TBI Sleep Monitor

Overview

This project focuses on the development of a remote patient monitoring (RPM) platform for the management of traumatic brain injury (TBI), including the development of algorithms for adaptive, patient-specific functional recovery predictions and symptom tracking, as well as technologies for the care of patients with chronic TBIs. The project's objectives include developing machine learning (ML) models for TBI detection and assessment, creating patient- and provider-friendly data visualization interfaces to display ML outputs and symptom data, and conducting system testing and integration.

Features

Onboarding

When users first onboard onto the app, they are required to enter their Glasgow Coma Scale from their clinical records, with the option to put their subscores (eye, verbal, and motor responses). They also have to input their Marshall Computed Tomography (CT) Score for more information on their current condition. To track their sleep patterns and monitor more continuously, users also have the option to pair their Apple Watch during onboarding.

Progress Resource Tracking

This app uses two primary data sources to track recovery for patients with traumatic brain injury: the daily Post-Concussion Symptom Scale (PCSS) Questionnaire (middle screen) and health data collected from the Apple Watch, the specific measurements of which users can view health data.

Using both of these data inputs, the machine learning-based prediction model will generate four scores, seen in the rightmost screen: the symptoms score (based on ratings in the PCSS Questionnaire), sleep score (score from an aggregate of sleep quality data), and future score (measurement of user's recovery rate), and recovery score (combination of the three previous scores).