o Specific Aims of the Real-vision project
Aim 1: Development of AI-generated Pareidolia test (Month 0-2)
Hypothesis 1: Development of AI-generated pareidolia testing will be feasible to incorporate in RealVision and feasibility of personalization based on GPS and self-report data on race and culture will be established.
Aim 2: Development of mobile app (Month 0-3)
Hypothesis 2: A mobile platform to adapt the paper-based test and provide an appropriate methodology for pareidolic evaluation in a non-clinical setting.
Aim 3: Examine the feasibility, acceptability, usability, and preliminary effectiveness of eye-tracking and NLP algorithms to identify pareidolias (Mon 3-4)
Hypothesis 3: Anonymized and tagged data generated by the participants, categorized by speech and eye tracking analysts will help elucidate flags specific to visual misperceptions.
Aim 4: Examine the correlation combined eye tracking and NLP algorithms and clinician-administered pareidolias testing (Month 3-6)
Hypothesis 4: Eye tracking and NLP algorithms can be used as prognostic markers of neuropsychiatric disease in participants presenting symptoms and people at risk of neuropsychiatric diseases.
Aim 5: Examine barriers and facilitators to use of RealVision in real-world settings through semi-structured interviews using the Consolidated Framework for Implementation Science (Month 2-6)
o Contribution of each team member
Dr. Revankar will lead the neuroscientific know-how of testing and evaluation of neuropsychiatric diseases with his background in Clinical Neurology. Dr. Fortuna will lead community-engaged user centered design efforts with her background in psychiatry and peer-support tech. Dr. Singh, with her background in Applied Physics, will develop an explainable computational framework integrating AI and Neuropsychiatry.
o Results obtained in response to the described plan.
1. App development - Was earlier planned to be built off existing app code. However, considering the specialized nature of our study, coding has been done from the ground up - both for Android and iOS platforms.
2. The name of the app has been changed to avoid potential conflicts with similar registered application.
3. Android version of the app has been completed. A web browser version for maintenance and management of participant/patient data has been additionally developed.
4. As per schedule, Aim-1 and Aim-2 are in completion phase. Aim-3, 4 and 5 are underway with the study plan on track for completion in the designated time. iOS version of the app and additional changes will be performed during Aim-3 and Aim-5 of the study.