AI-based camera tracking system for monitoring and enhancing toothbrushing effectiveness without sensors

Abstract

The present invention provides a system for monitoring and improving toothbrushing effectiveness. The system comprises a camera device 10 comprising a camera or similar visual sensor configured to cap

Description

The claimed invention is in the field of oral healthcare, in particular, the invention provides a system to assist users in improving their oral health by means of improving and enhancing their toothbrushing. Effective toothbrushing is critical to maintaining oral hygiene and preventing dental diseases like caries, gum disease, and plaque build-up. In currently used systems, the user would utilise a toothbrush that is equipped with one or more sensors configured to provide feedback regarding the user's toothbrushing. These toothbrushes are referred to as “smart” toothbrushes and comprise one or more accelerometers, pressure sensors, and magnetometers embedded within the toothbrush. When the toothbrush is in use these sensors can collect data detailing the motion of the toothbrush and the force applied to the brush. This information could then be used to determine the effectiveness of the user’s toothbrushing, which may include determining the user’s coverage, meaning that all of the teeth received adequate brushing, determining that there has been sufficient brushing duration, and / or determining that the user has supplied sufficient pressure to brush and has not applied excessive pressure which may damage the user’s teeth. One problem with such systems is that the embedded sensors it that such sensors may lose accuracy over time due in part to wear and tear as a result of the toothbrush’s repeated use. When such errors occur, the user would either need a means to recalibrate or would need to replace the brush entirely. The problem is that most users would not have access to the equipment required to perform the recalibration, and so would have no choice but to replace the toothbrush. It is noted that due to the embedded sensors, the smart toothbrushes can be relatively expensive therefore it may not be practical or affordable for all users to have to replace the smart toothbrush regularly. Another, potential problem with the use of smart toothbrushes is that they are all electric toothbrushes with vibrating heads, such head would also need to be regularly replaced. And in many cases, users may find the vibrations uncomfortable especially if they have sensitive teeth or gums, and therefore prefer the use of a manual toothbrush. Therefore, there is a need for an alternative to smart toothbrushes that still provides improved toothbrushing, but does not have the same high cost nor the hardware dependence as the smart toothbrush. Summary The present invention is in the field of oral healthcare. Specifically, the present invention provides a system to improve a user’s toothbrushing technique by utilising a system that provides video and images of the user brushing their teeth to a neural network which is configured to analyse the provided data and images to determine the effectiveness of the user’s toothbrushing and provide feedback on how their toothbrushing technique may be improved. In particular, the claimed invention provides a system that utilises a camera, or other suitable optical sensor that is configured to capture images and / or video of a user as they brush their teeth. These images provide data that can be fed into a feedback module to provide realtime feedback on the user’s toothbrushing technique. The data in the images may also be stored in a memory accessible to the feedback module to provide long-term feedback on the user’s toothbrushing technique using the stored information to determine patterns in the user’s tooth-brushing habits. The camera in question may be in the form of a digital camera or optical sensor which the user can mount in a position near the site where they brush their teeth such that the camera can capture video and / or images of the user brushing their teeth in real-time. This camera or sensor would also comprise a transmitter and receiver. Wherein the transmitter is configured to communicate the captured images and videos to a remote module that is configured to analyse the images and video to provide feedback on the user’s tooth brushing. The receiver is then configured to receive the feedback provided by the remote module. The camera or sensor would further comprise an interface configured to provide a display. Wherein the display would be configured to present the feedback received from the module. It is noted that the feedback may be presented as a visual prompt on the display or as an audio prompt. In some cases, the feedback may require the user to provide further input such as answering questions using buttons on the camera or a touchscreen display. For example, these questions may provide additional information that can help target or refine the feedback provided to that user. It is noted that users will likely brush their teeth in locations with running water, therefore it is preferable for the components of the camera or sensor to be housed within a waterproof housing to protect it from potential water damage. Wherein the housing would comprise suitable apertures to mount the camera and display such that the camera can view outside the housing and that the user may access the display to receive feedback. In some cases, the camera would be in the form of an application installed into the user’s mobile device, wherein the mobile device comprises one of a smartphone, laptop or tablet. It is noted that such mobile devices comprise each of the components described above. In these cases, the camera built into the mobile device will act as the camera sensor, the devices can then use communication channels available to the device, such as wi-fi or telecommunications to transmit and receive data to and from the remote module, with the device display and speakers being used to provide the feedback to the user. The remote module in communication with the camera would comprise a memory for storing the received images and videos, it would further comprise a suitable processor configured to perform analysis on the received data. In particular, the module would comprise an Al system, such as a neural network, contained within an Al software module, referred to herein as an Al module, configured to analyse the images and videos to monitor key variables in the user’s toothbrushing technique by tracking the motion of the user’s hands and the motion of the toothbrush relative to the user’s mouth, this may be achieved by tracking the user’s face in addition to their hand and the head of the toothbrush all of which should be present in the image. The parameters derived from the images and video may include the length of time the user brushes their teeth, the angle of their brush relative to their teeth, and the speed and motion of the toothbrush, from the above the system may track the coverage of the user’s tooth brushing technique, in this case referring to how much the of the surface of the user’s mouth had been brush and for how long. This information may be compared to a set of preferred values to determine the effectiveness of the user’s brushing technique and provide feedback on how the technique may be improved. The advice may include altering the relative angle or speed of the brush when cleaning a certain portion of their mouth or cleaning certain areas of their mouth for a longer duration. This feedback may be provided in real-time allowing the user to make these corrections immediately while brushing their teeth. It is noted that the provided advice may be stored in the camera device for the user to review at a later time or for the user to use as a prompt or reminder when they next brush their teeth. To achieve this effect the module processor would comprise suitable software for training and programming an Al module with training images and videos taken from a central database, and may also include the images and videos provided by other users. A program configured to store and apply algorithms to determine the relationships between the captured motions in the videos and images, with the toothbrush speed and moth coverage. These programs would be initially stored within the module and be trained using the stored data as described above. The module further comprises suitable tracking programs to identify and track key aspects in the captured images and video, such as the user’s face, hand and toothbrush. These tracking programs may include MediaPipe Face Detector for tracking the user’s face. It is noted that face tracking may be used to track the location of the user’s mouth and the relative size of their mouth. This information can be used to correlate the toothbrush movements to determine mouth coverage, it may also be used as a means to identify the user allowing the system to isolate the stored information relevant to that user, such as previously received videos, previously provided advice and the user’s mouth size and other details such as relevant oral conditions that may have been provided by the user. It is noted that the additional details may include oral health