Processing scan data

Abstract

There is provided a computer-implemented method (100) for processing scan data to facilitate identification of a first subject of first scan data of a first subject. The method comprises acquiring (102, 104): at least one first representation of first images included in the first scan data; and at least one second representation of second images included in second scan data. The second scan data is of a second subject. The method comprises determining (106) whether the first subject is the same as the second subject based on a difference between the at least one first representation and the at least one second representation.

Description

[0001] 2025P00030WG

[0002] 1

[0003] PROCESSING SCAN DATA

[0004] FIELD OF THE INVENTION

[0005] The disclosure relates to methods and apparatus for processing scan data to facilitate identification of a first subject of first scan data.

[0006] BACKGROUND OF THE INVENTION

[0007] Personal care routines often include the brushing of teeth at least twice a day using a toothbrush, such as a power toothbrush, to remove plaque from a person’s oral cavity. However, such toothbrushes do not typically provide at-home methods for controlling and monitoring the person’s oral health and brushing outcomes.

[0008] Therefore, in addition to the brushing of teeth, a personal care routine may include the use of a scanning device to take regular images (e.g., by performing regular scans) of the person’s oral cavity to obtain images of their teeth. The scanning device may be an intraoral scanner (IOS). The scanning device may be operated by the subject of the scan or an external user, such as a dental clinician. The images obtained from the scan may then be analysed (e.g., using computer vision algorithms) to generate insight and statistics regarding the oral health of the subject of the scan. The images may be collected and stored in a database of scans, such that they may be processed or analysed on-demand. For example, scans obtained at different times (e.g., on different days) for the subject may be retrieved from the scan database and analysed to identify and determine any changes in the oral health of the subject. As an example, a set, S, of n scans (where an individual scan is denoted by s) may be stored for each one of a set, U, of m users (where an individual user is denoted by it). For example, for each of

[0009] U ={!{■£, u2, ... um}, there may be stored a set, S = {s1,s2, ... sn}.

[0010] SUMMARY OF THE INVENTION

[0011] There currently exist certain challenges. In some cases, the identity of a subject of a scan that is to be stored in a scan database may be unknown (e.g., to a host of the scan database). That is, metadata identifying the subject of the scan may not be found in the database or may not be accessible to a host of the scan database. This may occur because of a corruption of data stored in the scan database, the metadata being lost when the scan and metadata were transmitted to the host of the scan database for storage, and / or because the metadata was never created in the first place. Similar considerations apply to scan data that is already stored in the scan database, particularly where any information identifying the subject of the scan data has been lost, corrupted or wasn’t saved in the first place. Again, similar2025P00030WG

[0012] 2

[0013] considerations apply when a scanning device is used to obtain scan for multiple subjects and, therefore, the subject of a given scan may be unknown or unspecified.

[0014] This can cause issues (e.g., failures) when scans of a given subject are to be retrieved from the scan database to be analysed (e.g., for health monitoring purposes). For example, a scan of a given subject may not be retrievable from the scan database if the scan cannot be identified as being associated with that subject. This may result in a poor user experience and a reduction in the quality and / or accuracy of the analysis.

[0015] Embodiments of the present disclosure address these and other challenges. In particular, embodiments disclosed herein enable a determination to be made as to whether a first subject of first scan data is the same as a second subject of second scan data. The determination is based on a difference between at least one first representation of first images included in the first scan data and at least one second representation of second images included in the second scan data. The embodiments described herein facilitate the maintenance and improved integrity of databases in which scan data of subjects is stored. The methods help ensure that scans for certain subjects can be reliably retrieved from a database, e.g., for analysis and health monitoring purposes.

[0016] According to a first aspect, there is provided a method of processing scan data to facilitate identification of a first subject of first scan data. The method comprises: acquiring at least one first representation of first images in the first scan data; acquiring at least one second representation of second images in second scan data of a second subject; and determining, based on a difference between the at least one first representation and the at least one second representation, whether the first subject is the same as the second subject.

[0017] As previously discussed, an advantage of the first aspect is that an unknown subject of scan data can be identified, facilitating the maintenance and improved integrity of databases in which scan data of subjects is stored.

[0018] In an example of the first aspect, the first subject is determined to be the same as the second subject based on whether the difference between the at least one first representation and the at least one second representation fulfils one or more criteria.

[0019] In an example of the first aspect, the first images are from a first scan performed on the first subject, the second images are from one or more second scans performed on the second subject, and the determination is based on inter-scan differences between the first scan and respective second scans of the one or more second scans.

[0020] In an example of the first aspect, an inter-scan difference between the first scan and a respective second scan is based on one or more inter-representation differences between: at least one first representation of first images from the first scan; and at least one second representation of second images from the respective second scan.

[0021] In an example of the first aspect, the inter-scan difference is based on an aggregate interrepresentation difference.2025P00030WG

[0022] 3

[0023] In an example of the first aspect, the aggregate inter-representation difference is determined by: identifying pairs of representations, each pair comprising one of the first representations of the first images from the first scan and one of the second representations of the second images from the respective second scan; determining, for each pair, an inter-representation difference between the representations included in the pair; and aggregating the inter-representation differences to obtain the aggregate inter-representation difference.

[0024] In an example of the first aspect, aggregating the inter-representation differences comprises determining a mean of the inter-representation differences.

[0025] In an example of the first aspect, determining whether the first subject is the same as the second subject comprises identifying, from the inter-scan differences, a minimum inter-scan difference.

[0026] In an example of the first aspect, if the first subject is determined to be the same as the second subject, the identity of the first subject is determined to be the same as an identity of the second subject.

[0027] In an example of the first aspect, if the first subject is determined to be the same as the second subject, the method further comprises storing, in a database: the first scan data; and an association between the first scan data and the identity of the second subject.

[0028] In an example of the first aspect, if the first subject is determined not to be the same as the second subject, the method further comprises one of: discarding the first scan data; and storing, in a database, the first scan data and an indication that the identity of the first subject is unknown in the database.

[0029] In an example of the first aspect, the at least one first representation comprises at least one first vector characterizing the first images and the at least one second representation comprises at least one second vector characterizing the second images.

[0030] In an example of the first aspect, the first images are of teeth in an oral cavity of the first subject, and / or the second images are of teeth in an oral cavity of the second subject.

[0031] According to a second aspect, there is provided an apparatus for processing scan data to facilitate identification of a first subject of first scan data. The apparatus comprises processing circuitry configured to cause the apparatus to: acquire at least one first representation of first images in the first scan data; acquire at least one second representation of second images in second scan data of a second subject; and determine, based on a difference between the at least one first representation and the at least one second representation, whether the first subject is the same as the second subject.

[0032] According to a third aspect, there is provided a computer program product comprising a computer readable medium. The computer readable medium has a computer readable code embodied therein. The computer readable code is configured such that, on execution by a suitable computer or processor, the computer or processor is caused to perform the method described earlier.

[0033] These and other aspects will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.2025P00030WG

[0034] 4

[0035] BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Exemplary embodiments will now be described, by way of example only, with reference to the following drawings, in which:

[0037] Fig. 1 is a flow chart illustrating a method according to an embodiment;

[0038] Figs. 2a and 2b are images of a set of teeth in an oral cavity;

[0039] Fig. 3 is a block diagram illustrating an example apparatus according to embodiments of the disclosure; and

[0040] Fig. 4 is a block diagram illustrating an example processor according to embodiments of the disclosure.

