colordifferent than a rendering color of the 3D model, and an intensity of the color decreases with a distance from the estimated epicenter of each inspection site.5. The computer-implemented method according to item 3, wherein the indication is apattern, and a density of the pattern decreases with a distance from the estimated epicenter of each inspection site.6. The computer-implemented method according to any of items 4 or 5, further comprisingcalculating the distance using a Euclidean distance measure.7. The computer-implemented method according to any of items 4 or 5, further comprisingcalculating the distance using a geodesic distance measure.8. The computer-implemented method according to any of items 4 or 5, further comprisingcalculating the distance using a facet distance measure.9. The computer-implemented method according to any of the previous items, wherein theone or more dental conditions comprise one or more of caries, gum recession, tooth wear, plaque, gum inflammation, and tooth crack.10. The computer-implemented method according to any of the previous items, whereinestimating the presence of the one or more dental conditions comprises processing the plurality of data points using a trained machine learning model for each of the one or more dental conditions.11. The computer-implemented method according to item 10, wherein the trained machinelearning model for each dental condition outputs for each data point a severity level of each dental condition, and wherein the severity level is associated with a number on a predefined scale of severity of each dental condition.12. The computer-implemented method according to item 11, wherein the predefined scale ofseverity of each dental condition comprises a minimum number which is indicative of an absent dental condition.13. The computer-implemented method according to item 12, wherein the minimum numberis zero.14. The computer-implemented method according to any of the previous items, whereindetermining the aggregated severity level of the one or more dental conditions for each data point comprises determining the severity level of each dental condition for each data point, and for each data point aggregating the determined severity level of each dental condition.15. The computer-implemented method according to item 14, wherein aggregating thedetermined severity level of each dental condition comprises summing the determined severity level of each dental condition.16. The computer-implemented method according to item 14, wherein aggregating thedetermined severity level of each dental condition further comprises determining a weighted sum of the determined severity level of each dental condition.17. The computer-implemented method according to item 16, wherein determining theweighted sum further comprises assigning a weight to the determined severity level of each dental condition based on a clinical significance of each dental condition.18. The computer-implemented method according to any of the previous items, wherein theplurality of data points comprises a point cloud.19. The computer-implemented method according to any of the previous items, wherein theplurality of data points comprises a plurality of facets of a triangle mesh.20. The computer-implemented method according to any of the previous items, wherein theplurality of data points comprises a plurality of vertices of a triangle mesh.21. The computer-implemented method according to any of the previous items, furthercomprising smoothing the one or more aggregated severity levels determined for each inspection site, whereby one or more smoothed aggregate severity levels are determinedfor each inspection site.22. The computer-implemented method according to item 21, wherein the one or moresmoothed aggregated severity levels determined for each inspection site linearly or exponentially decrease with the distance from the epicenter determined for eachinspection site.23. The computer-implemented method according to any of the previous items, furthercomprising determining a global epicenter of the one or more dental conditions of the 3Dmodel by determining a global maximum aggregated severity level among the one ormore aggregated severity levels determined for the one or more inspection sites, andassociating the one or more data points of the global maximum aggregated severity level with a location of the global epicenter.24. The computer-implemented method according to any of the previous items, furthercomprising normalizing the one or more aggregated severity levels of the one or more inspection sites with the determined global maximum aggregated severity level, whereby the normalized one or more aggregated severity levels of the one or more inspection sites vary between 1 and 0, wherein 1 is the normalized aggregated severity level of the global epicenter of the 3D model.25. The computer-implemented method according to any of the previous items, furthercomprising generating an ordered list of the one or more dental conditions estimated for each inspection site and displaying the ordered list on the GUI upon receiving an inputsignal.26. The computer-implemented method according to item 25, further comprising determiningan order of importance of the severity level of each of the one or more dental conditionsestimated for each inspection site, wherein the order is a descending