Shift happens: a fairness-oriented framework for medical classification under hidden bias

Many medical AI models perform unevenly across patient groups because they learn shortcuts from biased data. These hidden biases make models less reliable and less fair in real-world use. This work aims to develop a system that remains accurate and fair across different patient subpopulations, even when those groups are not explicitly labeled.

We introduce DPE-Former, a model that combines prototype-based learning with transformer attention. The system trains several complementary classifiers on balanced subsets of data, each capturing different aspects of the population. A transformer module then learns how to combine its outputs in an adaptive way, helping the model make more balanced decisions across unseen or minority groups.

Across diverse datasets, including prostate ultrasound, skin lesion images, and cardiac patient records, DPE-Former achieved higher accuracy on underrepresented groups and more consistent performance overall compared to standard training methods.

DPE-Former offers a simple yet effective approach to reduce hidden bias in medical AI. By improving fairness and reliability across both image and tabular data, it supports more equitable decision-making in clinical applications such as cancer diagnosis and cardiac care. Code is publicly available at https://github.com/minhto2802/prototypical-ensemble-med .