[0041] DETAILED DESCRIPTION OF EMBODIMENTS

[0042] The health of a subject may be monitored by obtaining scan data of the subject, particularly when the scan data comprises information regarding the health of the subject (that is, the scan data comprises health data). For example, the obtained scans of the subject may be processed for the purpose of providing insights into the health of the subject based on the processed scans. The scan data may be obtained using a scanning device. As an example, scans of an oral cavity of a subject may provide insights into the oral health of the subject, such as whether and where plaque is present in their oral cavity. Such scans may be obtained using an intraoral device (IOD), such as an IOS.

[0043] Scan data may comprise images, i, from one or more scans, s. of a subject, u. For the purpose of the present disclosure, the term “scan” is used to refer to a set of images obtained during a scanning procedure performed on a subject. For example, the result of a scan may be a list or set of images, / , where each individual image of the list or set is denoted by i. The images may form at least a part of a video stream. As an example, during a scan of an oral cavity, an IOS may be moved around the oral cavity such that an imaging device of the IOS captures or images the tooth enamel located in the oral cavity (e.g., captures or images the inner teeth surfaces, outer teeth surfaces, left teeth surfaces, front teeth surfaces, right teeth surfaces, upper teeth surfaces, lower teeth surfaces, etc).

[0044] After the scan data is collected, the scan data may be processed (e.g., using classical and / or machine learning (ML) algorithms) to produce a report on the health of the subject (e.g., whether and where plaque is present in an oral cavity). As such, it may be beneficial for the scan data to be collected and / or stored in a database (which may be referred to herein as a “scan database”), such that the scan data can be processed, using any suitable processing technique, on-demand.

[0045] The scan database may contain scan data for a set of subjects (e.g., U = u u2... um}). A subject may be considered a “user” of the scanning device if the subject operated the device. A subject may be considered a “user” of the database by virtue of their scan data being stored in the database.2025P00030WG

[0046] 5

[0047] Additionally or alternatively, the person causing the scan data to be stored in the database may be considered a user of the database.

[0048] One or more scans may be stored for a given subject in the scan database, with indications that the one or more scans belong to the given subject. For example, a set of n scans may be stored for a given subject, u (e.g., Su= {s1(s2, ... sn}). The indications may be identifiers of the subject. Alternatively, the indications may be respective identifiers of the one or more scans, and the database may comprise information identifying a mapping (e.g., a one-to-many mapping) between a subject identifier of the subject and the identifiers of the one or more scans.

[0049] The scan data stored for a subject may comprise a calibration scan (e.g., an initial scan stored for the subject in the scan database) and / or one or more routine scans. For example, one or two scans may be performed on a subject per week, and these scans may be stored in the scan database with indications of associations to the subject. The skilled person will appreciate that a scan database may store a large number of scans for a large number of subjects, depending on the server and / or the computing resources used to host the scan database. However, as previously discussed, there may be certain challenges associated with hosting such a database.

[0050] In some cases, the subject of scan data may be unknown or unidentifiable to a host of the database. For example, the host of the database may not be able to identify, based on information stored in the database, the subject of a given scan. This may be because the scan data arrived at the host of the scan database without any metadata identifying the subject of the scan data. In another example, the scan data may have been stored in the database, but any metadata identifying the subject of the scan data may have been lost or corrupted. In another example, no scan data for the subject may yet be stored in the database for the subject, and the database may need to identify that the scan data belongs to a new subject. In another example, even if metadata identifying a scanning device is associated with the scan data, the subject of a given scan obtained using the scanning device may be unknown if the scanning device is or has previously been used to obtain scans for multiple different subjects. Scan data of a subject that is not known or not identifiable to (a host of) the database may be referred to herein as scan data with unknown or unidentifiable provenance.

[0051] As a result, it may not be possible for certain scan data to be stored with an association to or identifier of the subject of the scan data, making it challenging for the scan data to subsequently be retrieved and / or processed on demand for monitoring the health of the subject. This can result in a poor user experience and less accurate and / or lower quality insights regarding the health of the subject.

[0052] Thus, there are situations where a subject of scan data may need to be identified.

[0053] There are provided herein techniques of processing scan data to facilitate identification of a first subject of first scan data. In particular, present embodiments provide a computer-implemented method for determining whether first scan data is of, is for, or otherwise relates to the same subject as second scan data already stored in the database. The second scan data is of, is for, or otherwise relates to a2025P00030WG

[0054] 6

[0055] subject referred to as the “second subject”. Until the subject that the first scan data relates to is identified, this subject is referred to as the “first subject” (e.g., u^).

[0056] The determination of whether the first scan data is of, is for, or otherwise relates to the same subject (the “second subject”, (e.g., u2)) as second scan data is made by determining differences between at least one first representation of first images included in the first scan data and at least one second representation of second images included in the second scan data. This determination enables the integrity and reliability of a database to be maintained, facilitating subjects to receive high quality feedback and guidance regarding their health based on their stored health data. For example, the feedback may be derived from an oral health history of the subject based on scans obtained for the subject, and may identify whether the subject’s teeth are moving, whether tartar has been removed, or how cavities are evolving overtime.

[0057] For example, if it is found that the first scan data is for the second subject, the first scan data can be stored in a database with an association to the second subject. In other examples, it may be found that the subject related to the first scan data is not the same as any subject for which scan data is stored in the database (e.g., the first subject may be recognised as a new user of the database), and the first scan data can be handled accordingly.

[0058] For the purpose of the present disclosure, a representation, z, of an image, i, may be any suitable data that can be used to represent, summarise, or otherwise indicate structural elements of the image. As an example, an image may be converted into an (e.g., compressed) representation by determining a vector embedding, such as a latent vector, for the image. This may be performed using an ML model trained to: identify one or more structural elements (e.g., teeth and / or plaque) in one or more input image; and output a representation of the one or more input images based on the one or more identified structural elements. The ML model may be trained on a public dataset of images (e.g., a public dataset of teeth, such as from a Common Objects in Context (COCO) dataset) or, more advantageously, trained and / or fine-tuned on a custom dataset of images. For example, the custom dataset may comprise images of teeth of a given subject and, as such, the ML model can be trained to identify features of the given subject (and embed them in representations) with improved accuracy. The images on which the ML model is trained may have been obtained using different scanning conditions. Scanning conditions may include: the type or model of scanning device used to perform the scan, the wavelength and / or intensity of light (i.e., a “light modality”) used by the scanning device to illuminate an area or surface being scanned (e.g., whether white or violet light is used), a distance between the imaging device of the scanning device and the area or surface being scanned, a location of the area or surface being scanned, settings of the imaging device, dimensions of the image sensor of the imaging device, light conditions in the environment in which the scan takes place, and the perspective of the scanned area or surface in the resulting image. This allows the ML model to robustly output / embed representations of images regardless of the scanning conditions under which the images were obtained.2025P00030WG

[0059] 7

[0060] Such an ML model may be referred to herein as an “embedding ML model”. The embedding ML model may use classifiers, object detectors, and / or computer vision techniques to detect and segment each structural element of interest. For example, the structural elements may be detected using bounding boxes and / or segmented using masks. The ML model may be a neural network, such as a supervised neural network. The terms ML model, ML algorithm, Artificial Intelligence (Al) algorithm, Al model, AI / ML (AIML) algorithm, and AIML model may be used interchangeably herein.

