← Back to article | Towards comprehensive benchmarking of medical vision language models