Current analyses methods are limited in features that can be determined from histology biopsies. Objectives were to develop a deep learning framework for mouse wound histology analysis and translate our approach for analysing human wound biopsies. Our data highlights the ability of our framework to accurately segment the various tissue types within mouse wound histology across 1-, 3-, 5-day timepoints of healing. We translate our framework to analyse human wound biopsies for the first time and our preliminary data provides strong support for use in clinical context. We demonstrate for the first time the analysis of all tissue types in mouse wound histology biopsies with accurate performance for the majority of tissue types and the use of deep learning for analysis of human wound biopsies.
Original languageEnglish
JournalPLoS ONE
Publication statusIn preparation - 2023


Dive into the research topics of 'Deep Learning Enables Accurate Automated Analysis of Wound Histology Biopsies'. Together they form a unique fingerprint.

Cite this