Benchmark RadarAI BENCHMARK PROFILE
LoMeVQA
LoMeVQA is a benchmark for longitudinal medical visual question answering, with 206K VQA pairs across five tasks: progress classification, progress description, progress report generation, differential region grounding, and differential region description.
- Released
- 2026-07-30
- Readiness
- Runnable
- Primary field
- Health & Life Sciences
Why it matters
Longitudinal medical reasoning is underexplored in MLLMs, and this benchmark provides a comprehensive testbed. It reveals limitations in temporal reasoning and supports future improvements in medical AI.
Motivation
In clinical practice, patients often undergo multiple imaging examinations over successive visits, yielding longitudinal data.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.