conditions the user suffers, location of crowns or filling in their teeth and any other conditions that may affect their tooth brushing technique. The Al module may utilise a model such as the YOLOv5, to perform real-time analysis on the received data from the user. This analysis may include detecting the motion of the user’s hand and / or toothbrush in the captured images. Using the information to determine the speed and position of the toothbrush and the angle of the toothbrush head, the system would use this collected data to determine the toothbrushing coverage, referring to the location of the user’s mouth that has been brushed and may also include a duration for which each section of the user’s mouth has been brushed for. Using this information the Al may compare the coverage to a set of ideal values that would be determined during the training of the IA. With this comparison, the Al may provide a list of commands that will improve the user’s brushing techniques such as changing the angle of the brush, changing their brushing speed or brushing a certain section of their mouth for longer. It is noted that the Al in the module would preferably be trained using a wide variety of images and videos, including those provided by the system’s various users. The training images would also include a wide range of video and images pre-provided to the system, to improve the training effectiveness these training images would include a range of different lighting conditions and images captured from different angles relative to the user, so long as the user’s face, hand and the hand and at least part of the toothbrush is visible in the captured image. As noted, the training data may also include a set of suggested ideal values for toothbrushing coverage such as an ideal speed, and time for brushing different sections of a user’s mouth. The training data may also include calculations that correlate the user’s hand or arm movement to the toothbrush movements, thereby allowing the system to use this data to determine mouth coverage. It is noted that the tracking model may include a plurality of additional programs to allow for further fine-tuned motion tracking to allow smaller movements to be detected. These programs may include OpenPose, Optical flow algorithms or similar programs. Using these programs the Al can track smaller movements in the received images such as changes in the angle of the toothbrush and or user's hand, or the motion of individual fingers that may change the orientation of the toothbrush or the pressure being applied to the toothbrush. The Al within the module may also utilise different systems to handle short-term and longterm analysis. For example, the feedback module may include Convolutional Neural Networks (CNNs) with Recurrent Neural Networks (RNNs), wherein the CNN is utilised to provide real-time feedback to the user while they are brushing their teeth based on the data derived from the real-time images and video received during their current toothbrushing system. In contrast, the RNN would be configured to track long-term patterns of the user’s toothbrushing technique over multiple sessions to find patterns in the user’s toothbrushing this may allow the module to provide immediate advice to the user to correct problems in their technique or provide other advice such as potential oral health problems or suitable products that may help the user in the long term. In some cases, the Al may also be configured to analyse still images provided by the user to derive additional information that would assist in the module’s analysis. For example, the user may provide images showing the head of the toothbrush. Such images may be used to assist the Al in tracking the position and angle of the toothbrush head when visible in received video data, this information can then be used to determine the location of the brush head throughout the video by tracking hand and brush motions. The images may also include one or more views of the toothbrush bristles before or after the user brushes their teeth so that the Al may track the degradation of the toothbrush over time, such that the Al may provide an alert when the user needs to change their toothbrush or to track if the user is applying too much or too little pressure when brushing their teeth. These images may also include images of the user’s mouth, these images may be used initially to determine the dimensions of the user’s mouth and teeth to assist in calibrating the Al calculations. The images may also be analysed to determine the user’s current oral health, as the Al may be trained to recognise objects in the user’s mouth such as crowns, fillings and braces that may affect their toothbrushing techniques and requirements, such as the need to apply less or more pressure to certain sections of their mouth. These mouth images may also be taken periodically, to indicate changes in the user's mouth such as their mouth growth, tooth loss or the addition of the objects set out above, such that the system may recalibrate accordingly. The Al may also be configured to diagnose certain oral health problems based on the images provided, such as indicating plague build-up or signs of gum disease and other oral health problems like gingivitis. To achieve this the training data may include a variety of reference images displaying such conditions from different angles and at different severities. It is noted that some images may use a plaque die to highlight the location and size of the plaque to help train the Al initially, it is noted that users may use such dies or toothpaste containing such dies to help the Al in recognising plaque more easily too. In response to such diagnostics, the Al may be programmed to adjust the suggested toothbrushing techniques to overcome the problems detected, provide product suggestions, such as tooth floss, kinds of toothpaste, mouthwashes or other products to help with the diagnosed condition, or may provide an alert indicating to the user that they should visit a dentist or oral health professional to receive treatment or additional advice. In these cases, the training data used when training the Al module would include a plurality of reference images to indicate different oral health conditions and different levels of toothbrush degradation, such images would include different lighting conditions and different angles relative to the camera. In operation, after the Al module has been trained the module may be configured to use the received data from a user to build a user profile. In some cases, the user may log in to the system through a user-specific account to ensure the module can assign the received data to a specific user. In other cases, the module may be configured to use recognition programs, such as facial recognition to identify users and assign the received data to that user’s profile. In other cases, the module may use metadata in the received data, such as the IP address that sent the received data to identify the user. It is noted that there are some cases where multiple users may use the same camera device to send data to the Al module, for example, a family who lives together may share a single account, therefore it may be preferable for the Al module to use both digital data, such as the account login or IP address, in combination with facial recognition when building the user profile as this will allow the system to differentiate between different users even when using the same camera device. In operation, the Al module may use the profiles to store client-specific details that would help to calibrate the system’s calculation along with client details required to provide longterm monitoring and advice to the client. For example, the module may be configured to store information such as the dimensions of the user’s mouth and arm to determine the arm motion the Al should see in the video to provide full brushing coverage. The dimensions of the user's hand may also be used to determine smaller arm motions such as finger and wrist movements that would affect the brush orientation, which would also be used when determining mouth coverage. The profile may also contain information regarding the user’s oral health, which may be inputted into the camera device display when setting up or updating the user’s profile or may be derived from images of the user’s teeth. Such information may include whether the user has braces, crowns, fillings or any oral health problems, again this information can be used to adjust the calculations and advice provided by the Al module, and it may also be used to track the oral health of the user over time to determine long term solutions and advice that can be provided to the user. When the Al module has recalled the stored information for an identified user it may analyse the video and images provided by the user in this session. The images may be analysed to determine if there have been any changes in the user information stored in the profile, such as the user dimension changing, which may be the case for a child user, changes in toothbrush degradation, or changes in the user's oral health such as new problems being identified. When these changes are identified the Al module can update the user profile and adjust