Machine learning-based diagnostic classifier
Systems and methods for utilizing machine learning to generate a trans-diagnostic classifier that is operative to concurrently diagnose a plurality of different mental health disorders using a single trans-diagnostic questionnaire that includes a plurality of questions (e.g., 17 questions). Machine learning techniques are used to process labeled training data to build statistical models that include trans-diagnostic item-level questions as features to create a screen to classify groups of subjects as either healthy or as possibly having a mental health disorder. A subset of questions is selected from the multiple self-administered mental health questionnaires and used to autonomously screen subjects across multiple mental health disorders without physician involvement, optionally remotely and repeatedly, in a short amount of time.
1. A system for evaluating a user, the system comprising:
a microphone;
a camera positioned to capture an image of the user and configured to output video data;
a memory containing machine readable medium comprising machine executable code having stored thereon instructions for performing a method of evaluating the user; and
a control system coupled to the memory comprising one or more processors, the control system configured to execute the machine executable code to cause the control system to:
record, by the camera, a set of test video data during a time period;
record, by the microphone, a set of test audio data during the time period;
process the video data to assign a plurality of pixels to a face of the user;
analyze the plurality of pixels to determine whether the face of the user is within a frame captured by the camera;
in response to determining that the face of the user is within the frame captured by the camera, process the plurality of pixels to output video features associated with the user;
process the audio data to identify sounds representing a voice of the user and output audio features associated with the user;
process, using a machine learning model, the audio and video features, wherein the machine learning model was previously trained with a set of training data comprising audio and video data recorded from a plurality of individuals with labels indicating whether each of the plurality of individuals has one of a plurality of characteristics; and
output an indication of whether the user has at least one of the plurality of characteristics.
2. The system of claim 1 , wherein the processing the audio data to identify sounds is in response to determining that the face of the user is within the frame captured by the camera during the time period.
3. The system of claim 1 , wherein the analyzing the plurality of pixels includes determining whether an entire face of the user is within the frame captured by the camera, and wherein the processing the plurality of pixels is in response to determining that the entire face of the user is within the frame captured by the camera.
4. The system of claim 3 , wherein the plurality of pixels is assigned to a boundary of the face of the user, and wherein the determining whether the entire face of the user is within the frame captured by the camera includes determining whether the boundary of the face of the user is within the frame captured by the camera.
5. The system of claim 3 , wherein the determining whether the entire face of the user is within the frame captured by the camera includes determining whether all of the plurality of pixels of the face of the user is within the frame captured by the camera.
6. The system of claim 1 , wherein the video features include facial expressions of the user.
7. The system of claim 1 , wherein the audio features include tone of voice of the user.
8. The system of claim 1 , wherein the recording, by the microphone, further includes initiating the recording upon determining, by the control system, that the user is speaking.
9. The system of claim 1 , wherein the control system is further caused to:
preprocess the recorded set of test video data to identify a plurality of video segments during the time period, each video segment corresponding to one question in a series of questions and comprising a time window; and
preprocess the recorded set of test audio data to identify a plurality of audio segments during the time period, each audio segment corresponding to one question in the series of questions and comprising a time window.
10. The system of claim 9 , wherein the control system is further caused to:
preprocess the plurality of audio segments and the plurality of video segments to identify overlapping time windows; and
output a set of integrated audio and video segments based on the identified overlapping time windows.
11. The system of claim 10 , wherein only the audio and video features associated with the set of integrated audio and video segments are processed using the machine learning model.
12. The system of claim 1 , wherein the time period corresponds to the user reading a text.
13. The system of claim 12 , further comprising a display configured to be placed in front of the user and displaying the text, such that the camera is positioned to capture the image of the user in front of the display.
14. The system of claim 12 , wherein the text includes a series of questions from questionnaires answers for each question, and wherein the questionnaires are associated with the plurality of characteristics.
15. The system of claim 14 , wherein each of the plurality of characteristics is indicative of a mental health disorder.
16. The system of claim 1 , wherein the machine learning model is at least one of: a generalized linear model, a regression model, a logistical regression model, and a supervised machine learning classification model.
17. The system of claim 16 , wherein the machine learning model includes a decision tree.
18. A system for screening mental health of a user, the system comprising:
a memory containing machine readable medium comprising machine executable code having stored thereon instructions for performing a method; and
a control system coupled to the memory comprising one or more processors, the control system configured to execute the machine executable code to cause the control system to:
receive a set of test video data representing a face of the user while the user is reading text;
process the set of test video data to output a set of video features associated with the face of the user;
receive a set of test audio data representing a voice of the user while the user is reading the text;
identify sounds representing the voice of the user;
process the set of test audio data to output a set of audio features based at least in part on the identified sounds presenting the voice of the user;
process, using a machine learning model, the set of video features and the set of audio features, to output an indication of the mental health of the user,
wherein the machine learning model is at least one of: a generalized linear model, a regression model, a logistical regression model, and a supervised machine learning classification model, and
wherein the machine learning model was previously trained with a set of training data comprising audio and video data recorded from a plurality of individuals with labels indicating whether each of the plurality of individuals has one of a plurality of mental health disorders; and
output an indication of whether the user has a mental health disorder.
19. The system of claim 18 , wherein the control system is further caused to determine whether the face of the user is within the set of test video data, and wherein the set of test video data is processed in response to determining that the face of the user is within the set of test video data.
20. The system of claim 19 , wherein the set of test video data is generated by a camera, and wherein the determining whether the face of the user is within the set of test video data includes determining whether all of a plurality of pixels of the face of the user is within a frame captured by the camera.
21. The system of claim 19 , wherein the set of test video data is generated by a camera, and wherein the determining whether the face of the user is within the set of test video data includes determining whether all of a plurality of pixels assigned to a boundary of the face of the user is within a frame captured by the camera.