[0061] The skilled person will appreciate that the term “image” may be used herein to refer to a complete image captured by an imaging device, or to a region of an image (e.g., an “image region”), particularly if the image region captures or otherwise comprises data relating to a structural element of interest, such as a tooth and / or plaque. That is, the term “image” may be used herein to refer to a subimage that may be incorporated in a larger, complete image. The complete image may be formed from multiple sub-images. An image may be divided into multiple sub-images if it contains more than one structural element of interest. For example, an image of an oral cavity may be divided into multiple subimages, where each sub-image captures a specific tooth in the oral cavity.

[0062] Similarly, the skilled person will appreciate that the term “image” may refer to a composite image formed from multiple, smaller (sub-)images captured by an imaging device. For example, the multiple smaller images may be converted, stitched, aligned, or otherwise combined to form the composite image.

[0063] The embodiments described herein may use one or more of a plurality of metrics to define the similarity of first scan data to second scan data to determine whether subjects of the first and second scan data are the same. For example, the embodiments may involve the use of a first metric, t(ii, i2), which may be used to define the difference (e.g., a distance) between two images and i2. h may be determined based on a difference (e.g. distance) between a representation, z1?of and a representation, z2. of i2. The embodiments may involve the use of a second metric, g s , s2). that may be used to define the difference (e.g., distance) between two scans s1and s2. The embodiments may involve the use of a third metric, f (slzSU2). that can be used to measure the similarity (e.g., distance) between a scan, s1?of a first subject, ultand a set of scans, SU2, of a second subject, u2.

[0064] Further detail regarding these metrics are provided below with reference to the method of Fig. 1. The skilled person will appreciate that the first, second, and third metrics may be used in combination to determine whether subjects of scan data are the same. For example, f may be based on one or more values of g. where g may be determined for a number of pairs of scans. Each value of g may be based on one or more values of h, where h may be determined for a number of pairs of images. However, the skilled person will appreciate that other metrics may be used to define the similarity of first scan data to second scan data to determine whether subjects of the first and second scan data are the same.2025P00030WG

[0065] 8

[0066] Fig. 1 illustrates a method 100 of processing scan data to facilitate identification of a first subject of first scan data according to an embodiment. More specifically, Fig. 1 illustrates a method 100 of operating an apparatus, such as the apparatus 300 described later with reference to Fig. 3, for processing scan data to facilitate identification of a first subject of first scan data. The apparatus 300 may be any suitable computing device and may comprise one or more of a computer, a server, a node of a cloud network, a host of a scan database, and any other suitable computing device. The method 100 illustrated in Fig. 1 is a computer-implemented method. For example, the method may be implemented by any software or application running on the apparatus 300. As described later with reference to Fig. 3, the apparatus 300 comprises one or more processors 302. The method 100 illustrated in Fig. 1 can generally be performed by or under the control of the one or more processors 302 of the apparatus 300 described later with reference to Fig. 3, such as a processor 400 described later with reference to Fig. 4.

[0067] With reference to Fig. 1, at block 102, the method comprises acquiring at least one first representation of first images in the first scan data.

[0068] At block 104 of Fig. 1, the method comprises acquiring at least one second representation of second images in second scan data of a second subject.

[0069] At block 106 of Fig. 1, the method comprises determining (106), based on a difference between the at least one first representation and the at least one second representation, whether the first subject is the same as the second subject.

[0070] If the first subject is determined to be the same as the second subject, the identity of the first subject may be determined to be the same as an identity of the second subject.

[0071] The method of Fig. 1 will now be discussed in more detail.

[0072] As previously mentioned, at block 102 of Fig. 1, the method comprises acquiring at least one first representation of first images in the first scan data (e.g., J first images). It will be appreciated that the term “at least one first representation” may be used inter-changeably herein with the more general term “first representations”.

[0073] The first scan data may be of unknown or unidentifiable provenance to a scan database in which the first scan data stored or is to be stored. For example, the subject may be unidentifiable based on metadata available for the first scan data. That is, at this stage of the method, it is not known which subject the first scan data relates to. This (unknown / unidentified) subject is referred to as the “first subject”. At this stage, it is not known whether other scan data for the first subject is stored in the scan database.

[0074] The first representations (as defined herein) may respectively represent one or more of the first images. For example, at block 102, the method may comprise obtaining a set of first representations, Z, for each of a set of first images, I. Alternatively, at least one of the first representations may represent a plurality of the first images (e.g., a plurality of consecutive images). As a result of the method at block 102, a set, Z, of first representations may be obtained (e.g., Z = z2, ... Zj}).2025P00030WG

[0075] 9

[0076] The first images may be from one or more first scans performed on the first subject. The first images may be of teeth in an oral cavity of the first subject. For example, the scan may have been performed in an oral cavity of the first subject to obtain first images of the teeth of the first subject. It will be appreciated that, in such embodiments, the first scan data may be useful for storage in a database given that they may help identify the presence of plaque in the oral cavity of the first subject.

[0077] The first images may have been obtained using a scanning device comprising an imaging device. The scanning device may be an IOS. The imaging device may comprise an image sensorthat is sensitive to visible light. The imaging device may be a camera.

[0078] Examples of first images are shown in Figs. 2a and 2b, which are images of a set of teeth 202. In particular, Fig. 2a is an image of the set of teeth 202 in an oral cavity captured using a camera that is sensitive to visible light. The set of teeth 202 in Fig. 2a is illuminated using white light such that tooth enamel is detectable in the image (however, any plaque on the set of teeth 202 is barely visible to the human eye and / or the camera). Fig. 2b is an image of the same set of teeth 202 that is captured using the same camera. In contrast to Fig. 2a, the set of teeth 202 in Fig. 2b is illuminated by violet light, which can cause the tooth enamel and any plaque / tartar on the teeth to undergo Quantitative Light-induced Fluorescence (QLF) at different wavelengths. As such, any plaque / tartar on the tooth enamel may become visible to the human eye and / or a camera. For example, if there is no plaque / tartar present on the surface of the teeth at regions 204 and 206, these regions will appear, to the human eye, to have a green colour, caused by tooth enamel in these regions fluorescing at a first wavelength. In contrast, if there is plaque / tartar present on the surface of the teeth at regions 208 and 210, these regions will be seen, to the human eye, to have a red or orange colour, caused by plaque / tartar on the tooth enamel in these regions fluorescing at a longer, second wavelength. The skilled person will appreciate that storage of Figs. 2a and 2b in a scan database will be useful for the purpose of monitoring the oral health of a subject, particularly the presence of plaque / tartar on, cavities in, and / or gum recession experienced by the set of teeth 202.

[0079] The at least one first representation may comprise at least one first vector characterizing the first images. For example, the at least one first vector may be a latent vector or an embedded vector. In such embodiments, the method 100 may further comprise generating the at least one first vector by generating an embedding space which encodes an image, i, into a latent vector, z. As discussed previously, the at least one first vector may be generated using an embedding ML model, such as a neural network.

[0080] In some examples, acquiring the first representations may comprise generating the first representations. Alternatively, acquiring the first representations may comprise receiving the first representations, e.g., from an entity, device, or apparatus that generated the first representations (e.g., one or more of a computer, a server, a node of a cloud computing network, a host of a scan database, etc.).

