AAIMC-OC Consortium

Democratising AI diagnostics for equitable cancer care

The Application of AI‑defined Molecular Classifications in Ovarian Cancer (AAIMC-OC) is a global research consortium aiming to develop and rigorously evaluate AI‑defined cancer diagnostics using federated learning.

The challenge

Targeted cancer treatments are transforming outcomes in ovarian cancer. However, many patients cannot access these therapies even when they are now becoming cheaper because the diagnostic tests required to guide treatment are costly, specialist, and unavailable in many parts of the world.

Artificial intelligence (AI) offers a potential solution by analysing standard pathology images (H&E slides and images), which are already used routinely for cancer diagnosis globally. Yet current AI tools are typically developed using data from high-income countries and may not perform reliably across diverse populations or healthcare systems.

This risks reinforcing existing global inequalities in cancer care.

Our approach

AAIMC-OC aims to address this challenge by combining international collaborative datasets and federated AI development and real-world validation.

Using federated learning, we enable AI models to be developed and evaluated across multiple countries without transferring patient private data. This preserves privacy, increasingly important to be legislatively compliant for AI development, while allowing models to learn from diverse populations.

Our work focuses not only on developing AI, but on ensuring that these tools are robust, fair, and usable in real healthcare settings worldwide.

What makes AAIMC-OC different

  • Global, real-world datasets
    Our collaboration spans high-income and low-and-middle-income countries, ensuring representativeness and relevance.

  • Focus on implementation, not just development
    We aim to address the barriers of data systems, as well as governance and workforce readiness for clinical integration

  • Federated, privacy preserving infrastructure
    Patient data remain securely within each institution, enabling collaboration whilst reducing data sharing.