order.27. The computer-implemented method according to any of items 25 or 26, furthercomprising storing the ordered list generated for each inspection site and loading theordered list upon receiving an input signal.28. The computer-implemented method according to any of items 25-27, wherein the inputsignal comprises hovering or clicking on an inspection site on the displayed 3D model with the graphical representation of the one or more aggregated severity levels.29. The computer-implemented method according to any of items 25-28, wherein the listcomprises the severity level of each of the one or more dental conditions estimated foreach data point in each inspection site.30. A computer program product comprising instructions which, when executed by acomputer, cause the computer to perform the method according to any of the preceding items.31. A non-volatile computer-readable medium comprising instructions which, when executedby a computer, cause the computer to perform the method according to any of the preceding items.32. A system comprising:^ an intraoral scanner; and^ a computer system comprising:- a display;- a communication interface;- one or more processors;- one or more memories containing a program content executable by the one ormore processors, the program content comprising executable instructions to: a. obtain a 3D model of a dental site, the 3D model comprising a plurality ofdata points collectively describing a surface of the dental site in 3D space; b. estimate a presence of one or more dental conditions defining one or moreinspection sites in the 3D model, wherein each inspection site comprises one or more of the plurality of data points; c. determine an aggregated severity level of the one or more dentalconditions for each of the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site; andd. determine an epicenter of the one or more aggregated severity levels ofeach inspection site by determining a maximum aggregated severity level of the inspection site and associate the one or more data points of the maximum aggregated severity level with a location of the epicenter.33. The system according to item 32, wherein the communication interface is configured toenable the computer system to exchange data with one or more external devices and with one or more external networks.34. The system according to item 33, wherein the one or more external networks comprise atleast one cloud network.35. The system according to item 34, wherein the at least one cloud network comprisesexecutable instructions to process the plurality of data points using a trained machine learning model for each of the one or more dental conditions.36. A computer-implemented method comprising the steps of:- obtaining a first three-dimensional (3D) model of a dental site comprising a firstset of data points, wherein the first set of data points describes a surface of the dental site at a first time; -obtaining a second 3D model of the dental site comprising a second set of datapoints, wherein the second set of data points describes the surface of the dental site at a second time; -comparing the first 3D model with the second 3D model to determine one or moredifferences therebetween; -determining one or more heat scores quantifying one or more clinical changes inthe dental site between the first time and second time by determining a heat score for each of the one or more differences; and -determining an epicenter of the one or more heat scores for each of the one ormore differences by determining a global maximum heat score for the corresponding difference. 37. The method according to item 36, wherein the second time is subsequent to the first time.38. The method according to any of items 36 or 37, further comprising obtaining firstaugmentation data indicating an estimated presence of one or more dental conditions at a first set of inspection sites in the first 3D model.39. The method according to any of items 36-38, further comprising obtaining secondaugmentation data indicating an estimated presence of the one or more dental conditions at a second set of inspection sites in the second 3D model.40. The method according to any of items 38-39, wherein the first augmentation data and thesecond augmentation data further comprise segmented tooth surface data and segmented gingiva surface data.41. The method according to any of items 36-40, wherein comparing the first 3D model withthe second 3D model comprises determining a spatial relationship between at least a part of the first 3D model and at least a part of the second 3D model.42. The method according to item 41, wherein determining the spatial relationship comprisesusing the first augmentation data and using the second augmentation data to determine that the at least a part of the first 3D model and the at least a part of the second 3D model represent a same part of the dental site.43. The method according to any of items 38-42, wherein at least a part of the firstaugmentation data and / or at least a part of the second augmentation data is obtained uponprocessing the first set of data points and / or the second set of data points by using a trained machine learning model for each of the one or more dental conditions.44. The method according to item 43, wherein the trained learning model is a pre-trainedmachine learning model for each of the one or more