Method to generate oral care advice

Abstract

The subject-matter of the present disclosure relates to a computer implemented method (400) to generate oral care advice. The method comprises receiving and storing (401) multimodal oral care data corresponding to a population of users of an oral care device, generating (402) population clusters based on the multimodal oral care data using a machine learning model, receiving (403), for a particular user in the population, an oral health goal, generating (404) oral care advice by performing collaborative filtering based on the generated population clusters and the received oral health goal, and controlling (405) an electronic apparatus to output the generated oral care advice.

Description

[0001] METHOD TO GENERATE ORAL CARE ADVICE

[0002] FIELD

[0003]

[0001] The subject-matter of the present disclosure relates to a computer implemented method for generating oral care advice, and a system therefore. In particular, the present disclosure relates to generate advice on oral care routines that users might perform using, e.g., a toothbrush, and device operation(s) pertaining thereto.

[0004] BACKGROUND

[0005]

[0002] Oral health refers to a healthy state of the oral cavity and related tissues that enables an individual to chew, speak, and confidently socialize without pain, discomfort, and or embarrassment. The maintenance of good oral health is integral to maintaining an individual’s overall health and negligence in this regard has profound consequences on one’s quality of life. Personalized oral health programs aim to improve the oral health of a targeted population group by implementing oral care behavioural changes.

[0006]

[0003] A large body of evidence has demonstrated that oral hygiene and diseases have strong causal relationships with a person’s dietary habits, salivary status, medical history, genetic factors as well as the socioeconomic status. Similarly, smoking, alcohol consumption, unbalanced diet, and psychological factors are among the oral health risks that are common to many chronic diseases, for example, cardiovascular diseases, type2 diabetes and vision problems.

[0007]

[0004] Effective oral self-care behaviour is fundamental for the successful prevention and treatment of various oral diseases. However, standardised general advice provided by dental practitioners (e.g., two minutes of brunching twice a day with fluoride toothpaste combined with flossing once per day) can be inadequate for meeting the complex needs of a society comprised of various diverse groups. Complex demographic groups include, for example, elderly, children, women during pregnancy, persons diagnosed with behavioural disorders, individuals going through chronic illnesses or health-conditions, people with gum sensitivities, population groups with dietary restrictions and preferences, people with impaired self-care capacity, and so on.

[0008]

[0005] As such, it is now desired to use technology to provide targeted, personalised, oral care guidance as an improvement over the prior art to drive oral health benefits at scale. SUMMARY

[0009]

[0006] The present invention is defined according to the claims.

[0010]

[0007] According to a first aspect of the present invention, there is provided a computer implemented method to generate oral care advice. The method comprises receiving and storing multimodal oral care data corresponding to a population of users of an oral care device, generating population clusters based on the multimodal oral care data using a machine learning model, receiving, for a particular user in the population, an oral health goal, generating oral care advice by performing collaborative filtering based on the generated population clusters and the received oral health goal, and controlling an electronic apparatus to output the generated oral care advice.

[0011]

[0008] In this way, the method allows for convenient and robust generation of personalised oral care advice, based on oral care behaviours, and controls, etc, which have been determined to be effective for a cluster of the population to which the particular user is most similar.

[0012]

[0009] The oral health goal may be, for example, at least one of even teeth coverage, teeth whitening, fresh breath, managing gum sensitivity, reducing risk of dental caries, and minimising dental plaque.

[0013]

[0010] The multimodal oral care data may comprise usage information of the oral care device (optionally captured by a sensor of the oral care device) and at least one of: dietary information of the user, demographic information of the user, oral health goal information, and oral health tracking information.

[0014]

[0011] The usage information may comprise at least one of time of use, pressure applied during use, and duration of use, operating signature, mouth shape, teeth shape, angle of use, pressure applied during use.

[0015]

[0012] Dietary information may comprise at least one of eating times, eating duration, and macro-nutrients in a user’s diet, and meal logs.

[0016]

[0013] Oral health tracking information may comprises data captured by at least one of an imaging sensor and a wearable oral sensor.

[0017]

[0014] Generating the oral care advice may comprise ranking a plurality of options for output as the oral care advice based on the oral health tracking information, and selecting the highest ranked option as the oral health advice.

[0015] The generated oral care advice may comprise at least one of a change in oral care behaviour and a change in oral care device usage.

