Patent Application
20230218371
The goal of the orthodontic treatment planning process is to determine where the post-treatment positions of a person's teeth (setup state) should be, given the pre-treatment positions of the teeth in a malocclusion state. This process is typically performed manually using interactive software and is a very time-consuming process. A need thus exists for an algorithm to display and evaluate treatment options.
A computer-implemented method for generating multiple orthodontic treatment options includes receiving a digital 3D model of teeth in malocclusion and generating a plurality of different orthodontic treatment plans for the teeth. The method displays a visual indication of each of the plurality of different orthodontic treatment plans for possible selection.
A computer-implemented method for displaying a user interface comprising orthodontic treatment options includes a visual indication of different orthodontic treatment plans with or without malocclusion overlays and optionally with a separate display of a digital 3D model of teeth in malocclusion corresponding with the treatment plans.
Embodiments include a system to display multiple orthodontic treatment options to a user (e.g., doctor, technician, or patient). A digital three-dimensional (3D) model of a patient's malocclusion is received input. Optional input includes one or more treatment guidelines, for example whether to apply IPR (interproximal reduction, also known as teeth shaving), duration of treatment, or others. Based on the input, the system generates several candidates for a treatment plan. These candidate treatments may be generated based on factors including user preferences, treatment appliances, treatment strategy, or types of interventions or features (e.g., IPR or attachments). The user can then select one of the candidate treatment plans from the set of treatment plan options and either make additional refinements to the selected treatment plan, begin the aligner tray manufacturing process directly from this output, or send the selected treatment plan to a lab or other facility for further refinement by a technician. Generating and displaying multiple orthodontic treatment options to a practitioner could enable greater choice, flexibility, and control by the doctor and patient, or other users.
The system receives a patient's malocclusion scan as input and displays multiple orthodontic treatment options to a user based on the input scan. Optionally, a set of doctor, technician, or patient preferences can also be included as input. For example, a doctor may wish to only see treatment options that do not require IPR, or only see treatment options that will take less than six months to complete.
Based on the inputs, a set of potential treatment plans can be displayed to the user. Some possibilities for treatment options include:
1. Appliances, for example dual arch aligners, brackets on one or both arches, brackets followed by aligners.
2. Treatment duration, for example six month treatment or two year treatment.
3. Treatment strategies, for example expansion, bite closure, or others.
4. Aligner features or lack thereof, for example attachments versus no attachments.
5. Doctor-specific treatment strategies, for example characteristic setups from a group of doctors.
6. Treatment finishing approaches, for example overcorrection.
7. Treatment interventions, for example IPR.
8. Doctor- or technician-specific preferences.
9. Patient-specific preferences, for example decreased pain (which could then increase the time of treatment) or lowered cost of treatment.
The overlay opacity tool can also be used to eliminate the shading in the view or show full shading on the digital 3D model of teeth. By varying the shading in such manner and selectively displaying the malocclusion separately, the user interface can be configured to display one or more of the following: the digital 3D model of teeth in malocclusion shown separately with malocclusion overlays; only the malocclusion overlays of the treatment plans without the digital 3D model of teeth in malocclusion shown separately; the digital 3D model of teeth in malocclusion shown separately with the treatment plans and without the malocclusion overlays; and the treatment plans without the digital 3D model of teeth in malocclusion shown separately and without the malocclusion overlays.
The user interface in
There are several ways to use a computerized system to learn multiple treatment options that can be displayed.
Final setups can be automatically created such that they adhere to a set of rules. For example:
Treatment strategies, for example expansion—move the molars in the positive direction normal to the arch form between the malocclusion and the setup.
Treatment interventions, for example IPR—expand the front teeth between malocclusion and setup to make space and do not apply any IPR in the setup, or, do not expand the front teeth and apply IPR.
An example of a rule-based approach to generate setups is disclosed in PCT Patent Application Publication No. WO 2019/069191, which is incorporated herein by reference as if fully set forth.
One optimization algorithm that creates final setups based on optimizing a set of metrics subject to some constraints in described in PCT Patent Application Publication No. WO 2020/026117, which is incorporated herein by reference as if fully set forth. Table 1 provides exemplary pseudocode for generating final setups for this optimization-based approach.
The method for this option can use metrics and/or constraints to customize final setups. For each treatment option, the method can either modify the optimal value for the metric, (change the penalty term (Pi) in the Scoring function), or modify the constraints (change the Constrain function), or do both simultaneously, to achieve the desired outcome. For example:
Treatment duration—Constraints: increase limits on how much teeth are allowed to move during treatment. The Constrain function would move the teeth in the current state to a position in which the movement between the maloccluded state and the current state is less than a certain amount.
Treatment strategy, e.g., expansion:
Constraints: require minimal amount of tooth movement normal to the arch form to be greater than a threshold amount. The Constrain function would move the teeth in the current state to a position in which the expansion amount between the maloccluded state and the current state is no less than a certain amount.
Metrics: penalize movements that are less than a threshold amount. The penalty term in the scoring function would measure how much less the current movement is than the ideal amount of expansion.
Aligner features, e.g., attachments—Metrics: to minimize the number of attachments that are necessary, penalize certain types of tooth movement that would require attachments to be placed. The penalty term in the scoring function would measure the amount of tooth movement for certain types of movement (e.g., root torque).
Treatment plans from previously treated patients are collected. These plans are separated into groups according to specific characteristics, for example:
Group 1: Cases with expansion between malocclusion and setup.
Group 2: Cases without expansion between malocclusion and setup.
Group 3: Cases with IPR in the setup.
Group 4: Cases with no IPR in the setup.
Group 5: Cases with fewer than 15 stages.
Group 6: Cases with greater than 15 stages.
Group 7: Cases with movement between malocclusion and setup for all tooth types.
Group 8: Cases with movement between malocclusion and setup for anterior teeth only.
Group 9: Cases planned by one member of a group of doctors.
From each of these (non-exhaustive) groups of cases, a different machine learning model is developed. When a new case is received by the system, each of these unique machine learning models can be applied to generate a treatment plan for each of these groups.
An example of a machine learning model is disclosed in co-pending Provisional Patent Application entitled, “System to Generate Staged Orthodontic Aligner Treatment,” and filed on even if fully set forth.
Preferred treatment strategies for specific users can also be learned by the system using one of several approaches identified below. These specific users can include, for example, doctors, technicians, patients, or others. Cases that adhere to specific preferred treatment methodologies can then be displayed as one or several of the options for treatment plans.
User feedback: Obtain preferred treatment strategies from users (e.g., doctors, technicians, or patients), or receive a completed form with their preferences.
Data analytics: Analyze cases from a specific user (e.g., doctor) and derive patterns.
Learn from data: Provide the user an exercise in which multiple setups for the same case are presented, and have the user either rank the setups or select which setup is preferred. Use this exercise to train a model to predict one or more of a set of setups that are most preferable to this user, and display these setups to that user.
Step 1: User (e.g, doctor or technician) has submitted a case and is offered the opportunity to participate in the exercise.
Step 2: User is instructed to select which setup they prefer.
Step 3: User is shown a malocclusion and three versions of a setup and can select which version they prefer. This step may be repeated multiple times with different cases.
Step 4: For a new case that has been submitted, the user (e.g., doctor or technician) is shown multiple treatment options with the option that is best customized to their preferences highlighted.
The user interfaces in
The user interfaces in
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2021/054027 | 5/11/2021 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 63033881 | Jun 2020 | US |
