AI Diagnoses Endocrine Cancers

An AI application has been developed to diagnose endocrine cancers with increased speed and accuracy. The AI, presented at ENDO 2025, is designed to be computationally efficient and universally accessible, even on smartphones. It aims to democratise expert-level cancer diagnostics.

The AI uses deep learning architectures like EfficientNet and ResNet to analyse medical data, including CT scans, MRIs, ultrasounds, and histopathology images. It has demonstrated high diagnostic accuracy, reportedly exceeding 99% in validation datasets across multiple endocrine cancer types. The AI was trained and validated using anonymised image datasets representing diverse populations.

By providing clinicians and primary care providers with expert diagnostic support, the technology can reduce diagnostic errors and accelerate treatment decisions. This offers hope for earlier detection, more precise treatment, and better survival rates, particularly in resource-limited settings.