Authors
Tianyue He, Wenyi Jing, Le Zhang, Qican Zhang, Tingdong Kou, Xin He, Yan Qiu, Yang Lu, Jian Cui, Yongfu Wen, Zhenrong Zheng, Hongying Zhang, Dayong Jin, Junfei Shen
Published in
Advanced science (Weinheim, Baden-Wurttemberg, Germany). Pages e76292. Jun 30, 2026. Epub Jun 30, 2026.
Abstract
Virtual staining will enable fast and reagent-free histological imaging toward accurate and consistent pathology diagnosis. However, most current developments rely on heuristic optical landmarks and blind reconstruction, which reduces interpretability, limits cross-task adaptations, and prevents their ultimate clinical translations. Here, we introduce an interpretable Plain-to-Stain framework, "Task-Adaptive Programmable Optics (TAPO)" that removes manual landmark selection by directly tying programmable optical control to task-specific diagnostic objectives. The approach operates under specially modulated visible light, and the highly programmable imaging system decomposes each optical step to execute optics-attention-based reconstruction, leading to task-adaptive multimodal optical encoding for virtual staining. We demonstrated that TAPO successfully passed rigorous validation across diverse tissue types (liver, lung), pathological states (normal, malignant), staining protocols (H&E, Masson's trichrome), magnifications (10×, 20×), and section-thickness settings including routine 4 µm sections and moderately thicker 10 µm sections. Blinded assessments by a panel of board-certified pathologists on multiple virtual slides showed that TAPO results had high clinical acceptance, especially in resolution, cytoplasmic clarity, and nuclear staining, and demonstrated considerable potential for clinical pathology practice.
PMID:
42378647
Bibliographic data and abstract were imported from PubMed on 01 Jul 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 10
- Comments 0