When studying the effect of clinical trial drugs on quality-of-life, researchers, pharmaceutical companies, and doctors often use ecological momentary assessment (EMA), a methodology for real-time assessment of experience and behavior in a subject's natural environment. Traditional methods of collecting EMA data have included paper-based patient diaries, shown to be ineffective in providing accurate data. Recently, electronic methods of collecting EMA data have been researched. We propose to study the feasibility of developing automated speech recognition (ASR) as a hardware/software application deploying real-time diary solutions in cancer control research.
The aims of this research are to: ? 1) Conduct research on existing technologies and methodologies for real-time data collection, conduct a feasibility analysis of the applicability of ASR technologies to data collection, and identify types of EMA most suitable for automated speech recognition systems. ? 2) Research and collect common grammars utilized in EMA. ? 3) Evaluate and design a system prototype for usability and applicability of EMA data collection. ? 4) Build a prototype ASR-based prototype for data collection from a simulated EMA protocol (10-12 questions). The goal is to achieve benchmark of 95% call sessions completed, 95% recognition rate, 95% accuracy. ? 5) Run a feasibility and benchmarking test. ? 6) Build a foundation for system expansion into Phase II. For Phase II, Spacegate intends to partner with companies with expertise in areas such as clinical data collection and call centers to design and test distributed architecture for the data collection process. The long term goal is to have a robust system that is cost effective, regulatory compliant and easy for cancer patients who are not able to use a computer to handle all aspects of real-time EMA data collection needs. ? ?