calculations and algorithms used to determine the effectiveness of the user’s toothbrushing accordingly. Once the profile is updated the Al module can analyse the video of the user brushing their teeth and determine the parameters of the user’s toothbrushing technique such as brushing speed, brushing direction and duration and mouth coverage, from which the Al may provide real-time feedback and advice to improve their brushing technique. It is noted that the Al module may store session-based information for each user, in this case, a session would refer to each time the user transmits the video of them brushing their teeth. This session-based information would include the above-mentioned parameters and analysis performed by the IA, any advice provided during that session and any parameter changes made in that session. By storing session details in the user profile, the Al module may identify changes over time more easily by comparing data across sessions for a specific user. This may include identifying recurring issues in the user brushing technique, worsening oral health problems and toothbrush degradation. From this information, the Al module may provide long-term advice for the user such as creating a toothbrushing schedule, providing brushing technique advice such as increasing brushing speed or brushing certain sections of their mouth more, at the start of a session based on recurring issues. It is also noted that using this session data the Al module may monitor the age of the data stored, this may help indicate when data in the system needs to be updated. More specifically, a user may not provide images of their mouth and toothbrush every session, therefore to ensure the data is up to date the system may determine when a predetermined number of sessions have passed without the mouth or toothbrush images being updated, at which point the Al may send an alert to the camera device instructing the user to take an updated picture of their mouth and / or toothbrush to update the user profile. It is also noted that if a user profile indicates that the user is suffering from an oral health condition, or worsening oral health, the system may reduce the number of sessions before the profile images should be updated to allow the system to monitor these problems more closely. The above profiles may be used to allow users to easily share their stored information with others. This may include sharing their profile records with other users to create a comparison across different profiles. For example, a child user’s profile may be shared with their parents’ profile so they can monitor their child's oral health. The profile may be shared anomalously to allow people to compare their toothbrushing effectiveness with other users either locally or globally, this may also include a feature to provide a comparison of the other user’s brushing technique to their own. In such cases, the Al module may provide a comparison of brushing speed, mouth coverage, toothbrush age or similar variables that can affect the toothbrushing effectiveness but would not share any personal data, nor the videos and images the other user had provided. Lastly, when an oral health problem has been identified, the system may be configured to alert the user to see a dentist or other oral health professional, as part of this alert the user may instruct the system to share their findings with a chosen oral health care professional and may also provide the user’s entire profile to the professional to allow them to more easily diagnose the problem or provide advice on how to address the issue. As noted above the claimed system is configured to provide feedback based on the analysis of the user’s video and images. To achieve this the Al module would be in communication with a user-specific feedback device that would configure a display that is suitable for providing audio and / or visual feedback. It is noted that the feedback device may be the same as the camera device or may be a separate device. In either case, the feedback device would be configured to provide feedback from the Al module, such feedback would include real-time instructions on how the user can improve their toothbrushing technique as they are brushing their teeth, toothbrushing advice and instructions based on previous toothbrushing sessions and common issues detected over time, and alerts with suggestions such as informing the user when they should change their toothbrush or when they should see a oral health professional for a potential oral health issue. It is noted that this feedback may be provided through visual and / or audio means. In the case of visual feedback instructions can be displayed on a display of feedback device such that the user may read the instructions, it may also include other visual cues such as highlighting sections of a mouth in different colours to indicate which sections have been brushed sufficiently and which areas require more attention, may also include images to indicate to the user that they need to change their toothbrushing speed and / or the orientation / angle of the toothbrush to provide better coverage. In the case of audio feedback, the instructions may be spoken through a speaker in the feedback device, the audio feedback may be in the form of spoken words and other haptic feedback such as buzzers or other noise. It is noted that haptic feedback may be used to indicate to the user when feedback has been received and / or when an instruction in the feedback has been achieved. The use of this feedback can allow the user to improve their toothbrushing technique without disrupting their tooth-brushing session. The display device may also store the user profile including their parameters, sessions, previous advice and any alerts or suggestions beyond their toothbrushing session such as suggestions to replace toothbrush, product suggestions, problem diagnosis or suggestions to see an oral health professional for the user to review between tooth brushing session to act on the suggestion or to track their changes and progress between sessions. In some cases, the feedback device may provide different types of feedback to child users. More specifically, the feedback device may provide “gamified” feedback for child users, wherein the feedback is configured to make the instructions more fun and interactive. This may include providing music to instruct the user when to brush their teeth faster or slower based on the tempo of the music, as well as indicating the duration they should be brushing their teeth for. The gamified feedback may also include setting objectives or goals and providing a reward when such goals have been met. These rewards can then be tracked in the user's profile. The feedback may also include lessons and quizzes to educate child users and promote good toothbrushing techniques. The feedback device may also comprise input means such as buttons or a touch screen to allow the user to provide inputs to the device. These inputs may include typing information to set up a user profile and adding additional information and parameters to their user profile. To initiate and end a current toothbrushing session, and as part of the session upload images from the camera device to the Al module and initiate a video recording from the camera device. The user may also use these inputs to review feedback, such as redisplaying feedback from a previous session, or opening their profile to see the long-term feedback and tracking and to review any alerts or suggestions provided by the Al module. In the gamified feedback, these input means would be used to provide responses to quizzes and games presented to the user and control the lessons being presented. The user may also use these inputs to customise the feedback they receive such as limiting feedback to audio or visual only, or changing the brightness or volume of the display to make the feedback easier to perceive. By using a system as described above the user is provided a means for improving their toothbrushing technique in real-time without the need for specialised hardware, which may be expensive to acquire and replace and may also be cumbersome to use. As such the claimed system is more accessible as it can use mobile devices the user already has access to alongside a regular toothbrush. The system also provides an easy means for a user to track their oral health and can also provide some diagnostics of potential oral health issues. And in response to the feedback provide the user with products and instructions that can help overcome the diagnosed issues and generally improve oral health practices. Wherein this advice and suggestions are targeted to a specific user’s need based on diagnostic image analysis, and short-term and long-term tracking of the user’s toothbrushing techniques based on the analysis of key variables within the images and videos provided by that user. In summary, the invention provides in the first aspect a system for monitoring and improving toothbrushing effectiveness, comprising a camera device configured to capture real-time footage of a user brushing their teeth; an Al-powered software module designed to analyse the footage using object recognition, pose estimation, and motion tracking algorithms to