FY24 SBIR PHASE IITOPIC NO. 423 PROJECT TITLE: "PASSIVE AND ACTIVE AI SCREENING FOR SOCIAL DETERMINANTS OF HEALTH AND INTERVENTIONS TO IMPROVE OUTCOMES IN THE CANCER PATIENT POPULATION".

It is increasingly clear that social determinants of health (SDOH) play an important, yet under-documented, role in the prognosis and treatment of cancer patients. Here an SBIR Phase II Research Program is proposed to develop artificial intelligence (AI) based tools to automate the identification of these determinants and facilitate their consideration within the clinical workflow. Technical objectives include the development of Pieces Screener made of two distinct components – an AI-driven tool that will passively, automatically, and accurately identify and categorize SDOH information from clinical notes – alongside Pieces Chat, a conversational AI agent designed to actively screen oncology patients for SDOH.
The objective is to demonstrate that a combination of these passive and active screening methods can successfully identify at least one SDOH in 35% of oncology patients, in a 12-month clinical trial at MetroHealth cancer centers. The study will also focus on evaluating user acceptance, analyzing performance metrics, and gathering patient perspectives on Pieces Screener. These metrics will permit assessment of the integration and impact of these tools in clinical settings, guide improvements of the AI chatbot, and measure the effectiveness of these tools in providing community services to patients with identified SDOH needs.

Grant Number: 75N91024C00086-0-9999-1
NIH Institute/Center: NIH
Principal Investigator: Ruben Amarasingham