Robust volumetric multiplex staining of centimeter-scale FFPE tissues guided by neural network-based optimization
Lin, Y.-H.; Huang, C.-Y.; Chen, Y.-H.; Chen, Y.-H.; Xu, Z.-W.; Ko, P.-L.; Hsu, H.-H.; Tung, Y.-C.; Chen, Y.-F.; Chen, H.-C.; Chiang, A.-S.; Fiock, K. L.; Wang, K.-C.; Lin, C.-H.; Hu, S.-H.; Chen, B.-C.; Chu, L.-A.
Show abstract
Neurodegenerative diseases involve structural and morphological alterations in tissue architecture that are difficult to capture in single thin sections. Three-dimensional multiplexed pathology, however, remains limited by the lack of clearing methods applicable to formalin-fixed paraffin-embedded (FFPE) clinical specimens. As the development of tissue-clearing methods requires the optimization of multiple parameters, we employed a neural network-based Complex System Response (CSR) approach to guide the design of FIDELITY, an epoxy-free delipidation and epitope-retrieval pipeline for whole FFPE specimens. FIDELITY preserves tissue rigidity, enhances immunostaining efficiency, and supports at least five rounds of multiplex labeling without deformation. It enables whole-brain atlas registration, quantitative neuronal profiling, and volumetric pathology of archived human Alzheimers and glioma specimens while remaining compatible with routine histology. Altogether, FIDELITY provides accurate 3D metrics and offers translational potential to bridge spatial mapping and conventional pathology.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.