detect the user’s hand, toothbrush, and face; a feedback interface that provides real-time visual, auditory, or haptic feedback based on brushing technique, including parameters such as angle, coverage, and movement patterns. Wherein the system operates independently of sensors, accelerometers, or physical markers embedded in or attached to the toothbrush, like those used in smart toothbrushes. The invention further provides the above system wherein the Al-powered software module uses pose estimation algorithms, including but not limited to OpenPose or MediaPipe, to detect and track the user’s hand movements and toothbrush trajectory, ensuring optimal brushing coverage. A system wherein the Al-powered software module employs optical flow tracking algorithms to analyse brushing speed, angle, and coverage over time. A system wherein the feedback interface provides real-time visual overlays that highlight areas of the mouth that have not been adequately brushed, issuing alerts when improper brushing angles or inadequate coverage are detected. A system wherein brushing pressure is inferred through camera data by analysing motion characteristics such as hand velocity, acceleration, and abrupt changes in toothbrush trajectory, thereby eliminating the need for physical pressure sensors. A system wherein a data analytics module aggregates session data, generating detailed brushing reports and personalised recommendations based on historical performance, utilising adaptive learning algorithms that refine feedback over time. A system configured to be adaptable to work with any type of toothbrush, whether manual or electric, and any consumer-grade camera-equipped device, including but not limited to smartphones, tablets, and webcams, regardless of lighting conditions and user environments. A system wherein the system includes adaptive learning algorithms that personalise brushing feedback based on the user's prior sessions, further enhancing long-term brushing effectiveness. A system configured to be capable of operating in multi-user environments by identifying individual users based on unique brushing patterns and facial recognition, providing personalised feedback and data logging for each user. A system wherein the Al module integrates Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to track and improve long-term brushing habits by analysing temporal brushing patterns, and providing feedback on repetitive brushing issues such as consistently poor angles or excessive pressure over time. A system wherein the Al module adapts feedback based on the user’s specific dental condition, including the presence of braces, crowns, or dental implants, providing tailored brushing guidance for these conditions. A system wherein the feedback interface includes gamification features for children, offering rewards, points, and progress tracking to encourage proper brushing habits. A system wherein the feedback system offers customisable visual, auditory, and haptic feedback to cater to users with hearing or vision impairments, ensuring accessibility for all users. A system wherein a cloud-based analytics platform allows users to compare their brushing habits against regional or national averages, positioning the system as a public health tool. Detailed Description The claimed invention is depicted in the following figures: Figure 1 - provides a simplified schematic drawing of the claimed invention Figure 2 - provide a flow diagram showing the general process of the claimed system Figure 3 - provides a flow diagram showing further steps performed during the video feedback steps depicted in Figure 2. The figures comprise the following features: 10 - camera device 20 - feedback device 30 - Al module 40 - Data Storage module 50 - Identification step 51 - Session start 52 - Image feedback step 53 - Video feedback step 54 - Session end 55 - Profile update 56 - Long-term feedback step 60 - Video start step 61 - Video processing 62 - Toothbrush detection step 63 - Mouth segmentation step 64 - Brush tracking step 65 - Coverage tracking step 66 - feedback step The present invention provides a system that utilises Al technology and machine learning to improve the oral health habits of the users. In particular, the system uses image and video analysis, with applied algorithms, to determine real-time feedback to improve the user’s toothbrushing technique. The system also stores information in a user profile to track the user’s feedback and oral health over time to provide more targeted feedback to improve their oral health practices. Figure 1 depicts a schematic showing a simplified version of the claimed system. The system comprises a plurality of camera devices 10 and feedback devices 20 which will be used by the various users. The camera devices 10 are configured to capture video and images of the user which can be supplied to the downstream system for analysis. In particular, the camera device would be used to capture images that can be used to determine key parameters for the user’s profile such as the dimensions of their mouth and hand which will be needed for calculations to determine brush movements and mouth coverage during toothbrushing. The images of the user's mouth may also be used to determine their current oral health and any potential problems. This may include identifying braces, crowns and filing in the user's mouth and issues such as cavities or plaque within the user's teeth, as the system may need to adjust its feedback to address these issues. The user may also provide images of their toothbrush head to allow the toothbrush to be identified in further images and videos and to track toothbrush degradation. The camera device 10 is then further configured to provide a video of the user brushing their teeth in real time to the downstream system. The downstream system will be configured to analyse the video to identify key features in the video feed such as the location of the user’s mouth, hand and toothbrush. The video analysis would further track the motion of the identified objects to derive variables based on the tracking of the user’s movement in the video. These variables will be used to determine the user's toothbrushing technique such as brushing duration and mouth coverage. The feedback device 20 would be configured to communicate with the downstream system to receive the results of the analysis of the video and images provided by the camera device 10. More specifically, the device 20 would be configured to present the feedback and results from the downstream system to the user. As such the feedback device 20 comprises a display that is configured to provide feedback to the user, this display would include elements to provide visual, audio and haptic feedback to the user. The device may also comprise control elements such as a touch screen or buttons to allow the user to provide additional information to the downstream system, respond to the received feedback, display the user’s profile and recall previous feedback from this session or a previous session for review. The user may also use the input means to adjust the settings of the feedback device display for example to prioritize a certain type of feedback, to activate audio descriptions or similar options to improve the user’s ability to perceive the feedback provided. It is noted that the above-mentioned camera device 10 and feedback device 20 may be a single device. In some cases, both devices 10 and 20 may have all their components housed in a single housing to form a single device, it is noted that this housing is preferably waterproof as the user would likely be cleaning their teeth in rooms with running water. However, it is noted that the users' mobile devices can also provide the combined camera device 10 and feedback device 20. In particular, instead of using separate devices, the user may install an application onto a mobile device such as a phone, tablet or laptop. When running the application, the built-in cameras and display of the mobile device may be used as the camera device 10 and display device 20 as described above. Regardless of the specific device arrangement used each of the camera devices 10, feedback devices 20 and combined devices will be in communication with a remote Al module 30, such that data can be exchanged between the Al module 30 and the devices using suitable wireless communications, such as wi-fi. In particular, when in use the camera device 10 would need to transmit videos and images to the Al module 30 for analysis and the display device 20 would be configured to receive analysis results and feedback from the Al module 30 to be displayed and to transmit inputted information to the Al module 30 when required. The Al software module 30 containing a suitable neural network for Al learning, herein referred to as the Al module 30, would be primarily configured to perform an analysis of the toothbrushing footage provided by each user. To achieve this the module 30 would comprise programs to allow object recognition and tracking. In particular, the module would be configured with facial recognition to identify the user in the video and locate the user’s mouth in the captured images. The module would further identify the users’ hands and the toothbrush they are