[0081] At block 104 of Fig. 1, the method comprises acquiring at least one second representation (as defined herein) of second images in second scan data of a second subject. It will be appreciated that the term “at least one second representation” may be used inter-changeably herein with2025P00030WG

[0082] 10

[0083] the more general term “second representations”. The second scan data may be of known or identifiable provenance to a scan database in which the second scan data is stored, e.g., based on metadata available for the second scan data. For example, a subject identifier may be associated with the second scan data, or a scan identifier for the second scan data may itself be associated and stored with a subject identifier for the second subject. The subject identifier may be stored in a database in which the second scan data is stored. For example, the subject identifier may be incorporated in or stored in association with the second scan data (e.g., as metadata). The scan identifier and subject identifier may be stored in a database in which the second scan data is stored (e.g., as metadata).

[0084] The second representations may respectively represent one or more of the second images. For example, at block 104, the method may comprise obtaining a second representation z for each of K second images. Alternatively, the second representation may represent a plurality of the K second images (e.g., a plurality of consecutive images). As a result of the method at block 104, a set of second representations (e.g., Z = {z1(z2, ... zK}) may be obtained.

[0085] The second images may be from one or more second scans performed on a subject, who is referred to as the “second subject”. That is, the second data may comprise a set, S, of respective scans, s. of the second subject. For example, a set of scans for the second subject may be SU2= {s1(s2, ... sn}. The second images may be of teeth in an oral cavity of the second subject. For example, the one or more second scans may have been performed in an oral cavity of the second subject to obtain second images of the teeth of second subject. It will be appreciated that, in such embodiments, the second scan data may be useful for storage in a database given that they may help identify the presence of plaque in the oral cavity of the second subject. Figs. 2a and 2b may also be considered examples of second images.

[0086] The second images may have been obtained using a scanning device (e.g., in a similar manner to the first images). As such, the skilled person is referred to the discussion of block 102 for further detail regarding this scanning device. The scanning device used to capture the first images may be the same scanning device used to capture the second images, or the scanning devices may be different. If different scanning devices are used, it may be beneficial to acquire the representations of the first and second images using the embedding ML model, which is robust enough to manage and account for differences in the images caused by the different scanning devices, as discussed above. In this case, the scanning devices may be the same type or model, or they may be different. The first images may have been captured at a different time to the second images. In some cases, the second images and / or the second scan data is already stored in the scan database prior to the method at block 102 being performed.

[0087] The at least one second representation may comprise at least one second vector characterizing the second images. That is, they at least one second vector may be a vector representation. For example, the at least one second vector may be a latent vector or an embedded vector. In such embodiments, the method 100 may further comprise generating the at least one second vector by generating an embedding space which encodes one or more of the second images into latent vectors. As2025P00030WG

[0088] 11

[0089] discussed previously, the second vectors may be generated using an embedding ML model, such as a neural network.

[0090] In some examples, acquiring the second representations may comprise generating the second representations. Alternatively, acquiring the second representations may comprise receiving the second representations, e.g., from an entity, device, or apparatus that generated the second representations (e.g., one or more of a computer, a server, a node of a cloud computing network, a host of a scan database, etc.), or retrieving the second representations from the scan database.

[0091] At block 106 of Fig. 1, the method comprises determining, based on a differences between the at least one first representation and the at least one second representation, whether the first subject is the same as the second subject. For example, the first subject may be determined to be the same as the second subject based on whether the difference between the at least one first representation and the at least one second representations fulfills one or more criteria.

[0092] For example, the first images may be from a first scan performed on the first subject, and the second images may be from one or more second scans performed on the second subject. For example, the second images may be from a set of scans, SU2= {sx, s2, • • •5n} • The determination may then be based on inter-scan differences between the first scan and respective second scans of the one or more second scans.

[0093] An inter-scan difference between the first scan and a respective second scan may be based on one or more inter-representation differences between: at least one first representation of the first images from the first scan; and at least one second representation of second images from the respective second scan. First representations of first images from the first scan may be referred to herein as a “first set” of the first representations, and second representations of second images from a (respective) second scan may be referred to herein as a (respective) “second set” of the second representations.

[0094] It will be appreciated, for the determination of the inter-scan difference, the at least one first representation may represent a plurality (e.g., all) of the first images from the first scan, and the at least one second representation may represent a plurality (e.g., all) of the second images from the respective second scan. In this case, the inter-scan difference may be based on a single interrepresentation difference between a first representation representing a plurality (e.g., all) of the first images from the first scan and a second representation representing a plurality (e.g., all) of the second images from the respective second scan.

[0095] Alternatively, the inter-scan difference between the first scan and a respective second scan may be based on a plurality of inter-representation differences between representations included in the first set and the respective second set. Each inter-representation difference may be a difference between a first representation of the first set and a second representation of the respective second set.

[0096] In general, an inter-representation difference may be defined as a difference between two representations (e.g., one of the first representations and one of the second representations). An inter-2025P00030WG

[0097] 12

[0098] representation difference may correspond to, indicate, or otherwise quantify a difference, h. (in particular, a distance) between the images represented by the two representations. The inter-representation may correspond to the first metric discussed above. As an example, an inter-representation difference may be defined as the cosine similarity between the two representations, particularly when the representations are vectors (e.g., vector embeddings). The cosine similarity is a mathematical metric used to measure the similarity between two vectors in a multi-dimensional space. The cosine similarity is determined by calculating the cosine of the angle between the two vectors. The cosine similarity enables the relationship between the two vectors to be understood by considering the direction they are pointing in, rather than just comparing them based on their individual magnitudes or values.

[0099] Hence, the inter-representation difference, which corresponds to the difference, h. between two images,

[0100]

[0101] and i2, may be represented by equation 1 :

[0102]

[0103] where z1and z2are respective vector representations of and i2. It will be appreciated that the closer the value of h is to 1, the smaller the difference between z1and z2. That is, the larger the similarity between the representations, the smaller the difference between the representations.

[0104] The skilled person will also appreciate that other methods and techniques for determining an inter-representation difference between two representations, particularly two vector representations, exist and may be utilized in the embodiments described herein. For example, the distance can be determined as a Euclidean distance between z1and z2and / or any other suitable metric quantifying a distance between z1and z2.

[0105] In some embodiments, equation 1 may be used to process a given scan by removing, from the scan, images that are determined to be dissimilar to other images within the scan. That is, equation 1 may be used to identify, within a given scan obtained for a subject, images that are not of the given subject (e.g., they do not contain image data for the oral cavity of the subject).

[0106] At block 106, determining whether the first subject is the same as the second subject may comprise determining a distance between a first image represented by one of the first representations and a second image represented by one of the second representations. For example, at block 106, the method may comprise using equation 1 to determine a difference between the one of the first representations and the one of the second representations.

[0107] The inter-scan difference may then be based on an aggregate inter-representation difference. For example, the aggregate inter-representation difference may correspond to the second metric previously discussed. In such embodiments, the aggregate inter-representation difference may be determined by identifying pairs of representations (e.g., N pairs of representations), where each pair comprises one of the first representations of the first set and one of the second representations of the2025P00030WG

[0108] 13

[0109] respective second set, such that an inter-representation difference therebetween can be calculated using equation 1.

[0110] For simplicity, the “first images from the first scan” (i.e., the first set) may be identified as S-L and the “second images from a respective second scan” (i.e., the second set) may be identified as s2. However, it should be appreciated that not all of the first representations of the first images from the first scan need be included in s1?and not all of the second representations of the second images from the respective second scan need be included in s2.

