Digital Themes

Generative AI in healthcare

Within the past year, generative AI (gen AI) has impacted nearly every industry, and healthcare is no exception. Even so, industry leaders emphasize that gen AI is not a replacement for human thought, medical analysis, or medical advice.

According to a paper published in June 2023, “The Emerging Role of Generative Artificial Intelligence in Medical Education, Research, and Practice,” gen AI performed several feats: It could predict renal illness, generate radiology reports, and predict the development of hematological disorders.

The paper’s authors cautioned that ChatGPT “has a limited role in critical thinking or making moral or ethical decisions.” gen AI tools lack a sociocultural background, and despite being relatively context-aware, they are “not trained to make difficult, contextual, evidence-based medical decisions when it comes to clinical decision-making for a fictional patient with common symptoms.” Most damagingly, the paper reported that current gen AI tools have made biased and discriminatory statements based on currently available data.

However, the above drawbacks have been accompanied by successes. ChatGPT, which debuted to great fanfare in late 2022, has become an efficient and inexpensive addition to customer services offerings. According to the June 2023 report in Cureus, a specialized gen AI can summarize patient data and provide well-written notes in standard English. Generative AI chatbots can also handle customer concerns and produce helpful responses.

Below are some business benefits of gen AI tools, as applied to healthcare: 

  • Improved patient engagement, via the use of chatbots and generated responses/text
  • The generation of therapeutics; per a Cell Reports Methods report: “For drug design, the models themselves would generate potential therapeutics for a particular disease state.”
  • Surgical planning and the virtual reality (VR) simulations of procedures
  • The creation of synthetic medical images to replace or accompany human-drawn interpretations
  • Greater empathy within patient-provider or patient-institution communications

 

References:

Badwan, Bara A., et al. “Machine learning approaches to predict drug efficacy and toxicity in oncology.” Cell Reports Methods. February 27, 2023. https://www.sciencedirect.com/science/article/pii/S2667237523000243

“Generative AI – Opportunities and Risks in Digital Health.” LinkedIn Pulse. LinkedIn. May 16, 2023. https://www.linkedin.com/pulse/generative-ai-opportunities-risks-digital-health-sonal-panda

Shoja, Mohammadali M., et al. “The Emerging Role of Generative Artificial Intelligence in Medical Education, Research, and Practice.” Cureus. June 24, 2023. https://pubmed.ncbi.nlm.nih.gov/37492829/

Related content