dental conditions, or the trained learning model is a pre-trained machine learning model for all of the one or more dentalconditions.45. The method according to any of items 43-44, wherein the trained machine learning modelfor each dental condition outputs for each data point in the first set of data points and / or for each data point in the second set of data points a severity level of each dental condition.46. The method according to any of items 38-45, wherein each inspection site of the first setof inspection sites comprises one or more data points within the first set of data points andeach inspection site of the second set of inspection sites comprises one or more datapoints within the second set of data points.47. The method according to any of items 38-46, wherein the first augmentation data furthercomprises one or more first severity levels of one or more dental conditions determinedfor each inspection sites in the first set of inspection sites, and the second augmentation data further comprises one or more second severity levels of the one or more dentalconditions determined for each inspection site in the second set of inspection sites.48. The method according to any of items 36-47, wherein comparing the first 3D model withthe second 3D model comprises determining a presence of at least a part of the dental sitein the second 3D model which is not present in the first 3D model.49. The method according to item 48, wherein the at least a part of the dental site comprises adental prosthesis.50. The method according to any of items 48-49, wherein the at least a part of the dental sitecomprises a dental filling.51. The method according to any of items 36-50, wherein comparing the first 3D model withthe second 3D model comprises determining one or more differences between secondcolor data comprised in the second set of data points and first color data comprised in the first set of data points.52. The method according to item 51, wherein determining the one or more differencesbetween the second color data and the first color data comprises: -computing one or more difference values, each difference value quantifying adifference between a color coordinate of a data point in the second set of data points and a color coordinate of a corresponding data point in the first set of data points; and- comparing each of the one or more difference values with a threshold value,wherein responsive to determining that a difference value among the one or more difference values is equal or greater than a threshold value the difference value isclassified as clinically relevant, and wherein responsive to determining that a difference value among the one or more difference values is smaller than the threshold value the difference value is classified as not clinically relevant.The method according to any of items 38-52, further comprising comparing the firstaugmentation data with the second augmentation data to determine the one or more differences therebetween.The method according to item 53, wherein comparing the first augmentation data with thesecond augmentation data comprises determining one or more preexisting inspection sitesby determining one or more inspection sites in the second set of inspection sites for which at least one corresponding inspection site in the first set of inspection sites is identified.The method according to item 54, further comprising:- computing a difference value for each of the one or more preexisting inspectionsites, the difference value quantifying a difference between a second size of the corresponding preexisting inspection site in the second set of inspection sites and a first size of the corresponding preexisting inspection site in the first set of inspection sites; and- comparing the difference value with a threshold value, wherein responsive todetermining that the difference value is equal or larger than the threshold value the difference value is classified as clinically relevant, and wherein responsive to determining that the difference value is smaller than the threshold value the difference value is classified as not clinically relevant.The method according to any of items 54-55, further comprising:- computing for each preexisting inspection site one or more difference values, eachdifference value quantifying a difference between a second severity level of each of the one or more dental conditions determined for the corresponding preexisting inspection site in the second set of inspection sites and a first severity level of each of the one or more dental conditions determined for the corresponding preexisting inspection site in the first set of inspection sites; and- comparing each of the one or more difference values with a threshold value,wherein responsive to determining that the difference value is equal or greaterthan the threshold value the difference value is classified as clinically relevant, and responsive to determining that a difference value is smaller than the threshold value the difference value is classified as not clinically relevant.The method according to any of items 53-56, wherein comparing the first augmentationdata with the second augmentation data further comprises determining one or moreinspection sites in the second set of inspection sites for which no corresponding inspection site in the first set of inspection sites is identified and / or determining one or more inspection sites in the first set of inspection sites for which no corresponding inspection site in the second set of inspection sites is identified.The method according to any of items 36-57, further comprising associating thedetermined one or more heat scores for the one or more differences with one or more data points of the second set of data points and rendering a view of the second 3D model on a GUI with a graphical representation of the one or more heat scores.The method according to any of items 36-58, further comprising associating thedetermined one or more heat scores for the one or more differences with one or more datapoints of the first set of data points and rendering a view of the first 3D model on a GUIwith a graphical representation of the one or more heat scores.The method according to any of items 58-59, wherein the graphical representation of theone or more heat scores for the one or more differences is displayed as an indication onthe view of the second 3D model and / or as an indication on the view of the first 3D model.The method according to item 60, wherein the indication is a color different than arendering color of the view of the second 3D model and different than a rendering color of the view of the first 3D model.62. The method according to any of items 60-61, wherein a different color is used for the oneor more heat scores determined for each of the one or more differences, and an intensity of the color decreases with a distance from the determined epicenter of the one or more heat scores for the corresponding difference.63. The method according to item 60, wherein the indication is a pattern different than arendering color of the view of the second 3D model and different than a rendering color of the view of the first 3D model.64. The method according to any of items 60 or 63, wherein a different pattern is used for theone or more heat scores determined for each of the one or more differences, and a densityof the pattern decreases with a distance from the determined epicenter of the one or moreheat scores for the corresponding difference.65. The method according to any of items 36-64, wherein the one or more differences defineone or more regions of clinical change in the first 3D model and / or in the second 3D model, and each of the one or more differences is associated with a type of clinicalchange.66. The method according to item 65, wherein the type of clinical change is associated withone of: a change in a size of a preexisting inspection site, a change in a severity level of a dental condition among one or more dental conditions, a color change in a region of dental site, or a change in the 3D geometry of a region of the dental site.67. The method according to any of items 36-66, further comprising determining one or moreaggregated heat scores for the one or more differences for each of the one or more regions of clinical change.68. The method according to item 67, wherein determining the one or more aggregated heatscores comprises summing the one or more heat scores for each of the one or moredifferences of the corresponding region of clinical change.69. The method according to item 67, wherein determining the one or more aggregated heatscores comprises computing a weighted sum of the one or more heat scores for each of the one of more differences of the corresponding region of clinical change.70. The method according to item 69, wherein computing the weighted sum comprisesassigning a weight to each of the one or more differences, wherein the weight is based ona clinical significance of the corresponding difference.71. The method according to any of items 38-70, wherein the first augmentation data furthercomprises a first rate of change of a severity of each of one or more dental conditionsdetermined at each inspection site in the first set of inspection sites, and the second augmentation data comprises a second rate of change of the severity of each of the one ormore dental conditions determined at each inspection site in the second set of inspection sites.72. The method according to item 71, wherein the first rate of change of each dentalcondition is determined based on a difference between a first severity level of thecorresponding dental condition determined at a first time and a second severity level of the corresponding dental condition determined at a second time, wherein the second time is subsequent to the first time.73. The method according to any of items 71-72, wherein the second rate of change of theseverity of each dental condition is determined based on a difference between a thirdseverity level of the corresponding dental condition determined at a third time and a fourth severity level of the corresponding dental condition determined at a fourth time, wherein the fourth time is subsequent to the third time.74. The method according to item 73, wherein the third time coincides with the second time,whereby the second severity level coincides with the third severity level.75. The method according to item 73, wherein the third time is subsequent to the second time.76. The method according to any of items 71-75, further comprising determining a differencebetween a second rate of change of the severity of each dental condition and a first rate of change of the severity of the corresponding dental condition.