[0111] The pairs of representations may be identified using the relatively simple approach of determining the cartesian product over s1and s2. That is, pairs of images included in s1and s2may be identified using equation 2:

[0112] Si X s2= {(%, y) \x G S-L and G b Equation 2 may result in all potential or possible pairs of representations in s1and s2being identified. Each pair of images (e.g., ij, ik) may correspond to a pair of representations of those images (e.g., Zj, zk). As such, the skilled person will appreciate that the pairs of representations may instead be identified, directly, by determining the cartesian product over the first set of first representations and the second set of second representations.

[0113] In another example, the pairs of representation may be identified by calculating or constructing a similarity matrix between the first set of first representations and the second set of second representations. A combinatorial optimization problem, such as an assignment problem (e.g., a linear assignment problem), may then be defined for the similarity matrix. An algorithm that solves the combinatorial optimization problem (e.g., a greedy algorithm, a global algorithm (e.g., a Hungarian algorithm), an auction algorithm, a linear sum assignment algorithm) may then be used to find suitable (e.g., optimal) one-to-one matches between the representations used to construct the similarity matrix.

[0114] The similarity matrix approach is useful as it means not every possible pair of representations representing images in in s1and s2need be considered when determining whether the first subject is the same as the second subject, which may help conserve computational resources. The similarity matrix approach is also useful as it may result in improved (e.g., optimal) matching of the representations included in the first set and the second set. For example, the similarity matrix approach allows first representations of the first set to be matched to (corresponding) second representations of the second set, particularly if the first scan data and second scan data were obtained according to similar or the same scanning configuration (i.e., a manner in which the scan data was obtained using a scanning device). Then, if the position of the scanning device when the scanning device was used to obtain the second images is known, the position of the scanning device when the scanning device was used to obtain the first images can be deduced. As such, plaque in an oral cavity detected in the first images can be2025P00030WG

[0115] 14

[0116] reported as present at specific locations within the oral cavity, based on the deduced position of the scanning device for each of the first images.

[0117] It will be appreciated that not all of the first representations in the first set may be paired to second representations of the second set (e.g., if the number of images in s1is smaller than or greater than the number of images in s2).

[0118] Once the pairs have been identified, the method of Fig. 1 may comprise determining, for each pair, an inter-representation difference between the representations included in the pair. For example, if the representations in the pairs are vector representations, this step may comprise determining a difference between the vectors included in the pair according to equation 1. For example, the interrepresentation determined for a given pair may be a cosine similarity between the two representations in the pair. As a result, a set of inter-representation differences (e.g., (h^, h2, ... hN}) may be obtained for a given second scan, where N is the number of identified pairs.

[0119] The method of Fig. 1 may then comprise aggregating the inter-representation differences obtained for a given second scan to obtain the aggregate inter-representation difference. The aggregate inter-representation difference g may represent a total or overall difference between the representations of images included s1and s2.

[0120] The aggregate inter-representation difference may be determined by aggregating the inter-representation differences. For example, the aggregate inter-representation difference may be calculated according to equation 3:

[0121] >

[0122]

[0123] for all matched pairs (ij,

[0124]

[0125] E (s1;s2). where N is the number of matched pairs. This aggregate interrepresentation difference may represent a difference or distance between s1and s2.

[0126] However, the skilled person will appreciate that the aggregate inter-representation difference may be any suitable metric that can be used to indicate or summarise a total, combined, or overall difference between a first scan and a second scan. For example, the aggregate inter-representation difference may be calculated using a function that identifies the minimum inter-representation difference from the set of inter-representation differences (e.g., a maximum cosine similarity) determined for a given second scan and / or takes into account the quality of the representations used to determine the set of interrepresentation differences determined for the given second scan.

[0127] As previously mentioned, the second images may be from (e.g., obtained from) a plurality or set, SU2, of second scans, performed on the second subject, and the determination at block 106 may be based on multiple inter-scan differences. In such examples, the method of Fig. 1 may comprise performing the above discussed approach for each of SU2= s2, s3, ... sn. That is, for a given second scan in SU2, pairs of representations may be identified (e.g., using equation 2). The method may then comprise2025P00030WG

[0128] 15

[0129] determining, for each identified pair for each second scan, inter-representation differences between the representations included in the pair. These inter-representation differences may be calculated using equation 1. As a result, a set of inter-representation differences (e.g.,

[0130]

[0131] h2, ... hN}) may be obtained for each second scan, where N is the number of identified pairs for a given second scan. Furthermore, a set of aggregate inter-representation difference (e.g., {g2, 93- ... 9N}) may be obtained for the plurality of second scans.

[0132] The method of Fig. 1 may then comprise determining whether the first subject is the same as the second subject by identifying, from the set of inter-scan differences, a minimum inter-scan difference. The minimum inter-scan difference may correspond to a maximum inter-scan similarity. For example, the first subject may be determined to be the same as the second subject based on whether the minimum inter-scan difference meets the one or more criteria. The one or more criteria may comprise a minimum threshold difference. If the minimum inter-scan difference meets the threshold, the first subject may be determined to be the same as the second subject.

[0133] For example, the method of Fig. 1 may comprise determining a scan-subject similarity between the first scan, s1?and the plurality of second scans of the second subject, SU2(e.g., all second scans of the second subject that are stored in a database). The scan-subject similarity may correspond to the third metric discussed above. The scan-subject similarity may then represent the similarity between a new scan (with unknown provenance) and SU2. As an example, a scan-subject similarity, tu, for a given second subject, it, may be defined according to equation 4:

[0134]

[0135] for all s 6 Su. If tufor the second subj ect (e.g., tu) meets the one or more criteria (e .g . , a threshold), the first subject may be determined to be the same as the second subject. The maximum scan-subject similarity may correspond to the minimum scan-subject difference.

[0136] However, the skilled person will appreciate that the scan-subject similarity may be any suitable metric that can be used to indicate or summarise a total, combined, or overall similarity between the first scan data and the second scan data. For example, the second aggregate difference may be a function that identifies the mean of the aggregate inter-representation differences and / or takes into account the quality of the representations used to determine the aggregate inter-representation differences.

[0137] It will be appreciated that in embodiments disclosed herein, the method of Fig. 1 may be repeated, where the second subject is a different subject for each repeat or iteration of the method (e.g., a different subject for whom second scan data is stored in a database). That is, the method of Fig. 1 may be repeated to compare the first scan data of the first subject to scan data of any subject that is stored in a database (e.g., any of the subjects included in the set {it2, u3... um)). In such embodiments, a set of scan-subject similarities may be obtained (e.g., {tM, tu, ... tu}). That is, a scan-subject similarity may be obtained for each of the (different) second subjects, via repeating the method of Fig. 1.2025P00030WG

[0138] 16

[0139] To determine which, if any, of the second subjects is the same as the first subject, the second subject having second scan data with the maximum similarity to the first scan data may be identified. This second subject may be identified as umax. and may correspond to the second subject for whom tuis maximised. As such, umaxmay be determined by identifying

[0140]

[0141] where

[0142]

[0143] (equation 5).

[0144] umflxmay then be determined by identifying the second subject associated with tmax.

