Ever since ChatGPT burst onto the scene a year ago, AI has gone from zero to sixty in a flash. AI is being used for many functions: content, clinical applications and administrative applications. Both Microsoft and Google Cloud have recently announced new AI platforms designed for healthcare use. 
Microsoft Azure and Fabric
Azure, which is a cloud division of Microsoft, has released these new capabilities:
- Patient timelines – used to extract specific components from unstructured data to be organized chronologically for a patient’s history.
- Clinical report simplification – used to simplify medical jargon into lay terms for better patient understanding.
- Azure AI Healthbot (currently being previewed) – when partnered with existing chatbots, can improve the chatbots’ ability to engage with and answer patient questions. It is set up in a way that ensures the information is only being pulled from authorized documents.
Fabric, an analytic platform, is being tied to Microsoft Cloud for Healthcare. While Cloud for Healthcare has been around for a few years already, connecting it to Fabric will enable healthcare organizations to bring in data from various sources – EHR, lab results, claims and medical devices – and combine it into a usable form.
Microsoft previously released Nuance DAX Copilot, a generative AI-enabled notetaker product, which has received good reviews from clinicians. Other functionalities coming from Microsoft include an integration of GPT into Epic EHR that will automatically formulate responses to messages from patients, as well as integrating AI into Epic’s hospital database to improve ease of asking questions and querying the database.
New from Google Cloud
New generative AI capabilities from Google Cloud include a medically-focused search capability to connect to clinical notes and other data sources within Vertex AI. This function is designed to help healthcare organizations improve pain points such as workforce shortages, burnout and administrative burden. Working in conjunction with Med-PaLM 2, searches can be connected to outside sources and patient medical records.