Claims

Claims 1. A computer-implemented method comprising the steps of:^ obtaining a three-dimensional (3D) model of a dental site, the 3D modelcomprising a plurality of data points describing a surface of the dental site in 3D space; ^estimating a presence of one or more dental conditions defining one or moreinspection sites in the 3D model, wherein each inspection site comprises one or more of the plurality of data points; ^determining an aggregated severity level of the one or more dental conditions foreach of the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site; and ^determining an epicenter of the one or more aggregated severity levels of eachinspection site by determining a maximum aggregated severity level of each inspection site and associating the one or more data points of the maximum aggregated severity level with a location of the epicenter.

2. The computer-implemented method according to claim 1, further comprisingassociating the determined one or more aggregated severity levels with the plurality of data points of the 3D model and rendering the 3D model on a graphical user interface (GUI) with a graphical representation of the one or more aggregated severity levels.

3. The computer-implemented method according to claim 2, wherein the graphicalrepresentation of the one or more aggregated severity levels is displayed as an indication on the 3D model.

4. The computer-implemented method according to claim 3, wherein the indication is acolor different than a rendering color of the 3D model, and an intensity of the color decreases with a distance from the estimated epicenter of each inspection site.

5. The computer-implemented method according to claim 3, wherein the indication is apattern, and a density of the pattern decreases with a distance from the estimated epicenter of each inspection site.

6. The computer-implemented method according to any of claims 4 or 5, furthercomprising calculating the distance using a Euclidean distance measure.

7. The computer-implemented method according to any of claims 4 or 5, furthercomprising calculating the distance using a geodesic distance measure.

8. The computer-implemented method according to any of claims 4 or 5, furthercomprising calculating the distance using a facet distance measure.

9. The computer-implemented method according to any of the previous claims, whereinthe one or more dental conditions comprise one or more of caries, gum recession, tooth wear, plaque, gum inflammation, and tooth crack.

10. The computer-implemented method according to any of the previous claims, whereinestimating the presence of the one or more dental conditions comprises processing the plurality of data points using a trained machine learning model for each of the one or more dental conditions.

11. The computer-implemented method according to claim 10, wherein the trainedmachine learning model for each dental condition outputs for each data point a severity level of each dental condition, and wherein the severity level is associated with a number on a predefined scale of severity of each dental condition.

12. The computer-implemented method according to claim 11, wherein the predefinedscale of severity of each dental condition comprises a minimum number which is indicative of an absent dental condition.

13. The computer-implemented method according to claim 12, wherein the minimumnumber is zero.

14. The computer-implemented method according to any of the previous claims, whereindetermining the aggregated severity level of the one or more dental conditions for each data point comprises determining the severity level of each dental condition foreach data point, and for each data point aggregating the determined severity level of each dental condition.

15. The computer-implemented method according to claim 14, wherein aggregating thedetermined severity level of each dental condition comprises summing the determined severity level of each dental condition.

16. The computer-implemented method according to claim 14, wherein aggregating thedetermined severity level of each dental condition further comprises determining a weighted sum of the determined severity level of each dental condition.

17. The computer-implemented method according to claim 16, wherein determining theweighted sum further comprises assigning a weight to the determined severity level of each dental condition based on a clinical significance of each dental condition.

18. The computer-implemented method according to any of the previous claims, whereinthe plurality of data points comprises a point cloud.

19. The computer-implemented method according to any of the previous claims, whereinthe plurality of data points comprises a plurality of facets of a triangle mesh.

20. The computer-implemented method according to any of the previous claims, whereinthe plurality of data points comprises a plurality of vertices of a triangle mesh.

21. The computer-implemented method according to any of the previous claims, furthercomprising smoothing the one or more aggregated severity levels determined for each inspection site, whereby one or more smoothed aggregate severity levels aredetermined for each inspection site.

22. The computer-implemented method according to claim 21, wherein the one or moresmoothed aggregated severity levels determined for each inspection site linearly or exponentially decrease with the distance from the epicenter determined for each inspection site.

23. The computer-implemented method according to any of the previous claims, furthercomprising determining a global epicenter of the one or more dental conditions of the3D model by determining a global maximum aggregated severity level among the one or more aggregated severity levels determined for the one or more inspection sites, and associating the one or more data points of the global maximum aggregated severity level with a location of the global epicenter.

24. The computer-implemented method according to any of the previous claims, furthercomprising normalizing the one or more aggregated severity levels of the one or more inspection sites with the determined global maximum aggregated severity level, whereby the normalized one or more aggregated severity levels of the one or more inspection sites vary between 1 and 0, wherein 1 is the normalized aggregated severity level of the global epicenter of the 3D model.

