Metraipolis › archive › dc_c7fa22b9311d

ViSTA: A Simple Bridge Extends Visual Alignment to Clinical Time-Series Understanding in Multimodal LLMs — arXiv:2609.31448

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.31448
License (as archived)
CC0-1.0
Archived
2026-10-08T01:39:57.117Z
SHA-256 of the archived text
75b7fbf3b09ee4dbca5d283e3d6ed358a457919d372c55cea84bce85211eaad2
Stable URL of this page
https://metraipolis.com/documento/dc_c7fa22b9311d

How to cite

ViSTA: A Simple Bridge Extends Visual Alignment to Clinical Time-Series Understanding in Multimodal LLMs — arXiv:2609.31448 — arXiv (en), https://arxiv.org/abs/2609.31448 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c7fa22b9311d (Metraipolis archive dc_c7fa22b9311d, sha256 75b7fbf3b09ee4dbca5d283e3d6ed358a457919d372c55cea84bce85211eaad2)
BibTeX
@misc{dcc7fa22b9311d,
  title = {ViSTA: A Simple Bridge Extends Visual Alignment to Clinical Time-Series Understanding in Multimodal LLMs — arXiv:2609.31448},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.31448},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c7fa22b9311d; sha256 75b7fbf3b09ee4dbca5d283e3d6ed358a457919d372c55cea84bce85211eaad2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.31448. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Clinical prediction models estimate risk from patient measurements, while large language models support medical text understanding and question answering. Yet their language capabilities do not ensure accurate prediction from structured, high-dimensional clinical time series. Improving this ability would connect risk estimation with flexible questions about a patient's evolving condition. We introduce ViSTA, a compact adapter that incorporates irregular numerical measurements into a pretrained vision-language model's chart representations. It learns corrections to visual tokens while leaving all pretrained parameters unchanged. On MIMIC-IV, ViSTA has the highest mean scores among the compared adaptations on all four metrics for acute kidney injury and mortality prediction across models with 2-9 billion parameters. With 0.516 million trainable parameters, the 2-billion-parameter model reaches an area under the ROC curve of 0.7376 for acute kidney injury, compared with GPT-5.6 Sol's 0.7380 with text input and high reasoning effort. Training for temporal question answering yields 69.27% accuracy at 4 billion parameters with over 90% fewer trainable parameters than low-rank adaptation using charts or numerical text, at a 2.82-4.88 percentage-point accuracy gap. ViSTA extends pretrained language models to numerical prediction and temporal questions.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-28 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.31448 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T01:12:14.597Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).