[0018]

[0016] The generated oral care advice may comprise a recommended operating setting of the oral care device, and the method may comprise controlling an operating state of the oral care device based on the recommended operating setting.

[0019]

[0017] The generated oral care advice may comprise coaching information on a usage technique of the oral care device, and the method further comprise determining that the oral care device is in operation, and wherein outputting the generated oral care advice comprising coaching information comprises outputting the coaching information in real time during operation of the oral care device.

[0020]

[0018] In a second aspect of the present invention, there is provided a system for generating oral care advice. The system comprises an oral care device, an electronic apparatus comprising at least one of a display and a speaker, and a first communicator for communicating with a server. The server comprises a second communicator configured to receive multimodal oral care data corresponding to a population of users of an oral care device, and an oral health goal for a user from the electronic apparatus, a storage for storing the multimodal oral care data, and a processor configured to: generate population clusters based on the multimodal oral care data using a machine learning model, generate oral care advice by performing collaborative filtering based on the generated population clusters and the received oral health goal, and control the second communicator to transmit the generated oral care advice to the electronic apparatus, and a signal for controlling the electronic apparatus to output the generated oral care advice on the at least one of the display and the speaker.

[0021]

[0019] The electronic apparatus may be communicatively coupled to the oral care device, and where the generated oral care advice comprises a recommended operating setting of the oral care device, the electronic apparatus may be configured to transmit, to the oral care device, a signal for controlling an operating state of the oral care device according to the recommended operating setting.

[0022]

[0020] These and other aspects of the present invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.

[0023]

[0021] As will be appreciated by one skilled in the art, the present techniques may be embodied as a system, method or computer program product. Accordingly, present techniques may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.

[0024]

[0022] Furthermore, the present techniques may take the form of a computer program product embodied in a computer readable medium having computer readable program code embodied thereon. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable medium may be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing.

[0025]

[0023] Computer program code for carrying out operations of the present techniques may be written in any combination of one or more programming languages, including object oriented programming languages and conventional procedural programming languages. Code components may be embodied as procedures, methods or the like, and may comprise sub-components which may take the form of instructions or sequences of instructions at any of the levels of abstraction, from the direct machine instructions of a native instruction set to high-level compiled or interpreted language constructs.

[0026]

[0024] Embodiments of the present techniques also provide a non-transitory data carrier carrying code which, when implemented on a processor, causes the processor to carry out any of the methods described herein.

[0027]

[0025] The techniques further provide processor control code to implement the abovedescribed methods, for example on a general purpose computer system or on a digital signal processor (DSP). The techniques also provide a carrier carrying processor control code to, when running, implement any of the above methods, in particular on a non- transitory data carrier. The code may be provided on a carrier such as a disk, a microprocessor, CD- or DVD-ROM, programmed memory such as non-volatile memory (e.g. Flash) or read-only memory (firmware), or on a data carrier such as an optical or electrical signal carrier. Code (and / or data) to implement embodiments of the techniques described herein may comprise source, object or executable code in a conventional programming language (interpreted or compiled) such as Python, C, or assembly code, code for setting up or controlling an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array), or code for a hardware description language such as Verilog (RTM) or VHDL (Very high speed integrated circuit Hardware Description Language). As the skilled person will appreciate, such code and / or data may be distributed between a plurality of coupled components in communication with one another. The techniques may comprise a controller which includes a microprocessor, working memory and program memory coupled to one or more of the components of the system.

[0028]

[0026] It will also be clear to one of skill in the art that all or part of a logical method according to embodiments of the present techniques may suitably be embodied in a logic apparatus comprising logic elements to perform the steps of the above-described methods, and that such logic elements may comprise components such as logic gates in, for example a programmable logic array or application-specific integrated circuit. Such a logic arrangement may further be embodied in enabling elements for temporarily or permanently establishing logic structures in such an array or circuit using, for example, a virtual hardware descriptor language, which may be stored and transmitted using fixed or transmittable carrier media.

[0029]

[0027] In an embodiment, the present techniques may be realised in the form of a data carrier having functional data thereon, said functional data comprising functional computer data structures to, when loaded into a computer system or network and operated upon thereby, enable said computer system to perform all the steps of the above-described method.

[0030]

[0028] The methods described above may be wholly or partly performed on an apparatus, i.e. an electronic device, using a machine learning or artificial intelligence model. The model may be processed by an artificial intelligence-dedicated processor designed in a hardware structure specified for artificial intelligence model processing. The artificial intelligence model may be obtained by training. Here, "obtained by training" means that a predefined operation rule or artificial intelligence model configured to perform a desired feature (or purpose) is obtained by training a basic artificial intelligence model with multiple pieces of training data by a training algorithm. The artificial intelligence model may include a plurality of neural network layers. Each of the plurality of neural network layers includes a plurality of weight values and performs neural network computation by computation between a result of computation by a previous layer and the plurality of weight values.

[0031]

[0029] As mentioned above, the present techniques may be implemented using an Al model. A function associated with Al may be performed through the non-volatile memory, the volatile memory, and the processor. The processor may include one or a plurality of processors. At this time, one or a plurality of processors may be a general purpose processor, such as a central processing unit (CPU), an application processor (AP), or the like, a graphics-only processing unit such as a graphics processing unit (GPU), a visual processing unit (VPU), and / or an Al-dedicated processor such as a neural processing unit (NPU). The one or a plurality of processors control the processing of the input data in accordance with a predefined operating rule or artificial intelligence (Al) model stored in the non-volatile memory and the volatile memory. The predefined operating rule or artificial intelligence model is provided through training or learning. Here, being provided through learning means that, by applying a learning algorithm to a plurality of learning data, a predefined operating rule or Al model of a desired characteristic is made. The learning may be performed in a device itself in which Al according to an embodiment is performed, and / or may be implemented through a separate server / system.

[0032]

[0030] The Al model may consist of a plurality of neural network layers. Each layer has a plurality of weight values, and performs a layer operation through calculation of a previous layer and an operation of a plurality of weights. Examples of neural networks include, but are not limited to, convolutional neural network (CNN), deep neural network (DNN), recurrent neural network (RNN), restricted Boltzmann Machine (RBM), deep belief network (DBN), bidirectional recurrent deep neural network (BRDNN), generative adversarial networks (GAN), and deep Q-networks.

[0033]

[0031] The learning algorithm is a method for training a predetermined target device (for example, a robot) using a plurality of learning data to cause, allow, or control the target device to make a determination or prediction. Examples of learning algorithms include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0034] BRIEF DESCRIPTION OF DRAWINGS

[0035]

[0032] The embodiments of the present inventions may be best understood with reference to the accompanying figures, in which:

[0036]

[0033] Figure 1 shows an example oral care device;

[0037]

[0034] Figure 2 shows an example system for generating oral care advice; and

[0038]

[0035] Figure 3 shoes an example method for generating oral care advice;

[0039]

[0036] Figure 4 shows further steps of an example method for generating oral care advice. DESCRIPTION OF EMBODIMENTS

[0040]

[0037] Figure 1 shows an example of a personal care device. In the following example embodiments the personal care device is embodied as an oral care device 100, such as a toothbrush, which is generally used for the maintenance of oral health such as gum and teeth hygiene. It will however be appreciated that the example embodiments are not limited thereto and may be readily applied to any personal care device having the requisite features and functionality.

[0041]

[0038] The oral care device 100 comprises a handle 102 and an operational head 104. The operational head may take a variety of forms depending on the intended usage of the oral care device 100. In these examples the operational head 104 takes the form of a brush head for brushing teeth, gums, and so on. The operational head 104 is a separate entity to the handle section 102 and is suitably configured to be driven by the handle 102 once attached. The handle 102 comprises an engagement member 106 configured to couple the head operational head 104 to the handle 102.

[0042]

[0039] The handle 102 also comprises a drive train assembly 108, housed inside the handle 102, which is configured to drive the operational head 104. As will be appreciated by those in the art, the drive train assembly 108 may comprise a motor or electromagnet(s) that generates oscillating movement of a spring assembly, which is subsequently transmitted to the operational head 104 via the engagement member 106 which extends from the drive train and acts as a drive shaft. Drive train assembly 108 can include components such as a power supply, an oscillator, and one or more electromagnets, among other components, as desired to perform its function of driving the operational head 104. Examples of drive train assemblies suitable for use in an oral care device can be found in commercially available sonic electric toothbrushes and are described in W02015 / 101898, WO2018 / 046760, and WO2018 / 046429, all of which are incorporated by reference in their entirety. More specifically, the drive train assembly 108 is configured to cause the engagement member to vibrate to drive particular motion of the operational head 104. As used herein, vibration encompasses reciprocating, oscillating, and / or rotational motions.