25. The computer-implemented method according to any of the previous claims, furthercomprising generating an ordered list of the one or more dental conditions estimated for each inspection site and displaying the ordered list on the GUI upon receiving aninput signal.

26. The computer-implemented method according to claim 25, further comprisingdetermining an order of importance of the severity level of each of the one or moredental conditions estimated for each inspection site, wherein the order is a descending order.

27. The computer-implemented method according to any of claims 25 or 26, furthercomprising storing the ordered list generated for each inspection site and loading theordered list upon receiving an input signal.

28. The computer-implemented method according to any of claims 25-27, wherein theinput signal comprises hovering or clicking on an inspection site on the displayed 3D model with the graphical representation of the one or more aggregated severity levels.

29. The computer-implemented method according to any of claims 25-28, wherein the listcomprises the severity level of each of the one or more dental conditions estimated foreach data point in each inspection site.

30. A computer program product comprising instructions which, when executed by acomputer, cause the computer to perform the method according to any of the preceding items.

31. A non-volatile computer-readable medium comprising instructions which, whenexecuted by a computer, cause the computer to perform the method according to any of the preceding items.

32. A system comprising:^ an intraoral scanner; and^ a computer system comprising:- a display;- a communication interface;- one or more processors;- one or more memories containing a program content executable by the one ormore processors, the program content comprising executable instructions to: e. obtain a 3D model of a dental site, the 3D model comprising a plurality ofdata points collectively describing a surface of the dental site in 3D space; f. estimate a presence of one or more dental conditions defining one or moreinspection sites in the 3D model, wherein each inspection site comprises one or more of the plurality of data points; g. determine an aggregated severity level of the one or more dentalconditions for each of the one or more of the plurality of data points, thereby determining one or more aggregated severity levels for each inspection site; and h. determine an epicenter of the one or more aggregated severity levels ofeach inspection site by determining a maximum aggregated severity level of the inspection site and associate the one or more data points of themaximum aggregated severity level with a location of the epicenter.

33. The system according to claim 32, wherein the communication interface is configured toenable the computer system to exchange data with one or more external devices and with one or more external networks.

34. The system according to claim 33, wherein the one or more external networks comprise atleast one cloud network.

35. The system according to claim 34, wherein the at least one cloud network comprisesexecutable instructions to process the plurality of data points using a trained machine learning model for each of the one or more dental conditions.

36. A computer-implemented method comprising the steps of:- obtaining a first three-dimensional (3D) model of a dental site comprising a firstset of data points, wherein the first set of data points describes a surface of the dental site at a first time; -obtaining a second 3D model of the dental site comprising a second set of datapoints, wherein the second set of data points describes the surface of the dental site at a second time; -comparing the first 3D model with the second 3D model to determine one or moredifferences therebetween; -determining one or more heat scores quantifying one or more clinical changes inthe dental site between the first time and second time by determining a heat score for each of the one or more differences; and -determining an epicenter of the one or more heat scores for each of the one ormore differences by determining a global maximum heat score for the corresponding difference.

37. The method according to claim 36, wherein the second time is subsequent to the firsttime.

38. The method according to any of claims 36 or 37, further comprising obtaining firstaugmentation data indicating an estimated presence of one or more dental conditions at a first set of inspection sites in the first 3D model.

39. The method according to any of claims 36-38, further comprising obtaining secondaugmentation data indicating an estimated presence of the one or more dental conditions at a second set of inspection sites in the second 3D model.

40. The method according to any of claims 38-39, wherein the first augmentation data andthe second augmentation data further comprise segmented tooth surface data and segmented gingiva surface data.

41. The method according to any of claims 36-40, wherein comparing the first 3D modelwith the second 3D model comprises determining a spatial relationship between at least a part of the first 3D model and at least a part of the second 3D model.

42. The method according to claim 41, wherein determining the spatial relationshipcomprises using the first augmentation data and using the second augmentation datato determine that the at least a part of the first 3D model and the at least a part of the second 3D model represent a same part of the dental site.