using. The Al module 30 would then comprise programs for tracking the movements of the user’s toothbrush, and the hand holding it relative to the user’s mouth. From this tracking, the Al module 30 can derive key parameters relating to the user’s toothbrushing technique, such as the speed of the toothbrushing, the angle of the toothbrush compared to the user’s teeth, and the duration of brushing for each toothbrush orientation and direction. From this information, the Al module 30 would be configured to user algorithms stored in the module to determine scores for the users’ mouth coverages, both as a total coverage score and separate scores for individual sections of the user’s mouth and toothbrushing effectiveness. These scores can then be presented to the user as feedback on the feedback device 20 as written words, audio outputs or images when appropriate. The Al module 30 would be further configured to analyse the above results to determine suggestions to improve the user’s scores this may include having the user change their brushing speed, brush certain sections of their mouth more, or use different brushing angles. It is noted that the Al module may contain a set of ideal values to which the scores are compared, in some cases the set of ideal scores may be adjusted based on the best performances stored by other users in the system. These improvements would be sent to the feedback device to be displayed to the user in real-time to allow the user to improve their brushing techniques as they are brushing their teeth. As with the previous feedback, these suggestions may be presented as audio or visual feedback on the display device 20. It is noted that to provide this feedback accurately, the Al module would be configured with Al learning software, such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). These neural networks would be configured with appropriate algorithms to derive the performance parameters from the parameters determined from the received images and videos, the network may also be provided target values / scores for these parameters based on common dental practice and oral health recommendations, though it is noted that the scores may be adjusted if users are found to achieve better results in practice compared to these hypothetical values. It is noted that in addition to the algorithm and expected values uploaded into the neural network, the Al module can be trained with a plurality of reference images and videos uploaded onto the system, this training would then be further refined with additional images and videos supplied by the users’ camera devices 10. The reference video and images would include images with a range of different lighting conditions and a range of different angles relative to the camera device 10, so as to train the Al module 30 to detect key features, namely the people in the images, their faces, mouth, hands and toothbrush and track the motions of these features in the video, thereby training the module 30 to apply the installed programs and algorithms over the wider range of images and video conditions. The training images may also include a range of different toothbrushes and kinds of toothpaste on said brushes to allow the Al module to identify them and track them more easily. The training images may also include examples of toothbrush heads at different stages of degradation to train the Al module when a user’s toothbrush should be replaced. The images may also include a range of reference images for different oral health issues, such as braces, crowns, filings, gingivitis, cavities and other problems that may affect the ideal values of the user’s toothbrushing technique and may alter the suggestions provided to the user. In some cases, the training images may include different levels of plaque buildup, these images may include the mouth with the addition of a plaque die to make the plaque more visible and programming into the module the varying values for the key parameters required to address different levels of plaque and to address plaque in different locations of the mouth. It is also noted that in these cases, the Al module 30 may be trained to diagnose different oral health conditions based on the images provided by the user and provide suggestions such as oral health products and instructing the user to see an oral health professional. In some embodiments, the system would be configured to build user-specific profiles. These profiles would include key parameters for the user such as reference images of their mouth and body so that they can be identified in received videos more easily. The profile would also include derived parameters to be used in the algorithm to determine the ideal values for their toothbrushing scores as described above. This will allow the Al module 30 to derive a score for the user’s toothbrushing technique more easily as the system can use the stored values rather than recalculating key parameters for each video provided. It is noted that a user may manually log into their profile when initiating the system, or the Al module may use facial recognition to automatically correlate the user to their specific profile. The user profile may also store a record of the user’s performance between sessions to allow long-term monitoring of the user’s performance. This would allow users to monitor if their scores are improving over time or if there is any common advice provided in multiple sessions to give the user a target to improve. The Al module 30 may be configured to track the user profile over multiple sessions to determine if there are common issues, referring to advice that is given to the user repeatedly such that the Al module 30 may provide the repeated advice at the start of the users’ current session. Long-term monitoring may also be used to track the condition of a user’s toothbrush and oral health issues more accurately by providing a timeline of toothbrush degradation and replacements and also allowing the severity of any diagnosed mental health problems to be tracked over time. The user’s profile would also allow information regarding the user’s toothbrushing performance and general oral health to be shared more easily. More specifically, the user can share the content of their profile with dentists or other oral health professionals to provide targeted advice or to confirm the diagnosis performed by the Al module. The users may also share their scores to allow users to compare their performance with other users, comparing each other’s techniques such as brushing speed, directions and durations to allow users to improve their scores and techniques. The Al module may also be configured to have separate protocols when configuring the profile for child users. The system may be configured to recognise child users and note this in their profile as the requirements for a child will differ from an adult. As such, the system can use a different set of ideal parameter values for a child profile. Additionally, the child will be more likely to grow meaning the Al module may be configured to recalibrate the user dimensions more often for a child profile, this would include mouth dimensions and user arm length which would affect the motions the Al module 30 would be programmed to track. The child profiles may also be linked to an adult profile, that belongs to the child user's parent or guardian to allow the parent or guardian to monitor the child’s oral health more easily. The feedback provided to a child's account may also be altered. Not only will the feedback be adjusted to a different set of ideal values, but may also be gamified to make the system more engaging for a child user. For example, when advising the user to user to brush their teeth longer the feedback device may use music instead of or in addition to a timer. The feedback instructions may be presented as goals which provide the user rewards, such as stars or merits, which can be tracked on their profile and the users' scores may be presented as a game score, which may include a user’s high score or leaderboard to encourage them to achieve a higher score. It is noted that the latter feature is also present for adult users when performing profile reviews and comparisons, but for a child user, it may be presented as a game which may provide a reward for matching or beating the previous high scores, or for having a streak of achieving the ideal score. The child profile may also be provided with interactive educational materials such as lessons and quizzes to promote improved toothbrushing techniques. To store the vast amount of data disclosed above including the training data, user profiles and the images and videos provided, the system would require one or more data storage systems 40 configured to store this data in a suitable format such that the Al module may recall data when required, such as recalling parameter values, user profiles and algorithms to be applied to the currently received videos and images. It is noted that the storage system 40 may include databases that correlate user profiles with images and videos containing that user such that the Al module may isolate the data for a specific user when performing the long-term monitoring determination as described above. The storage system