Health AI Developer Foundations Overview

Developing AI for health applications can be challenging for several reasons:

  • data availability is limited;
  • modelling experimentation requires a lot of compute
  • expert labeling is expensive; and
  • the number of potential applications to tackle is vast.

Health AI Developer Foundations (HAI-DEF) turbocharges AI development for healthcare applications with open weight models specialized for health domains. These models can be adapted for the use-cases you care about requiring less data and compute than traditional approaches. HAI-DEF includes rich examples of how to use the models in the form of Google Colab Notebooks. It also includes libraries to streamline ingestion of processing of specialized medical data like 3D imaging and digital pathology. HAI-DEF is tightly integrated into Google Cloud so developers can take ideas from research through to productionization on the same platform. HAI-DEF is an evolving effort; we will continue to add new models and resources to cover more healthcare applications.