43. The method according to any of claims 38-42, wherein at least a part of the firstaugmentation data and / or at least a part of the second augmentation data is obtainedupon processing the first set of data points and / or the second set of data points by using a trained machine learning model for each of the one or more dental conditions.

44. The method according to claim 43, wherein the trained learning model is a pre-trainedmachine learning model for each of the one or more dental conditions, or the trained learning model is a pre-trained machine learning model for all of the one or moredental conditions.

45. The method according to any of claims 43-44, wherein the trained machine learningmodel for each dental condition outputs for each data point in the first set of datapoints and / or for each data point in the second set of data points a severity level of each dental condition.

46. The method according to any of claims 38-45, wherein each inspection site of the firstset of inspection sites comprises one or more data points within the first set of datapoints and each inspection site of the second set of inspection sites comprises one ormore data points within the second set of data points.

47. The method according to any of claims 38-46, wherein the first augmentation datafurther comprises one or more first severity levels of one or more dental conditionsdetermined for each inspection sites in the first set of inspection sites, and the second augmentation data further comprises one or more second severity levels of the one ormore dental conditions determined for each inspection site in the second set of inspection sites.

48. The method according to any of claims 36-47, wherein comparing the first 3D modelwith the second 3D model comprises determining a presence of at least a part of thedental site in the second 3D model which is not present in the first 3D model.

49. The method according to claim 48, wherein the at least a part of the dental sitecomprises a dental prosthesis.

50. The method according to any of claims 48-49, wherein the at least a part of the dentalsite comprises a dental filling.

51. The method according to any of claims 36-50, wherein comparing the first 3D modelwith the second 3D model comprises determining one or more differences betweensecond color data comprised in the second set of data points and first color data comprised in the first set of data points.

52. The method according to claim 51, wherein determining the one or more differencesbetween the second color data and the first color data comprises: -computing one or more difference values, each difference value quantifying adifference between a color coordinate of a data point in the second set of datapoints and a color coordinate of a corresponding data point in the first set of data points; and- comparing each of the one or more difference values with a threshold value,wherein responsive to determining that a difference value among the one or more difference values is equal or greater than a threshold value the difference value isclassified as clinically relevant, and wherein responsive to determining that a difference value among the one or more difference values is smaller than the threshold value the difference value is classified as not clinically relevant.

53. The method according to any of claims 38-52, further comprising comparing the firstaugmentation data with the second augmentation data to determine the one or more differences therebetween.

54. The method according to claim 53, wherein comparing the first augmentation datawith the second augmentation data comprises determining one or more preexistinginspection sites by determining one or more inspection sites in the second set of inspection sites for which at least one corresponding inspection site in the first set of inspection sites is identified.

55. The method according to claim 54, further comprising:- computing a difference value for each of the one or more preexisting inspectionsites, the difference value quantifying a difference between a second size of the corresponding preexisting inspection site in the second set of inspection sites and a first size of the corresponding preexisting inspection site in the first set of inspection sites; and - comparing the difference value with a threshold value, wherein responsive to determining that the difference value is equal or larger than the threshold value the difference value is classified as clinically relevant, and wherein responsive to determining that the difference value is smaller than the threshold value the difference value is classified as not clinically relevant.

56. The method according to any of claims 54-55, further comprising:- computing for each preexisting inspection site one or more difference values, eachdifference value quantifying a difference between a second severity level of eachof the one or more dental conditions determined for the corresponding preexisting inspection site in the second set of inspection sites and a first severity level of each of the one or more dental conditions determined for the corresponding preexisting inspection site in the first set of inspection sites; and- comparing each of the one or more difference values with a threshold value,wherein responsive to determining that the difference value is equal or greaterthan the threshold value the difference value is classified as clinically relevant, and responsive to determining that a difference value is smaller than the threshold value the difference value is classified as not clinically relevant.