may also include suitable programs to compress the video and images received by the user to allow them to be stored without using up excessive portions of the storage memory. Preferably these programs would be configured to allow the compressed images and videos to be restored or at least partially restored to allow the Al module to analyse the images and videos for training the neural network and making determinations. Figures 2 and 3 provide flow diagrams summarising the processes described above. Figure 2 provides a flow diagram of the general process when using the claimed system. In step 50 the user or users about to perform the toothbrushing are identified. This step may involve the users signing into a personal account through inputs via the feedback device 20. Alternatively, the user may initiate a connection with the Al module which starts streaming a video from the camera device 10 to the Al module 30, wherein the Al module may use facial recognition to identify the user or users in the video stream, or may use metadata in the stream such as an IP address or device ID for the camera device to determine the user’s identity. Once the user is identified the Al module 30 may recall the user’s profile to add the new data to the profile and use the user profile to provide feedback and adjust parameter values. At step 51 the user starts a session this informs the Al module 30 to begin capturing the camera device 10 video for analysis, it may also provide the user with the option to take and upload specific images to the Al module 30, as shown in step 52, such as their current toothbrush to allow better tracking of toothbrush motion and degradation, the user may also upload images of their teeth and gums to diagnose any issues or track any pre-diagnosed issues. The Al module 30 may also be configured to determine how long it has been since any such images have been received, if the time is above a predetermined threshold the Al module 30 may instruct the user to upload such images before they begin brushing their teeth to improve the system’s tracking and monitoring. Regardless if any images are uploaded, the Al module 30 will proceed to step 53 and will begin recording the video transmitted by the camera device 10, wherein the Al module will analyse the video as described in Figure 3 to provide the user with feedback, both real-time feedback to improve their toothbrushing technique while brushing their teeth as well as generalised feedback based on trends in the user’s data between sessions such as suggesting oral health products or instructs to follow next time they brush their teeth. Once the user has finished brushing their teeth, they may move to step 54 they may end the session to stop the Al module from further capturing and analysing the video feed. The Ai module will then advance to step 55 wherein it will update the user profile adding the data from the current session including key parameters, technique scores, any alerts or diagnoses provided and a record of the advice given. This stored data can then be tracked with the stored data from previous sessions to determine the user’s state of oral health and to determine any additional advice and instructions that can be provided to the user, in step 56 to improve their oral health, such as additional products they could use, replacing their toothbrush, different techniques to use in the next session or setting an appointment with an oral health professional when necessary. Figure 3 provides a flow diagram that details the steps that occur during the video analysis in step 53 in Figure 2. At step 60, a video input is transmitted from the camera device 10 to the Al module 30 for analysis. In step 61, the Al module 30 performs its initial preprocessing this may include using object identification to identify the user in the video and locate the user’s mouth, hand and toothbrush in the video feed, such that the position and orientation of these objects can be tracked. The Al module may also be configured to take images during this process for analysing the user's toothbrush for degradation and to detect the toothpaste on the user’s toothbrush and determine if there is sufficient toothpaste on the brush, if not the Al module 30 may be configured to instruct the user to apply more toothpaste to the brush before they start brushing their teeth. In steps 62 and 63, the Al module 30 will prefer some analysis of the initial video images to determine further variables that need to be determined to allow for accurate tracking. These determinations include determining the dimensions of the toothbrush in the feed and estimates for the size of the user's mouth sections or segments, to provide dimensions for the user's teeth that need to be brushed. It is noted that the user may provide images of their toothbrush and the inside of their mouth to assist in these determinations. However, on receiving the video feed the Al module would need to determine the relative size of these dimensions in the angle viewed by the camera device 10 to adjust the parameters to the specific video feed being received that session. In step 64, the Al module will analyse the video feed in real-time to track the motion of the user’s hand and toothbrush. In particular, the Al module 30 will be tracking the portion of the toothbrush relative to the user's mouth, the toothbrush angle and orientation relative to the user's mouth, and the velocity of the tracked motions of the user’s hand, or the toothbrush if visible in the video images, to track the speed and direction of the user’s motions over time. From these values, the Al module can determine the brushing speed and mouth coverage of the user’s tooth brushing to determine its effectiveness, by comparing these values to a set or range of ideal values, it is noted that these ideal values may be adjusted to a user’s size or age. The Al module may also be configured to use the motions of the users hands and the toothbrush head to determine the amount of force and pressure being applied to the user’s teeth, which may be compared to a set or range of ideal values to ensure that the user is applying sufficient force to clean their teeth but will also ensure that the user is not applying excessive pressure that may damage their teeth. However, it is noted that the toothbrush itself may not always be visible in the video therefore it is important for the Al module to also track the movement of the user’s hand holding the toothbrush. This includes the overall motion of the hand to determine brushing speed, but also smaller movements like the fingers and wrist which may be used to change the angle and orientation of the brush. From this the Al module 30 may track motions in the video to determine the user's mouth coverage, brushing duration and overall brushing effectiveness. In step 65, the Al module may be configured to separate the values determined in step 64 into region-specific values. This is to say that the Al module may be configured to track the user’s toothbrush motions at different regions of the user's mouth. Thereby the Al module may determine the speed at which the users have brushed each section of their mouth and the duration each section has been brushed for. The Al module may also use the tracking programs to determine the toothbrush angles used when brushing each section of the user’s mouth to determine tooth coverage for each section. With these determinations, the Al module may provide more targeted advice and feedback for the user’s toothbrushing, as the results will be broken down for each section of a user’s mouth rather than an overall score. The feedback may therefore include more specific instructions such as instructing the user to clean a specific region of their mouth for longer or to use different brushing angles when cleaning a specific section of their mouth, thereby providing more targeted specific feedback to the user when they are brushing their teeth. At step 66 the system provides the feedback based on the determinations and tracking described above to the user. In particular, the Al module 30 will provide the feedback to the user’s feedback device 20 to be displayed to the user as audio, haptic or visual feedback. This feedback may include the user’s current scores for effectiveness and tooth coverage, average brushing speed and these same values for specific mouth regions. In addition to these results, the feedback will include real-time instructions that the user may apply to improve the scores while brushing their teeth such as changing brushing speed, changing the pressure applied to the toothbrush, changing the brushing angle or brushing a specific portion of their mouth for a longer duration. This way the system can improve the effectiveness of the user’s toothbrushing technique in real time. By using the system as described above the user is provided with a means of improving their toothbrushing and general oral health that is not hardware dependent, and may use devices they already possess providing a cheaper alternative to a smart toothbrush that will not require additional maintenance. This will allow a wider range of people to access a system to effectively review and improve their toothbrushing to improve their oral health.