[0145] However, it will be appreciated that, even if umaxis identified, this second subject may not actually be the same as the first subject. For example, no scan data of the first subject may be already stored in the database (e.g., the first subject is a new user of the database), meaning umaxcannot be the same as the first subject, but instead simply represents the closest match. In such scenarios, the first subject should not be determined to be the same as any of the set {it2, u3... um). Similarly, in some circumstances, multiple (different) second subjects (e.g., uxand uy) may be determined to be same as the first subject via the method of Fig. 1. For example, the second scan data of uxand uymay have a similar difference or distance to the first scan data (e.g.,

[0146]

[0147] To account for these and similar circumstances, a threshold tmaxvalue, tthreshoid-maY be defined. Then, tmaxmay need to meet tthreshoid inorder for the first subject to be determined to be the same subject as the subject associated with the scan data that resulted in umax. That is, (only) if tmaxis equal to or greater than tthreshoid, a match between the first subject and the subject associated with the scan data that resulted in umaxmay be considered to be identified.

[0148] It will be appreciated that, if tudetermined for each second subject is below tthreshoid, the first scan data may be considered to be scan data of a subject for whom scan data is not currently stored in a database (e.g., the first subject is a new user). As such, the first scan data may be discarded or stored in the database with an indication that the identity of the first subject is unknown in the database, as previously discussed.

[0149] tthreshoidmaY be determined manually (e.g., heuristically), or based on second scan data stored in the database, to determine a typical similarity of second scans included in second scan data for a given second subject. For example, the second scans stored in the database for a given second subject may be compared, to determine their typical similarity. For example, the method of Fig. 1 may be performed using (only) second scan data for a second subject for whom second scan data is stored in the database, where the “first scan data” is taken to be one of the second scans of the second scan data (i.e., a known scan of the second subject), such that a value of tmaxis obtained, which indicates an example similarity between second scan data of the given subject. This value may be used as tthreshoid.2025P00030WG

[0150] 17

[0151] However, a more robust approach would be to repeat the above discussed method for a plurality of second subjects for whom second scan data is stored in a database, such that a set of tmaxvalues are obtained, threshold may be determined based on this set of tmaxvalues. For example, ^threshold may be the mean or average of the these values of tmax. which would indicate an expected similarity between second scan data for a given subject.

[0152] As such, tthreshoidcan be defined such that, when the method of Fig. 1 is performed for a known subject, the known subject is correctly identified based on its scan data and the scan data of second subjects stored in a database. This may involve a known mapping between second subjects and second scan data stored in a database being iteratively removed and re-determined via a method corresponding to the method discussed with reference to Fig. 1.

[0153] Note that the above method for determining tthreshoidmay also indicate how complex the second scan data is and how well the method of Fig. 1 determines similarity of subjects for whom scan data is acquired. For example, it may indicate how well the first and / or second scan data is represented by (e.g., embedded in) the first and / or second representations, respectively. Similarly, it may indicate how well the distance (metric) space is defined.

[0154] If the first subject is determined to be the same as the second subject, the method of Fig. 1 may further comprise storing, in a database: the first scan data; and an association between the first scan data and the identity of the second subject. The association may be indicated by a subject identifier stored in or with the first scan data. Additionally or alternatively, the association may be indicated by a scan identifier, and the database may include information identifying a mapping between subject identifiers and scan identifiers. For example, the first scan data may be stored in the database along with metadata (e.g. an identifier for the second subject) that indicates that the first scan data comprises scan data of the second subject. In this way, the subject of the first scan data has been identified.

[0155] If the first subject is determined not to be the same as the second subject, the method of Fig. 1 may further comprise one of: discarding the first scan data; and storing, in a database, the first scan data and an indication that the identity of the first subject is (at least currently) unknown in the database. The indication may indicate an identity of a new user of the database. The indication may act as a placeholder identity until further (meta)data is obtained from the first subject.

[0156] It will be appreciated that the embodiments described herein provide methods for matching scan data with unknown provenance to scan data with known provenance. For example, the scan data with known provenance may be stored in a database in which the scan data with unknown provenance is stored or is to be stored.

[0157] The embodiments described herein may be useful for determining whether subjects of oral scan data are the same. Advantageously, this facilitates the (re-)identification of a subject of scan data obtained from a device operable to image or scan oral cavities (e.g., IOS devices). Thus, such embodiments may form part of an integrated solution for imaging, detecting, and storing data regarding2025P00030WG

[0158] 18

[0159] the health of an oral cavity of a subject. It should be noted that the embodiments described provide robust processing steps that enable scan data for a wide range of subjects to be stored in (and / or retrieved from) a scan database (e.g., the methods described herein are robust enough to handle “anomalies” in scan data of oral cavities, such as braces and implants).

[0160] For example, the first scan data may be useful to store in a database as it may facilitate the detection of plaque in an oral cavity of the first subject, particularly if the first scan data was obtained whilst tooth enamel and / or plaque in the oral cavity of the first subject was caused to undergo QLF (e.g., by being illuminated with violet light having a peak wavelength between 400 nm and 405 nm, such as using light emitted from a blue Light Emitting Diode (LED)). Whilst undergoing QLF, plaque may have a red / orange appearance whilst tooth enamel may have a green appearance, making the plaque easier to detect in comparison to when the oral cavity is illuminated using white light and imaged using a standard camera. If the presence of plaque in an oral cavity can be identified using the first scan data, this information can be used to determine a brushing configuration for the first subject. The brushing configuration may be used to guide the first subject on where they should be brushing more effectively. This is useful, as unremoved plaque can turn into tartar. Tartar build-up in oral cavities can lead to serious dental issues such as gingivitis and periodontitis.

[0161] Furthermore, the first scan data may be useful to store in the database as it may facilitate the differentiation of plaque and tartar in an oral cavity, which is challenging if only an individual image or scan of the oral cavity is available. That is, plaque in an oral cavity can be differentiated from tartar in the oral cavity by observing multiple scans of the oral cavity obtained over time, as plaque changes over time whilst tartar build-up is more constant / stable.

[0162] To provide further detail, an example implementation of the method of Fig. 1 is now described. Steps 102 and 104 may be performed such that each tooth in first images of an oral cavity of an (unknown) first subject is converted into vector embeddings. For example, a set of images {i1(... , ij} forming the first scan data may be converted into vector embeddings and a set of images {i i> ■■■>

[0163]

[0164] forming the second scan data may be converted into vector embeddings, such as through using a selfsupervised Al or ML neural network. Step 106 may then be performed (e.g., using a matching linear algorithm) such that the first subject is determined to be the second subject by (re-)identifying the teeth of the first subject in the second images of the second subject (e.g., identifying that the first and second images may comprise images of the same set of teeth). This may be based on the localization of the teeth of the first / second subject in the first / second images. A global distance metric between two scan of the first and second scan data may be defined, where the global distance metric is based on a combination of distances in an embedding space. The first subject may be determined to be the second subject if the global distance metric meets a specific identification threshold. However, the skilled person will appreciate that the embodiments described herein may be used for determining whether subjects of scan data are the same based on any suitable type of scan data (e.g., scan data of skin regions of the subjects).2025P00030WG

[0165] 19

[0166] For example, the embodiments described herein may be used for the (re ^identification of a subject of skin region scan data, which may be useful if the subject is undergoing multiple Intense Pulsed Light (IPL) treatment sessions. Such embodiments may rely on scale -invariant feature transform (SIFT) features of skin regions that are in (e.g., imaged or captured by) the first and second images. The first and second representations may characterise the SIFT features, and differences between these representations may indicate whether these SIFT features are the same or different (e.g., whether they are different because they have simply been shifted, rotated, and / or scaled between different images, or because they characterise skin regions of different subjects).