57. The method according to any of claims 53-56, wherein comparing the firstaugmentation data with the second augmentation data further comprises determiningone or more inspection sites in the second set of inspection sites for which no corresponding inspection site in the first set of inspection sites is identified and / or determining one or more inspection sites in the first set of inspection sites for which no corresponding inspection site in the second set of inspection sites is identified.

58. The method according to any of claims 36-57, further comprising associating thedetermined one or more heat scores for the one or more differences with one or more data points of the second set of data points and rendering a view of the second 3D model on a GUI with a graphical representation of the one or more heat scores.

59. The method according to any of claims 36-58, further comprising associating thedetermined one or more heat scores for the one or more differences with one or more data points of the first set of data points and rendering a view of the first 3D model on a GUI with a graphical representation of the one or more heat scores.

60. The method according to any of claims 58 or 59, wherein the graphical representationof the one or more heat scores for the one or more differences is displayed as anindication on the view of the second 3D model and / or as an indication on the view of the first 3D model.

61. The method according to claim 60, wherein the indication is a color different than arendering color of the view of the second 3D model and different than a rendering color of the view of the first 3D model.

62. The method according to any of claims 60-61, wherein a different color is used for theone or more heat scores determined for each of the one or more differences, and an intensity of the color decreases with a distance from the determined epicenter of the one or more heat scores for the corresponding difference.

63. The method according to claim 60, wherein the indication is a pattern different than arendering color of the view of the second 3D model and different than a rendering color of the view of the first 3D model.

64. The method according to any of claims 60 or 63, wherein a different pattern is usedfor the one or more heat scores determined for each of the one or more differences, and a density of the pattern decreases with a distance from the determined epicenter ofthe one or more heat scores for the corresponding difference.

65. The method according to any of claims 36-64, wherein the one or more differencesdefine one or more regions of clinical change in the first 3D model and / or in the second 3D model, and each of the one or more differences is associated with a type ofclinical change.

66. The method according to claim 65, wherein the type of clinical change is associatedwith one of: a change in a size of a preexisting inspection site, a change in a severity level of a dental condition among one or more dental conditions, a color change in a region of dental site, or a change in the 3D geometry of a region of the dental site.

67. The method according to any of claims 36-66, further comprising determining one ormore aggregated heat scores for the one or more differences for each of the one or more regions of clinical change.

68. The method according to claim 67, wherein determining the one or more aggregatedheat scores comprises summing the one or more heat scores for each of the one ormore differences of the corresponding region of clinical change.

69. The method according to claim 67, wherein determining the one or more aggregatedheat scores comprises computing a weighted sum of the one or more heat scores for each of the one of more differences of the corresponding region of clinical change.

70. The method according to claim 69, wherein computing the weighted sum comprisesassigning a weight to each of the one or more differences, wherein the weight is basedon a clinical significance of the corresponding difference.

71. The method according to any of claims 38-70, wherein the first augmentation datafurther comprises a first rate of change of a severity of each of one or more dentalconditions determined at each inspection site in the first set of inspection sites, and the second augmentation data comprises a second rate of change of the severity of each ofthe one or more dental conditions determined at each inspection site in the second set of inspection sites.

72. The method according to claim 71, wherein the first rate of change of each dentalcondition is determined based on a difference between a first severity level of thecorresponding dental condition determined at a first time and a second severity level of the corresponding dental condition determined at a second time, wherein the second time is subsequent to the first time.

73. The method according to any of claims 71-72, wherein the second rate of change ofthe severity of each dental condition is determined based on a difference between athird severity level of the corresponding dental condition determined at a third time and a fourth severity level of the corresponding dental condition determined at a fourth time, wherein the fourth time is subsequent to the third time.

74. The method according to claim 73, wherein the third time coincides with the secondtime, whereby the second severity level coincides with the third severity level.

75. The method according to claim 73, wherein the third time is subsequent to the secondtime.

76. The method according to any of claims 71-75, further comprising determining adifference between a second rate of change of the severity of each dental condition and a first rate of change of the severity of the corresponding dental condition.

Images