As it stands, doctors spend about 6 hours (or over 50 percent of their workday) in Electronic Health Records (EHR), taking notes on their interactions with patients. Consequently, less time is available for actual patient care. But with voice recognition technology, Google hopes that new tools could “be used to document patient-doctor conversations and help doctors and scribes summarize notes more quickly.”
In a new research paper titled “Speech Recognition for Medical Conversations,” Google demonstrates that existing technology can be used to create Automatic Speech Recognition (ASR) models for transcribing medical conversations. Currently, many of the ASR solutions in the medical field are geared toward transcribing doctor dictations, which is to say, a single person using predictable medical terminology. Google’s research suggests that an ASR model may be able to understand and transcribe actual conversations, with topics ranging from the weather to medical diagnoses.
Google plans to build upon its technology by working alongside doctors and researchers at Stanford University who have already conducted research on how best to improve the notation process in the healthcare profession. “We hope these technologies will not only help return joy to practice by facilitating doctors and scribes with their everyday workload, but also help the patients get more dedicated and thorough medical attention, ideally leading to better care,” Katherine Chou, a Google product manager, and Chung-Cheng Chiu, a software engineer on the Google Brain Team, noted in a blog post.
In fact, there’s already a pilot study underway, and teams are looking to determine what sorts of clinically relevant information can be deduced from medical conversations, helping cut down on the amount of time doctors must spend with EHR. Don’t worry, the tech giant notes, “The study is fully patient-consented and the content of the recording will be de-identified to protect patient privacy.”