[0167] Fig. 3 is a schematic diagram illustrating an apparatus 300 for processing scan data to facilitate identification of a first subject of first scan data according to an embodiment. The apparatus 300 may comprise one or more computing / processing devices, such as one or more of a computer, a server, a node of a cloud computing network, a wireless device (e.g. a mobile phone, a smart phone, a tablet, a laptop, or any other wireless device), and any other suitable computing / processing device. The apparatus 300 may host a scan database, or be communicatively coupled to a host of the scan database. The apparatus 300 may be communicatively coupled (directly or indirectly) to a scanning device used to obtain the first images and / or the second images.

[0168] As illustrated in Fig. 3, the apparatus 300 comprises one or more processors 302. The one or more processors 302 can be implemented in numerous ways, with software and / or hardware, to perform the various functions described herein. The one or more processors 302 can comprise a plurality of software and / or hardware modules, each configured to perform, or that are for performing, individual or multiple steps of the method described herein.

[0169] The one or more processors 302 may comprise, for example, one or more microprocessors, one or more multi -core processors and / or one or more digital signal processors (DSPs), one or more processing units, and / or one or more controllers (e.g. one or more microcontrollers) that may be configured or programmed (e.g. using software or computer program code) to perform the various functions described herein. The one or more processors 302 may be implemented as a combination of dedicated hardware (e.g. amplifiers, pre-amplifiers, analog -to-digital convertors (ADCs) and / or digital -to-analog convertors (DACs)) to perform some functions and one or more processors (e.g. one or more programmed microprocessors, DSPs and associated circuitry) to perform other functions.

[0170] The one or more processors 302 can be configured to perform the method described herein, such as the embodiments of the method described with reference to Fig. 1. For example, the one or more processors 302 may be configured to cause the apparatus 300 to perform the embodiments of the method described herein, such as the embodiments of the method described with reference to Fig. 1.

[0171] As illustrated in Fig. 3, the apparatus 300 may comprise at least one memory 306.

[0172] Alternatively or in addition, at least one memory 306 may be external to (e.g. separate to or remote from) the apparatus 300. For example, another apparatus may comprise at least one memory 306 according to some embodiments. A hospital database may comprise at least one memory 306, at least one memory 3062025P00030WG

[0173] 20

[0174] may be a cloud computing resource, or similar. The one or more processors 302 of the apparatus 300 may be configured to communicate with and / or connect to at least one memory 306. The at least one memory 306 may comprise any type of non-transitory machine-readable medium, such as cache or system memory including volatile and non-volatile computer memory such as random access memory (RAM), static RAM (SRAM), dynamic RAM (DRAM), read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), and electrically erasable PROM (EEPROM). At least one memory 306 can be configured to store program code that can be executed by the one or more processors 302 of the apparatus 300 to cause the apparatus 300 to operate in the manner described herein.

[0175] Alternatively or in addition, at least one memory 306 can be configured to store information required by or resulting from the method described herein. For example, at least one memory 306 may be configured to store any one or more of the first representations, the second representations, the first images, the second images, and any other information, or any combination of information, required by or resulting from the method described herein. The memory may form at least a part of the scan database. The one or more processors 302 of the apparatus 300 can be configured to control at least one memory 306 to store information required by or resulting from the method described herein.

[0176] As illustrated in Fig. 3, the apparatus 300 may comprise at least one user interface 308. Alternatively or in addition, at least one user interface 308 may be external to (e.g. separate to or remote from) the apparatus 300. The one or more processors 302 of the apparatus 300 may be configured to communicate with and / or connect to at least one user interface 308. One or more processors 302 of the apparatus 300 can be configured to control at least one user interface 308 to operate in the manner described herein.

[0177] A user interface 308 can be configured to render (or output, display, or provide) information required by or resulting from the method described herein. For example, one or more user interfaces 308 may be configured to render (or output, display, or provide) one or more of the first representations, the second representations, the first images, the second images, or any other information, or any combination of information, required by or resulting from the method described herein.

[0178] Alternatively or in addition, one or more user interfaces 308 can be configured to receive a user input. For example, one or more user interfaces 308 may allow a user to manually enter information or instructions, interact with and / or control the apparatus 300. Thus, one or more user interfaces 308 may be any one or more user interfaces that enable the rendering (or outputting, displaying, or providing) of information and / or enables a user to provide a user input.

[0179] The user interface 308 may comprise one or more components for this. For example, one or more user interfaces 308 may comprise one or more switches, one or more buttons, a keypad, a keyboard, a mouse, a display or display screen, a graphical user interface (GUI) such as a touch screen, an application (e.g. on a smart device such as a tablet, a smart phone, or any other smart device), or any other visual component, one or more speakers, one or more microphones or any other audio component, one or more lights (e.g. one or more light emitting diodes, LEDs), a component for providing tactile or haptic2025P00030WG

[0180] 21

[0181] feedback (e.g. a vibration function, or any other tactile feedback component), a smart device (e.g. a smart mirror, a tablet, a smart phone, a smart watch, or any other smart device), or any other user interface, or combination of user interfaces. One or more user interfaces that are controlled to render information may be the same as one or more user interfaces that enable the user to provide a user input.

[0182] As illustrated in Fig. 3, the apparatus 300 may comprise at least one communications interface (or communications circuitry) 310. Alternatively or in addition, at least one communications interface 310 may be external to (e.g. separate to or remote from) the apparatus 300. A communications interface 310 can be for enabling the apparatus 300, or components of the apparatus 300 (e.g. one or more processors 302, one or more sensors 304, one or more memories 306, one or more user interfaces 308 and / or any other components of the apparatus 300), to communicate with and / or connect to each other and / or one or more other components (e.g., components of a host of the scan database). For example, one or more communications interfaces 310 can be for enabling one or more processors 302 of the apparatus 300 to communicate with and / or connect to one or more sensors 304, one or more memories 306, one or more user interfaces 308 and / or any other components of the apparatus 300.

[0183] The communications interface 310 may enable the apparatus 300, or components of the apparatus 300, to communicate and / or connect in any suitable way. For example, one or more communications interfaces 310 may enable the apparatus 300, or components of the apparatus 300, to communicate and / or connect wirelessly, via a wired connection, or via any other communication (or data transfer) mechanism. In some wireless embodiments, for example, one or more communications interfaces 310 may enable the apparatus 300, or components of the apparatus 300, to use radio frequency (RF), Bluetooth, or any other wireless communication technology to communicate and / or connect.

[0184] Fig. 4 is a block diagram illustrating an example processor 400 according to embodiments of the disclosure. Processor 400 may be used to implement one or more processors described herein, for example, processor 302 shown in Fig. 300. Processor 400 may be any suitable processor type including, but not limited to, a microprocessor, a microcontroller, a digital signal processor (DSP), a field programmable array (FPGA) where the FPGA has been programmed to form a processor, a graphical processing unit (GPU), an application specific circuit (ASIC) where the ASIC has been designed to form a processor, or a combination thereof.

[0185] The processor 400 may include one or more cores 402. The core 402 may include one or more arithmetic logic units (ALU) 404. In some embodiments, the core 402 may include a floating point logic unit (FPLU) 406 and / or a digital signal processing unit (DSPU) 408 in addition to or instead of the ALU 404.

