Kazi Nazmul Hasan, Master’s Defense
- Thursday, November 18, 2021 from 1:45pm to 2:45pm
Abstract: Behavioral health including mental health is one of the most common but expensive healthcare conditions in the world. Yet, more than half of the patients go untreated due to various reasons such as lack of access to resources and clinicians. One key underlying reason is the time-consuming and tedious nature of the clinical documentation process, which is a key component of clinical practice and plays a vital role in ensuring the consistency of care, illustrating the provided care for reimbursement, and improving the overall healthcare system. While providers rely on Electronic Health Records (EHRs) to compile and share clinical notes, time-consuming data entry is considered one of the prime downsides of EHRs. Many practitioners are spending more time in EHR documentation than direct patient care which adds to patient dissatisfaction and clinician burnout.
In this thesis work, we explore the feasibility of developing an end-to-end clinical transcription tool that fully automates the documentation process for behavioral health clinicians. We divide the task into several sub-tasks and mainly focus on the following: 1) extraction and classification of important information from patient-provider conversations, and 2) generation of clinical notes from extracted information. First, we audio record a large number of simulated provider-patient conversations using remote conferencing platforms. We then use existing speech-to-text techniques to generate digital transcripts of these conversations. For extracting and classifying the information, we train a transformer language model. Furthermore, we develop a rule-based natural language generation module that formalizes the extracted information and synthesizes them into clinical notes. The overall pipeline shows the potential of automatically generating draft clinical notes and reducing the documentation time by 70-80%. The findings of this work have implications for health behavioral care providers as well as machine learning and natural language processing application developers.
- Gianforte School of Computing