[0043]

[0040] The handle 102 may also house a storage 110 configured to store drive profiles, or routines, for operation of the drive train 108. The drive profiles may set operating parameters for various modes of operation during a usage session, such as frequency and / or amplitude of vibration, duration of usage session, duration of operation at particular frequencies and / or amplitudes during a usage session, sequencing of particular frequencies and / or amplitudes during a usage session, and so on, and may be associated with a particular oral care requirement / procedure.

[0044]

[0041] Suitably, a user may set an operating setting of the drive train assembly 108 based on one of these routines. In one example the setting may be selected via operation of a switch on the handle section 102. In another example, the setting may be selected using an external electronic apparatus, such as a mobile phone, and communicated to the oral care device 100; suitably the oral care device 100 may comprise a communicator 112, which may be housed within the handle 102, for communicating with an external apparatus. The communicator 112 may be configured to communicate wirelessly using one or more wireless communication protocols such as Bluetooth, ZigBee, WiFi, near field communication, and so on.

[0045]

[0042] The apparatus 100 may also comprise a sensor unit 114, which may be configured to capture information related to usage of the apparatus 100 while it is in operation. For example, the sensor unit 114 may be configured to detect at least one of brushing signature, a shape of a user’s mouth, a shape of a user’s teeth, an angle of brushing (optionally determined per tooth), pressure used during operation (again, optionally determinable per tooth), overall duration of the operation (e.g., a total teeth brushing time), and a time of operation on an individual tooth (or segment of teeth).

[0046]

[0043] The sensor unit 114 may comprise optical based sensors, audio based sensors, and accelerometer based sensors. In an example, the sensor unit 114 may comprise an inertial measurement unit (IMU), with information such as tooth shape, brushing angle, pressure, etc, inferred (i.e., calculated) based on data provided by the IMU. In an example, the sensor unit 114 may comprise an image sensor configured to capture a plurality of images during operation of the apparatus 100 (including, optionally a video recording of a user’s full brushing procedure), with relevant characteristics (e.g., teeth shape, time of brushing, etc) determined based on the captured plurality of images.

[0047]

[0044] Data captured by the sensor unit 114 may be transmitted to an external device via the communicator 112.

[0048]

[0045] Currently, the oral-health care guidance provided for smart powered brushes and other oral care devices is based on general guidelines. These guidelines are too generic and do not address the needs, challenges, and concerns of various population groups. Furthermore, the standard advice has a low engagement potential as it might fail to address the goals and challenges one is facing at individual level. In contrast, a personalized advice can alter the self-care behaviour much more effectively owing to its high engagement prospects. The personalized approach however is not viable at scale without the use of automation.

[0049]

[0046] Suitably, the following embodiments describe a system that automates generation of a personalized oral-health care advice (for each entrant) through multi-modal data fusion, which is employed to group people into various demographic manifolds (or manifolds based on other characteristics). Each individual manifold is then subjected to collaborative filtering and learning schemes that extract insights and advice befitting the underlying group.

[0050]

[0047] In collaborative filtering, the goal is to provide personalized recommendations that leverage user-level information. A user-based collaborative filtering starts with a user, then finds users who have similar set of demography, behaviours, constraints, and strengths and finally makes a recommendation to the initial user based on what items the similar users benefit the most from, in the present context the items are the controls and behavioural adaptations of the users. An item-based collaborative filtering starts with a set of behaviours and controls, common among a group, and recommends similar items, based on their prevalence, that are beneficial for the group. Collaborative filtering, therefore, may use similarities between users and items simultaneously to provide recommendations. Furthermore, embeddings and population manifolds can be learned automatically, using machine learning, without relying on hand-engineering of features.

[0051]

[0048] The present embodiments provide an automated and engaging oral self-care guidance which is specialized for a certain population group. The engagement roots from the system’s ability to hierarchically group people with similar demographics as well as oral goals and then collaboratively generate advice through an automated virtual counselling (collaborative filtering). The tracking of goals at individual level creates an additional intra-group hierarchy, based on goal progression and the captured oral self- care behaviours. This hierarchy is used to generate evidence-based insights and to identify degenerative oral-care behaviours along with coping and adaptation mechanisms for the group.

