Exploring Voice-Based AI in Healthcare

Updated on September 30, 2023
Illustration of the words A.I. (artificial intelligence)

Voice recognition technology has become commonplace in society, with personal assistant programs gaining frequent use across countless mediums. As described in Forbes, AI and ML are increasingly used in healthcare applications to expand clinical efficiency, boost diagnosis speed and accuracy, reduce physicians’ duties, and improve patient outcomes. Healthcare AI can also help with administrative tasks like appointment scheduling and insurance claims processing, freeing up time for clinicians to focus on patient care. Especially during the pandemic, voice AI bots helped to screen patients; for instance, Apple’s Siri guided users through the CDC’s COVID-19 evaluation questions. 

As a result of voice-based technology, patients can get prompt answers to their questions about symptoms, therapies, and pharmaceutical availability using smart healthcare voice assistants like Alexa and Google Home. Voice assistants can also offer patients up-to-date information on their mental health; audio acoustics change in response to a person’s health. These audio qualities can help diagnose various physical and mental diseases and conditions. Chatbots—computer programs that simulate and process written or spoken human conversation—also allow people to interact with digital devices as though they were communicating with a live person.

Current Trends

According to the National Center for Biotechnology Information (NCBI), in 2020, the World Health Organization developed a chatbot via WhatsApp and Facebook to find answers about how to protect themselves from COVID-19 and how to understand news and facts related to the disease. This started a trend in chatbots in the healthcare market; health chatbots are anticipated to grow by over $498 million by the end of 2029. Chatbot technology can assist consumers in scheduling appointments and acting as health consultants.

Recently, ChatGPT, a natural language processing tool driven by AI developed by the company OpenAI, has attracted the attention of researchers and medical professionals. ChatGPT collects web data stored on a server by a human programmer. The algorithm can anticipate which word will appear in a sentence based on context; it can answer questions even if it has not encountered a specific sequence of words before. ChatGPT generates a response based on facts on its server. According to the company, ChatGPT may also acknowledge errors, respond to follow-up inquiries, and dismiss inappropriate ones. 

Dr. Victor Tseng co-authored a 2023 study in which ChatGPT demonstrated that it could pass a medical licensing exam. He claimed that his coworkers experimented with ChatGPT and were intrigued when it correctly identified fictitious patients in hypothetical situations. He asserted that it could pass the three-part medical exam that U.S. medical students must pass to be licensed to practice medicine. Tseng’s research group used 305 questions from the June 2022 sample test made available to the public. They were not part of the data set used to train ChatGPT, and without any specialized training, the chatbot nearly passed all the exam’s components. 

The Future

A.J. Ghergich, a Forbes Technology Council member, recently outlined four major ways he believes voice recognition could improve healthcare:

1. More Data: Getting data into electronic health records (EHRs) must be easier, faster, and more accurate. Voice recognition can enable placing data into EHRs faster and more accurately.  With appropriate privacy and security safeguards, big data can further advance machine learning and artificial intelligence technologies in healthcare.

2. Better Experience: Voice recognition, as it continues to improve, will support faster and easier transactions and exchanges, improving patient experience. This technology could translate to easier appointment-making, quick and more accurate intake of medical histories and other patient data, and the application of insurance, billing, and payment matters. 

3. Improved Care: Voice recognition can reduce the burden of information exchange and data collection between patients and physicians and allow more attention to the patient-professional relationship.

4. New Insight: As voice recognition technology advances, it will yield insight beyond what the patient says to provide more reliable evidence about how the patient is doing. The patient’s voice will become a data source, similar to blood pressure, temperature, and other vital signs.

Those working in healthcare and technology should continue to follow these trends, looking for ways to collaborate to learn how to best serve patients using voice-based and other intelligent technologies.

Dr. Rekha Bhandari
Dr. Rekha Bhandari
Founder at MedElite Group

Dr. Rekha Bhandari is a well-respected and accomplished doctor and Founder of MedElite Group, based in New York City. Her specialty is in geriatric care and the integration of digital health technologies in the hospital.