[0186] The processor 400 may include one or more registers 412 communicatively coupled to the core 402. The registers 412 may be implemented using dedicated logic gate circuits (e.g., flip-flops) and / or any memory technology. In some embodiments the registers 412 may be implemented using static memory. The register may provide data, instructions and addresses to the core 402. In some embodiments, processor 400 may include one or more levels of cache memory 410 communicatively2025P00030WG

[0187] 22

[0188] coupled to the core 402. The cache memory 410 may provide computer-readable instructions to the core 402 for execution. The cache memory 410 may provide data for processing by the core 402. In some embodiments, the computer-readable instructions may have been provided to the cache memory 410 by a local memory, for example, local memory attached to the external bus 416. The cache memory 410 may be implemented with any suitable cache memory type, for example, metal -oxide semiconductor (MOS) memory such as static random access memory (SRAM), dynamic random access memory (DRAM), and / or any other suitable memory technology.

[0189] The processor 400 may include a controller 414, which may control input to the processor 400 from other processors and / or components included in a system and / or outputs from the processor 400 to other processors and / or components included in the system. Controller 414 may control the data paths in the ALU 404, FPLU 406 and / or DSPU 408. Controller 414 may be implemented as one or more state machines, data paths and / or dedicated control logic. The gates of controller 414 may be implemented as standalone gates, FPGA, ASIC or any other suitable technology. The registers 412 and the cache 410 may communicate with controller 414 and core 402 via internal connections 420A, 420B, 420C and 420D. Internal connections may implemented as a bus, multiplexor, crossbar switch, and / or any other suitable connection technology.

[0190] Inputs and outputs for the processor 400 may be provided via a bus 416, which may include one or more conductive lines. The bus 416 may be communicatively coupled to one or more components of processor 400, for example the controller 414, cache 410, and / or register 412. The bus 416 may be coupled to one or more components of the system.

[0191] The bus 416 may be coupled to one or more external memories. The external memories may include Read Only Memory (ROM) 432. ROM 432 may be a masked ROM, Electronically Programmable Read Only Memory (EPROM) or any other suitable technology. The external memory may include Random Access Memory (RAM) 433. RAM 433 may be a static RAM, battery backed up static RAM, Dynamic RAM (DRAM) or any other suitable technology. The external memory may include Electrically Erasable Programmable Read Only Memory (EEPROM) 435. The external memory may include Flash memory 434. The External memory may include a magnetic storage device such as disc 436. In some embodiments, the external memories may be included in a system, such as a system comprising the apparatus 300 and a host of the scan database.

[0192] There is provided a computer program comprising instructions which, when executed by a processor (such as one or more processors 302 of the apparatus 300 or the processor 402), cause the processor to perform at least part of, or all of, the method described herein.

[0193] There is provided a computer program product comprising a computer readable medium. The computer readable medium has a computer readable code embodied therein. The computer readable code is configured such that, on execution by a suitable computer or processor (such as one or more processors 302 of the apparatus 300 or the processor 402), the computer or processor is caused to perform the method described herein. The computer readable medium may be, for example, any entity or device2025P00030WG

[0194] 23

[0195] capable of carrying the computer program product. For example, the computer readable medium may include a data storage, such as a ROM (such as a CD-ROM or a semiconductor ROM) or a magnetic recording medium (such as a hard disk). Furthermore, the computer readable medium may be a transmissible carrier, such as an electric or optical signal, which may be conveyed via electric or optical cable or by radio or other means. When the computer program product is embodied in such a signal, the computer readable medium may be constituted by such a cable or other device or means. Alternatively, the computer readable medium may be an integrated circuit in which the computer program product is embedded, the integrated circuit being adapted to perform, or used in the performance of, the method described herein.

[0196] There is thus provided herein an apparatus, method, and computer program product of processing scan data to facilitate identification of a first subject of first scan data, which address the limitations associated with the existing techniques.

[0197] It will be understood that at least some or all of the method steps described herein can be automated. That is, at least some or all of the method steps described herein can be performed automatically. The method described herein can be a computer-implemented method.

[0198] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments.

[0199] Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the principles and techniques described herein, from a study of the drawings, the disclosure and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfil the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. A computer program may be stored or distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

Claims

2025P00030WG24CLAIMS:

1. A computer-implemented method (100) of processing scan data to facilitate identification of a first subject of first scan data, the method comprising:acquiring (102) at least one first representation of first images in the first scan data; acquiring (104) at least one second representation of second images in second scan data of a second subject; anddetermining (106), based on a difference between the at least one first representation and the at least one second representation, whether the first subject is the same as the second subject.

2. The method as claimed in any claim 1, wherein the first subject is determined to be the same as the second subject based on whether the difference between the at least one first representation and the at least one second representation fulfills one or more criteria.

3. The method as claimed in claim 1 or 2, wherein the first images are from a first scan performed on the first subject, and wherein the second images are from one or more second scans performed on the second subject, and the determination is based on inter-scan differences between the first scan and respective second scans of the one or more second scans.

4. The method as claimed in claim 3, wherein an inter-scan difference between the first scan and a respective second scan is based on one or more inter-representation differences between: at least one first representation of first images from the first scan; and at least one second representation of second images from the respective second scan.

5. The method as claimed in claim 4, wherein the inter-scan difference is based on an aggregate inter-representation difference.

6. The method as claimed in claim 5, wherein the aggregate inter-representation difference is determined by:identifying pairs of representations, each pair comprising one of the first representations of the first images from the first scan and one of the second representations of the second images from the respective second scan;determining, for each pair, an inter-representation difference between the representations included in the pair; and2025P00030WG25aggregating the inter-representation differences to obtain the aggregate interrepresentation difference.

7. The method as claimed in claim 6, wherein aggregating the inter-representation differences comprises determining a mean of the inter-representation differences.

8. The method as claimed in any of claims 3 to 7, wherein determining whether the first subject is the same as the second subject comprises identifying, from the inter-scan differences, a minimum inter-scan difference.

9. The method as claimed in any preceding claim, wherein if the first subject is determined to be the same as the second subject, the identity of the first subject is determined to be the same as an identity of the second subject.

10. The method as claimed in claim 9, wherein if the first subject is determined to be the same as the second subject, the method further comprises storing, in a database:the first scan data; andan association between the first scan data and the identity of the second subject.

11. The method as claimed in claim 9 or 10, wherein if the first subject is determined not to be the same as the second subject, the method further comprises one of:discarding the first scan data; andstoring, in a database, the first scan data and an indication that the identity of the first subject is unknown in the database.

12. The method as claimed in any one of the preceding claims, wherein the at least one first representation comprises at least one first vector characterizing the first images and the at least one second representation comprises at least one second vector characterizing the second images.

13. The method as claimed in any one of the preceding claims, wherein the first images are of teeth in an oral cavity of the first subject, and / or the second images are of teeth in an oral cavity of the second subject.

14. An apparatus (300) for processing scan data to facilitate identification of a first subject of first scan data, the apparatus comprising processing circuitry (302) configured to cause the apparatus to:acquire (102) at least one first representation of first images in the first scan data;2025P00030WG26acquire (104) at least one second representation of second images in second scan data of a second subject; anddetermine (106), based on a difference between the at least one first representation and the at least one second representation, whether the first subject is the same as the second subject.

15. A computer program product comprising a computer readable medium, the computer readable medium having a computer readable code embodied therein, the computer readable code being configured such that, on execution by a suitable computer or processor, the computer or processor is caused to perform the method as claimed in any of claims 1 to 13.

Images