[0052]

[0049] Figure 2 shows an example system 300 for generating oral care advice, which incorporates an oral care device 100 (e.g., toothbrush) as outlined above. The system also comprises an electronic apparatus 200, which may be embodied as a smart device such as a mobile phone, and an external server 250, which may be embodied as cloud computing architecture or other high performing computing infrastructure.

[0053]

[0050] The apparatus 200 comprises a display 210 configured to be controlled by a processor 206 to generate a visual output to a user of the apparatus 200. The displayed output might include, for example, a guided user interface (GUI). The apparatus also comprises a speaker 212 configured to be controlled by the processor 206 to generate an audio output, for example, a voice utterance.

[0054]

[0051] The apparatus 200 comprises a first communicator 202, and the server 250 comprises a second communicator 252, the first and second communicators 202, 252 being suitably arranged to communicate with each other to send and receive data transmissions to / from the other component. The first and second communicators 202, 252 may be configured to communicate wirelessly using long range communication protocols such as wired or wireless internet connection, cellular telecommunications (e.g., 3G, 4G, 5G), and so on.

[0055]

[0052] The server 250 comprises a storage 256 configured to store data received from the apparatus 200 as well as instructions to be performed by a processor 254.

[0056]

[0053] Figure 3 shows a computer implemented method 400 for generating oral care advice; for example, a method as may be performed by the server 250.

[0057]

[0054] At step 401 , the method comprises receiving, by the server 250, multimodal oral care data corresponding to a population of users of an oral care device. The population of users may be a population comprising exclusively users of the same oral care device, such as device 100 described above, or may be a population of users of a similar class of oral care device of which the device 100 is a member. For example, the class of devices may be “electric tooth brush”, and the population of users may all use an electric toothbrush as their preferred oral care device, although not necessarily the same electric toothbrush. In another example, the class of oral care device may be a device which connects to a particular application executable on the apparatus 200; for example, the Philips Sonicare app.

[0058]

[0055] For ease of explanation, in the following it is assumed that the population of users all the type of oral care device 100, and same type of electronic apparatus 200. In other words, along with a population of users, it can be assumed that there is a corresponding population of oral care devices 100, and a corresponding population of electronic apparatuses, or client devices, 200.

[0059]

[0056] Suitably, the multimodal data is collected by the sensors 114 provided on the population of oral care devices 100. Sensor data collected by a sensor unit 114 is transmitted to the apparatus 200, which in turn is transmitted to the server 250 via the apparatus 200. The sensor data may be sent to the server 250 substantially immediately upon receipt by the apparatus 200, or may be stored on the apparatus 200 for a predetermined (of flexible) amount of time before being transmitted. In some examples, the sensor data may be stored on the apparatus 200 until the apparatus 200 receives a request for the sensor data from the server.

[0060]

[0057] Suitably, the multimodal data may comprise usage information of the oral care device, as collected by the sensor(s) 114. By way of example, the usage information may be determined by reference to IMU data, which may indicate usage information such as shape of a tooth 302, brushing angle, pressure applied, usage time, and so on. In general, the usage information may include at least one of time of use (total or partial), pressure applied during use, and duration of use, operating signature, mouth shape, teeth shape, angle of use, pressure applied during use. Here it will be appreciated that ‘use’ is synonymous with operation of the oral care device 100.

[0061]

[0058] In another example, usage information may also be collated via user input into the apparatus 200, for example via a user interface displayed on display 210.

[0062]

[0059] In addition, the multimodal data may include one or more of dietary information for one or more of the population of users, demographic information for one or more of the population of users, oral health goal information for one or more of the population of users, and oral health tracking information for one or more of the population of users. It will be appreciated that, where a particular data type is included in the multimodal data, a data entry may be provided for each member of the population, however this is not a requirement, and entries for certain data types for certain users may be left empty (for example, if a member of the population has declined to share that particular information).

