Work

Precision medicine’s inevitable trajectory toward rare-disease-sized cohorts: implications for machine learning and deep learning

Year: 2026

Type: review

Abstract: Although precision medicine redefines health care by tailoring treatments to highly specific patient subgroups, the rise of rare-disease-sized cohorts (RDSCs) presents a formidable challenge. RDSCs are the smaller fragmented datasets induced by the precision medicine paradigm; they strain traditional deep learning and machine learning approaches, which typically thrive on large-scale data. In this Viewpoint, we explore the growing tension between the promise of personalised care and the limitations of current data infrastructure and associated algorithmic approaches, drawing lessons from rare disease research. We highlight the need for new methods that can handle the complexities of small, high-dimensional, and heterogeneous datasets. Importantly, although technical innovation is essential, systemic changes in biobanking, data standardisation, and cross-institutional collaboration are equally vital. Without investment in data collection, usage, and governance frameworks, the full potential of precision medicine will remain inaccessible. This Viewpoint attempts to offer a comprehensive roadmap for researchers, clinicians, and policy makers committed to advancing precision medicine in an era where "n=1" may become the new normal.

Source: The Lancet Digital Health

DOI: 10.1016/j.landig.2026.101000

Authors: Andrew Janowczyk, Doron Merkler, Olivier Michielin, Anant Madabhushi

Institutions: Georgia Institute of Technology, Emory University, University Hospital of Geneva, University of Geneva, Atlanta VA Medical Center

Language: English

Topic: Artificial Intelligence in Healthcare and Education

Subfield: Health Informatics

Field: Medicine

Domain: Health Sciences

SDGs: Good health and well-being, Reduced inequalities

Open Access status: gold

Estimated APC paid: $6,680