Claims

1. A system for monitoring and improving toothbrushing effectiveness, comprising:A camera device comprising a camera or similar visual sensor configured to capture real-time footage of a user brushing their teeth;An Al-powered software module designed to analyse the footage using object recognition, pose estimation, and motion tracking algorithms to detect the user’s hand, toothbrush, and face;A feedback device that comprises a feedback interface that provides real-time visual, auditory, or haptic feedback based on brushing technique, including parameters such as angle, coverage, and movement speed and movement patternsWherein the feedback comprises instructions to the user to alter their toothbrushing technique to bring the parameters closer to a predetermined set of ideal values.

2. The system of claim 1, wherein the Al-powered software module uses pose estimation algorithms to detect and track the user’s hand movements and toothbrush trajectory, ensuring optimal brushing coverage.

3. The system of claims 1 and 2, wherein the Al-powered software module employs optical flow tracking algorithms to analyse brushing speed, angle, and coverage over time.

4. The system of any preceding claim, wherein the feedback interface provides real-time visual overlays that highlight areas of the mouth that have not been adequately brushed, issuing alerts when improper brushing angles or inadequate coverage are detected.

5. The system of any preceding claim, wherein brushing pressure is inferred through camera data by analysing motion characteristics such as hand velocity, acceleration, and abrupt changes in toothbrush trajectory, thereby eliminating the need for physical pressure sensors.

6. The system of any preceding claim, wherein the Al software module is in communication with a data storage system configured to aggregate session data, generating detailed brushing reports and personalised recommendations based on historical performance, utilising adaptive learning algorithms in the Al software module to refine feedback over time.

7. The system of claim of any preceding claim, wherein the system is adaptable to work with any type of toothbrush, whether manual or electric, identifying the type of toothbrush from the camera device video feed and adjusting the ideal parameters and algorithm accordingly.

8. The system of any preceding claim wherein the camera device and feedback device comprise the same device.

9. The system of claim 8, wherein the combined camera and feedback device comprises any consumer-grade camera-equipped device, including smartphones, tablets, and webcams.

10. The system of any preceding claim, further configured to be capable of operating in multi-user environments by identifying individual users based on unique brushing patterns and facial recognition in a single video fed, providing personalised feedback and data logging for each user.

11. The system of any receding claim, wherein the Al software module integrates Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to track and improve long-term brushing habits by analysing temporal brushing patterns, providing feedback on repetitive brushing issues such as consistently poor angles or excessive pressure over time.

12. The system of any preceding claim, wherein the Al software module is configured to use images of the user’s mouth to identify specific dental conditions including the presence of braces, crowns, or dental implants; andIs configured to adapt feedback based on the user’s specific dental condition, providing tailored brushing guidance for the identified conditions.

13. The system of any preceding claim, wherein a user in the video feed can be identified as a child, adjusting the ideal values for a child user; andwherein the feedback interface includes gamification features for child users;wherein the gamification features include providing rewards or points for achieving ideal brushing parameters, and progress tracking to encourage proper brushing habits.

14. The system of any preceding claim, wherein the feedback system offers customisable visual, auditory, and haptic feedback to cater to users with hearing or vision impairments.

15. The system of claim 1, wherein the Al software module provides a platform allowing users to compare their brushing habits against regional or national averages, positioning the system as a public health tool.

16. The system of any preceding claim wherein the camera device is configured to provide still images of the user to calibrate the Al software module programs, including images of the user for calibrating facial recognising and hand detection.