[0063]

[0060] Dietary information may comprise at least one of eating times, eating duration, and macro-nutrients in a user’s diet, and meal logs. Dietary data may be logged by the user in the population directly into the apparatus 200 (again, via a suitable GUI) using a suitable application that allows users to log diet and daily routines. The logs may be used to extract the dietary habits and the underlying macro-nutrients, which has direct implications for one’s oral health.

[0064]

[0061] Demographic data might be obtained based on manual input of the user into the apparatus 200 in form of surveys and questionnaires output to each user in the population via their respective display 210. That is, collation of demographic data may be made on request by the server 250 to client device 200 to be completed by each user.

[0065]

[0062] The oral health goal information of the population of users may include information such as “even teeth coverage”, “teeth whitening”, “fresh breath”, “managing gum sensitivity”, “reducing risk of dental caries”, and “minimising dental plaque”, and the like. Such information may collated using a suitable GUI on the aforementioned application.

[0066]

[0063] Oral health tracking information may track progress towards a listed goal. Similar to demographic data, health tracking information may be collated via survey and questionnaires. In an example, however, the health tracking information may be determined automatically based on data captured by the sensor 114 of the oral care device 100, or a sensor of an ancillary device. For example, oral health tracking information may comprise data captured by at least one of an imaging sensor 304 and a wearable oral sensor 306. Such data may provide data corresponding to gums health, teeth heath, and overall mouth state. In combination with the other data being collected, it may also be possible to determine periodic behavioural patterns in order to track habits. Mining of habits, at an individual level, is an important step for ensuring user engagement. The user can monitor not only their goal progression but also the evolution of relevant behaviours. This creates positive reinforcement and thus provides them with the required confirmation.

[0067]

[0064] The multi-model data of the population of users is stored in the storage 256, as appropriate for a cloud based architecture or other centralised high performance computing infrastructure. Centralised storage allows for ease of data mining and data fusion, as well as providing data retention and archiving capabilities to the system 300. Data retention allows creating and managing of the historical records for each member of the population which is relevant to the production of suitable advice, while archiving allows re-entrants to make use of their former records when desired.

[0068]

[0065] At step 402, the method includes generating, by the server 250, population clusters based on the multimodal oral care data. In particular, the population clusters may be determined using a machine learning (ML) model applying a clustering technique.

[0066] More specifically, using the fused multi-modal data, the method comprises generating population manifolds through a manifold learning algorithm, which performs a non-linear dimensionality reduction of the data. A machine learning algorithm is employed to this end that clusters the high-dimensional fused data into various groups. The grouping is performed by computing intrinsic embeddings, which are meaningful for the task of advice generation. In addition, a continuous feedback loop may be employed where, by tracking the efficacy of the generated advice, the manifold learning algorithm may be self-steered towards stabilization and better outcomes.

[0069]

[0067] In this context, although clustering may be performed based on usage data alone, the additional multimodal data mentioned above allows for improved clustering results. That is, ancillary information as captured from, e.g., the interactive health surveys and wearable health-trackers, provides additional information on the subjects, which results in additional dimensions in the population manifolds. These provide the basis to cluster subjects on similarity in much more detail, thus allowing a more targeted advice for such groups.

[0070]

[0068] At step 403, the method comprises receiving, for a particular user in the population, an oral health goal. The oral health goal may correspond to the same oral health goals included within the multimodal data. In some examples, however, the oral health goal may be a goal not included in the multimodal data.

[0071]

[0069] It may be envisaged that the particular user is a new entrant into the population, thereby setting up their data for the first time and inputting a health goal to obtain, or that the user is currently part of the population but is changing their oral health goal to focus on a new outcome. Thus the user is desirous of appropriate oral care advice to achieve this (new) goal.

[0072]

[0070] In other words, while the previous steps have related to the population as a whole, the following method steps now focus on one of those users (i.e. , the particular user and, accordingly, a particular oral care device 100 and particular apparatus 200, associated with that particular user).

[0073]

[0071] It should be appreciated, however, that the particular user may be implementing the method at the same time as other users, and therefore the method may be envisaged as being concurrent for all users in the population performing step 403 substantially at the same time. Moreover, the particular user may be taken to be an analogue for a group of users all sharing the same oral health goal.

[0072] At step 404, the method comprises generating oral care advice by performing collaborative filtering. The collaborative filtering is suitably based on the generated population clusters and the received oral health goal.

