Newest papers

Expandable Modality Experts for Medical Vision-Language Learning

DM BAPPY, Eunsu Kim, Hanjoo Lee, Lee Sh, Sangjeong Ahn · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.29881 · cited by 0 works (OpenAlex) · open-access PDF

Spanish

El aprendizaje a través de las modalidades de imagen médica requiere el uso de imágenes sin texto emparejado y el alojamiento de nuevos dominios al mismo tiempo que se mantienen las representaciones existentes. Proponemos una técnica de visión médica construida alrededor de un banco expandible de expertos en modalidad. Su Transformer Multiway combina la computación visual compartida con expertos de feed-forward deterministas y una tomografía de texto separada en los datos emparejados. Un tokenizador visual consciente de la modalidad proporciona objetivos para la verificación previa enmascarada en 19,4 millones de líneas de imagen no emparejadas en siete modalidades; la propuesta de alineamiento de contraste luego utiliza 3,23 millones de pares de imágenes-texto. Los expertos de fondos adicionales y de coherencia óptica aprenden de datos empare

Machine translation (Meta M2M100); the English original follows.

Abstract

Learning across medical imaging modalities requires making use of images without paired text and accommodating new domains while retaining existing representations. We propose a medical vision-language technique built around an expandable bank of modality experts. Its Multiway Transformer combines shared visual computation with deterministically routed feed-forward experts and a separate text pathway. A modality-aware visual tokenizer supplies targets for masked pretraining on 19.4 million unpaired image rows across seven modalities; contrastive alignment then uses 3.23 million image-text pairs. Additional fundus and optical coherence tomography experts learn from paired data while the original image and text embedding paths remain frozen. With a ViT-Large backbone, the proposed technique achieves the best or tied-best result in 24 of 25 benchmark settings across nine modalities: 10 of 11 zero-shot classification benchmarks, all 10 linear probes, and all four retrieval benchmarks. Mean balanced accuracy is 0.674 for zero-shot classification and 0.856 for linear probing, and mean bidirectional retrieval recall is 20.70%. A separate three-seed sequential experiment with matched adaptation data yields fundus and OCT retrieval gains of 14.95 and 9.31 percentage points, with no measured loss on preceding fundus or pathology tasks. The preservation argument requires unchanged existing computation paths and inference conditions, and feature-identity checks verify retention on evaluated inputs. The proposed technique provides a modular approach to learning and expanding medical vision-language representations.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Artificial intelligence, Medical imaging, Modular design, Modality (human–computer interaction), Encoder

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange