According to the latest news, technology companies similar to Google are developing language models optimized specifically for the medical field. According to the Microsoft research team, large language models like GPT-4 have great potential in the medical field. These models are expected to accelerate medical processes, such as by processing large-scale unstructured patient data to improve the efficiency of cancer drug development.
Although GPT-4's training data is not specific medical data, but general-purpose Internet data, it excels at constructing complex clinical studies that can be organized according to specified criteria. Compared to systems currently developed for this task, such as Criteria2Query, GPT-4 performs better. In addition, GPT-4 can also play a role in processing medical images and other biological data.
Microsoft has also introduced language models such as BioGPT specifically for medical tasks, but will rely heavily on GPT-4 in the future. Another application of AI in medicine is multimodal models, which can process genetic data, proteins, and other biological data in medical images that are not limited to text.
To provide biomedical imaging data to medical professionals, Microsoft is developing LLaVA-Med, a specialized chatbot. At the same time, Google also introduced the Med-Palm M, a multimodal medical model that solves multiple medical tasks and provides chat capabilities.
This vision represents a dream of evidence-based precision health, and large-scale language models will play a key role in driving us towards this exciting and transformative future.