[0074]

[0073] Whereas oral-health surveys and smart oral-care sensors, for example, a guided imaging device to capture the gums, teeth, and mouth state, are integral part of the data collection.

[0075]

[0074] The multi-modal data establishes and tracks individuals’ goals, on one hand, and provides a basis for ranking of subjects, based on their similarity, on the other hand.

[0076]

[0075] Clustering of individuals provides fine possibilities for a hierarchical collaborative filtering, wherein a hierarchical weighting process is employed to assign distinct weights to user behaviours that originates from distinct groups, while the group-level performances can also be benchmarked and ranked likewise. This enables the generated advice to be tailored to the needs of a particular group, for example, by making sure the behavioural shift is not too large for a particular group, while still incentivising the global best behaviours.

[0077]

[0076] The collaborative filtering can also be visualized as a hierarchical improvement plan, where the individuals strive to be best in a group, by adopting best practices within the group, but the group itself is adapting and learning from other groups.

[0078]

[0077] The outcome of hierarchical collaborative filtering may be a consolidated ranking of oral health advice comprising various behaviours and controls which can be recommend to the particular user (and indeed, a group within the population to which the particular user belongs). That is, the collaborative filtering may comprise ranking a plurality of options for output as the oral care advice based on the oral health tracking information, and selecting the highest ranked option as the oral health advice. A superset of such controls and behaviours can be obtained through mining of association rules but can also be defined beforehand, though the latter is suboptimal.

[0079]

[0078] In this context, controls may correspond to operational characteristics of the oral care device 100. Put another way, the controls may relate to the usage information on the oral care device 100. That is, the generated oral care advice may comprise recommending a change in oral care device usage. Some examples of controls include, brushing style adaptation, use of a particular brushing head type, use of a certain toothpaste, and so on. Suitably, in one example, the generated oral care advice may comprise coaching information on a particular usage technique of the oral care device 100. In another example, the generated oral care device may comprise a recommended operating setting of the oral care device; for example, lower or higher frequency vibration.

[0080]

[0079] Behaviours may relate to characteristics of other data types captured in the multimodal data, such as dietary information. That is, the oral care advice may comprise recommending a change in behaviour. For example, a behaviour may be the addition or omission of certain nutrients or food items in the diet, eating at a specific time before or after brushing teeth, etc.

[0081]

[0080] At step 405, the electronic apparatus 200 is controlled to output the generated oral health advice (i.e., as determined by the collaborative filtering) via at least one of the display 210 or microphone 212; that is, as a visual and / or audio output. More specifically, the second communicator 252 may be configured to transmit the generated oral care advice to the first communicator 202 of the apparatus 200. In an example, the second communicator 252 may also transmit (concurrently or sequentially) a control signal for controlling the apparatus 200 to output the generated advice.

[0082]

[0081] Although the method above has been described with reference to a particular user, it will be appreciated that by virtue of the particular user being a ‘new’ entrant into the group associated with the oral health goal(s) selected by the particular user, that the oral advice for the group of users interested in that oral care goal may change. Suitably, it will be appreciated that the generated oral care advice may apply not just to the particular user, but to the group as a whole, and therefore the step of controlling to output the generated oral care advice may include transmitting the generated oral care advice to each apparatus 200 of users with the same oral health goal as the particular user, and controlling those devices to output the generated oral care advice too.

[0083]

[0082] Figure 4 shows further optional steps which may be performed in the present method.

[0084]

[0083] At step 406 where the oral care advice comprises a recommended operating setting of the oral care device 100, the method may further comprise, after outputting the oral care advice, controlling an operating state of the oral care device 100 according to the recommendation. Controlling the operating state may include, for example, switching the oral care device from an inactive to active mode (e.g., “off” to “on”), and controlling a vibration level output by the drive train assembly 108, although this list is not exhaustive. The operating state may be received from the apparatus 200 via the communicator 112. Controlling the operating state may be made responsive to a user input, for example, by touching an appropriate part of a GUI shown on the display 210.

[0085]

[0084] Further optionally, at step 407, the method may comprise determining that the oral care device 100 is in operation. In some examples, this step may be performed after performing step 406. In other examples, this step may be performed without performing step 406 (i.e. , direct after step 405).

[0086]

[0085] At step 408, the method may comprise determining that the oral care advice included coaching information, and if so controlling the display 210 (or speaker 212) to output the coaching information in real time while the device 100 is in operation.

