Potential Projects for MS Students

Online recommendation engine for mental health technologies

Project summary: The Child Mind Institute, in conjunction with groups at the Pasteur Institute in Paris and Google, are just starting to build an online resource for anyone in the world to find what technologies exist to help people with specific mental disorders. The technologies will include smartphone applications, sensors, wearable devices, and computer games. “Helping” can mean assessment, diagnosis, monitoring progression of a disease, tracking symptom severity over time, predicting outcome measures, etc. The student(s) would help develop such an online resource to make it useful and easy to use for patients and clinicians alike. One of our goals is to create a rating and recommendation system that a user would use to filter and find technologies according to various criteria, including scientific validity, applicability to a given condition, user engagement, burden, cost, and availability.  This is principally an informatics project, so the student(s) would either need to have computational skills and an interest in comparative literature reviews to determine how best to rate technologies, or programming experience for full-stack web app (in particular JavaScript) development.

Mapping psychiatric symptoms and behaviors to sensors for monitoring mental health

Project summary: The Child Mind Institute needs to evaluate what types of devices can be used to obtain objective measures for assessing and monitoring childhood psychiatric illnesses over time. The student(s) would help evaluate which sensor technologies (microphone, camera, accelerometer, skin conductance electrodes, etc.) in what currently available devices are best positioned to assess and monitor physiological and behavioral symptoms in specific childhood psychiatric illnesses.  The student(s) would need to have a strong interest in learning about child psychiatric illnesses and strong organizational skills for helping to develop a database of information that maps sensors to behavioral and physiological symptoms.

Analyze voice data -- extract features, perform predictive modeling

Project summary: The Child Mind Institute has been collecting voice data from children with selective mutism, and will be collecting a lot more voice data from children with different psychiatric illnesses.  The student(s) would use existing software and write new software where necessary to analyze voice data to extract features and find patterns in the features to better understand how speech is affected by psychiatric illnesses and how context affects their speech.  For this project, the student(s) would need to have an interest in analyzing voice and other mobile health data and very strong computational and programming skills (Python preferred).


500 W. 120th St., 351 Engineering Terrace, New York, NY 10027    212-854-4460                 
©2016 Columbia University