17. The system of claim 16, wherein the provided images comprise periodic images of the user’s toothbrush to calibrate toothbrush detection and tracking, especially when the toothbrush has been replaced; andwherein the Al module is configured to use the periodic images to track toothbrush degradation and alert the user, via the feedback device when their toothbrush needs to be replaced.

18. The system of claims 16 and 17, wherein the provided images include images of the inside of the user’s mouth;Wherein the Al software module is configured to determine the dimensions of the user’s mouth and specific mouth segments from the images; andWherein the Al software module is configured to diagnose different oral health conditions from the mouth images, and track the severity of the diagnosed condition over time;wherein the Al software module may adjust the ideal parameter values to address the diagnosed issues and wherein the feedback provided by the Al module includes suggesting products or instructing users to have an appointment with a profession to address the issue.

19. The system of claim 18, wherein the oral health conditions identified include at least one of plaque buildup, cavities, gingivitis or gum disease.

20. The system of any preceding claim wherein the Al module is configured to detect the presence of toothpaste on the toothbrush from a supplied image or an image of the video feed and determine if there is a sufficient amount present, based on the size of the user’s toothbrush in the image, if not the Al module will provide feedback comprising an instruction to apply more toothpaste to the toothbrush.AMENDMENTS TO THE CLAIMS HAVE BEEN FILED AS FOLLOWS:-Claims1. A system for monitoring and improving toothbrushing effectiveness, comprising:A camera device comprising a camera or similar visual sensor configured to capture real-time footage of a user brushing their teeth with a toothbrush;An Al-powered software module designed to analyse the footage using object recognition, pose estimation, and motion tracking algorithms to detect the user’s hand, toothbrush, and face;A feedback device that comprises a feedback interface that provides real-time visual, auditory, or haptic feedback based on brushing technique, including parameters such as angle, coverage, and movement speed and movement patternsWherein the feedback comprises instructions to the user to alter their toothbrushing technique to bring the parameters closer to a predetermined set of ideal values andWherein the system operates independently of sensors, accelerometers, or physical markers embedded in or attached to the toothbrush.

2. The system of claim 1, wherein the Al-powered software module uses pose estimation algorithms to detect and track the user’s hand movements and toothbrush trajectory, ensuring optimal brushing coverage.

3. The system of claims 1 and 2, wherein the Al-powered software module employs optical flow tracking algorithms to analyse brushing speed, angle, and coverage over time.

4. The system of any preceding claim, wherein the feedback interface provides real-time visual overlays that highlight areas of the mouth that have not been adequately brushed, issuing alerts when improper brushing angles or inadequate coverage are detected.

5. The system of any preceding claim, wherein brushing pressure is inferred through camera data by analysing motion characteristics such as hand velocity, acceleration, and abrupt changes in toothbrush trajectory, thereby eliminating the need for physical pressure sensors.

6. The system of any preceding claim, wherein the Al software module is in communication with a data storage system configured to aggregate session data, generating detailed brushing reports and personalised recommendations based on historical performance, utilising adaptive learning algorithms in the Al software module to refine feedback over time.

7. The system of claim of any preceding claim, wherein the system is adaptable to work with any type of toothbrush, whether manual or electric, identifying the type of toothbrush from the camera device video feed and adjusting the ideal parameters and algorithm accordingly.

8. The system of any preceding claim wherein the camera device and feedback device comprise the same device.

9. The system of claim 8, wherein the combined camera and feedback device comprises any consumer-grade camera-equipped device, including smartphones, tablets, and webcams.

10. The system of any preceding claim, further configured to be capable of operating in multi-user environments by identifying individual users based on unique brushing patterns and facial recognition in a single video fed, providing personalised feedback and data logging for each user.

11. The system of any receding claim, wherein the Al software module integrates Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to track and improve long-term brushing habits by analysing temporal brushing patterns, providing feedback on repetitive brushing issues such as consistently poor angles or excessive pressure over time.

12. The system of any preceding claim, wherein the Al software module is configured to use images of the user’s mouth to identify specific dental conditions including the presence of braces, crowns, or dental implants; andIs configured to adapt feedback based on the user’s specific dental condition, providing tailored brushing guidance for the identified conditions.

13. The system of any preceding claim, wherein a user in the video feed can be identified as a child, adjusting the ideal values for a child user; andwherein the feedback interface includes gamification features for child users;wherein the gamification features include providing rewards or points for achieving ideal brushing parameters, and progress tracking to encourage proper brushing habits.

14. The system of any preceding claim, wherein the feedback system offers customisable visual, auditory, and haptic feedback to cater to users with hearing or vision impairments.

15. The system of claim 1, wherein the Al software module provides a platform allowing users to compare their brushing habits against regional or national averages, positioning the system as a public health tool.

16. The system of any preceding claim wherein the camera device is configured to provide still images of the user to calibrate the Al software module programs, including images of the user for calibrating facial recognising and hand detection.

17. The system of claim 16, wherein the provided images comprise periodic images of the user’s toothbrush to calibrate toothbrush detection and tracking, especially when the toothbrush has been replaced; andwherein the Al module is configured to use the periodic images to track toothbrush degradation and alert the user, via the feedback device when their toothbrush needs to be replaced.

18. The system of claims 16 and 17, wherein the provided images include images of the inside of the user’s mouth;Wherein the Al software module is configured to determine the dimensions of the user’s mouth and specific mouth segments from the images; andWherein the Al software module is configured to diagnose different oral health conditions from the mouth images, and track the severity of the diagnosed condition over time;wherein the Al software module may adjust the ideal parameter values to address the diagnosed issues and wherein the feedback provided by the Al module includes suggesting products or instructing users to have an appointment with a profession to address the issue.

19. The system of claim 18, wherein the oral health conditions identified include at least one of plaque buildup, cavities, gingivitis or gum disease.

20. The system of any preceding claim wherein the Al module is configured to detect the presence of toothpaste on the toothbrush from a supplied image or an image of the video feed and determine if there is a sufficient amount present, based on the size of the user’s toothbrush in the image, if not the Al module will provide feedback comprising an instruction to apply more toothpaste to the toothbrush.

Images