[0087]

[0086] Further optionally, at step 409 the method may comprise transmitting the generated oral care advice to a medical practitioner (more specifically, a computing device associated with the medical practitioner), and receiving a review (which may include further recommendations) of the advice from the medical practitioner. The step 409 may be performed after any of steps 405, 406, or 408.

[0088]

[0087] As the above advice is generated purely through an automated virtual counselling of the population of users, there might be situations where the progress of the particular user (or group of users) is halted and needs assistance from the oral-care professional. Moreover, for safety and legal reasons the advice might need to be reviewed and endorsed by an oral-health care professional. Thus step 409 allows for expert oversight to be incorporated into the method / system when needed. The main task of this oversight is to ensure that the advice is not degenerative, that it is safe for the users and is adequately substantiated by the data and insights.

[0089]

[0088] 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.

[0090]

[0089] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, 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 measured cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

Claims

CLAIMS1 . A computer implemented method (400) to generate oral care advice, comprising: receiving (401 ) and storing multimodal oral care data corresponding to a population of users of an oral care device (100); generating (402) population clusters based on the multimodal oral care data using a machine learning model; receiving (403), for a particular user in the population, an oral health goal; generating (404) oral care advice by performing collaborative filtering based on the generated population clusters and the received oral health goal; and controlling (405) an electronic apparatus to output the generated oral care advice.

2. The method of claim 1 , wherein the multimodal oral care data comprises usage information of the oral care device and at least one of: dietary information of the user, demographic information of the user, oral health goal information, and oral health tracking information.

3. The method of claim 2, wherein usage information of the oral care device comprises sensor data received from an oral care device.

4. The method of claim 2 or 3, wherein the usage information comprises at least one of time of use, duration of use, operating signature, mouth shape, teeth shape, angle of use, pressure applied during use.

5. The method of any of claims 2 to 4, wherein dietary information comprises at least one of eating times, eating duration, macro-nutrients in a user’s diet, and meal logs.

6. The method of any of claims 2 to 5, wherein the oral health tracking information comprises data captured by at least one of an imaging sensor (306) and a wearable oral sensor (304).

7. The method of any of claims 2 to 6, wherein generating the oral care advice comprises ranking a plurality of options for output as the oral care advice based on the oral health tracking information, and selecting the highest ranked option as the oral health advice.

8. The method of any preceding claim, wherein the oral health goal comprises at least one of even teeth coverage, teeth whitening, fresh breath, managing gum sensitivity, reducing risk of dental caries, and minimising dental plaque.

9. The method of any preceding claim, wherein the generated oral care advice comprises at least one of a change in oral care behaviour and a change in oral care device usage.

10. The method of any preceding claim, wherein the generated oral care advice comprises a recommended operating setting of the oral care device.11 . The method of claim 10, further comprising controlling an operating state of the oral care device based on the recommended operating setting.

12. The method of any preceding claim, wherein the generated oral care advice comprises coaching information on a usage technique of the oral care device.

13. The method of claim 12, further comprising determining that the oral care device is in operation, and wherein outputting the generated oral care advice comprising coaching information comprises outputting the coaching information in real time during operation of the oral care device.

14. A system (300) for generating oral care advice, comprising: an oral care device (100); an electronic apparatus (200) comprising at least one of a display (210) and a speaker (212), and a first communicator (202) for communicating with a server (250); the server comprising: a second communicator (252) configured to receive multimodal oral care data corresponding to a population of users of an oral care device, and an oral health goal for a user from the electronic apparatus, a storage (256) for storing the multimodal oral care data, and a processor (254) configured to: generate population clusters based on the multimodal oral care data using a machine learning model; generate oral care advice by performing collaborative filtering based on the generated population clusters and the received oral health goal; and control the second communicator to transmit the generated oral care advice to the electronic apparatus, and a signal for controlling the electronic apparatus to output the generated oral care advice on the at least one of the display and the speaker.

15. The system of claim 14, wherein the electronic apparatus is communicatively coupled to the oral care device, the generated oral care advice comprises a recommended operating setting of the oral care device, and the electronic apparatus is configured to transmit, to the oral care device, a signal for controlling an operating state of the oral care device according to the recommended operating setting.

Images