arXiv cs.AI· Amirmohammad Mohammadi, Rupak Khadka, Kathleen A. Grant, Mary Lauren Benton, Joshua Peeples·· 3 小时前AI 评分35
用参数高效微调从非人灵长类肝脏病理切片预测饮酒状态：少样本视觉-语言模型概念验证
Predicting Ethanol Drinking Status from Nonhuman Primate Liver Histology: A Proof of Concept for Parameter-efficient Fine-tuning in Few-Shot Vision-Language Modeling
AI 导读
一项发表于 IEEE BIBM 2026 的研究验证了在计算病理学中用参数高效微调（PEFT）结合视觉-语言模型进行少样本全切片图像分类。在 MATRR 猴肝组织数据子集上，四-shot 设置下 ABMIL 基线的平均平衡准确率为 43.5%，加入 adapter 后提升至 78.1%。注意力分析显示文本通路适配产生局部化热图，视觉通路适配则更弥散。
正文 · 原文
Abstract:Combining vision-language models (VLMs) with parameter-efficient fine-tuning (PEFT) offers a promising approach to few-shot whole-slide image (WSI) classification in computational pathology. This study evaluates this approach for predicting ethanol drinking status from nonhuman primate liver histology. Adapters are evaluated against non-PEFT configurations on a subset of the Monkey Alcohol Tissue Research Resource (MATRR). All evaluated PEFT configurations achieve higher mean balanced accuracy than the non-PEFT baselines. For example, in the four-shot setting, attention-based multiple instance learning (ABMIL) alone achieves 43.5% mean balanced accuracy, compared with 78.1% after adding an adapter. Qualitative attention analysis shows that textual pathway adaptations produce localized attention heatmaps, while visual pathway adaptations produce more diffuse ones. A pathologist's review identified similar histological features across adaptations.
| Comments: | 3 pages, 2 figures. Accepted for publication in the proceedings of IEEE BIBM 2026. The code is publicly available: this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.17105 [cs.CV] |
| (or arXiv:2611.17105v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.17105 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Amirmohammad Mohammadi [view email]
[v1]
Thu, 12 Nov 2026 17:02:27 UTC (3,983 KB)
来源：arXiv cs.AI · arxiv.org