Deep learning-based classification of the capillary ultrastructure in human skeletal muscles

TL;DR

This study uses deep learning to classify capillary structures in muscle biopsies, showing better accuracy than traditional methods in distinguishing healthy individuals from those with systemic diseases.

Contribution

A deep learning model (ResNet101) outperforms manual morphometric analysis in classifying capillary ultrastructure in human skeletal muscle biopsies.

Findings

The CNN achieved 79% diagnostic accuracy, significantly higher than manual BM thickness analysis (AUC 0.657).

The CNN's predictions were primarily based on pericyte debridement patterns.

The model's performance suggests it can generate hypotheses about capillary changes in systemic diseases.

Abstract

Capillary ultrastructure in human skeletal muscles is dynamic and prone to alterations in response to many stimuli, e.g., systemic pathologies such as diabetes mellitus and arterial hypertension. Using transmission electron microscopy (TEM) images, several studies have been conducted to quantify the capillary ultrastructure by means of morphometry. Deep learning techniques like convolutional neural networks (CNNs) are utilized to extract data-driven characteristics and to recognize patterns. Hence, the aim of this study was to train a CNN to identify morphometric patterns that differ between capillaries in muscle biopsies of healthy participants and patients with systemic pathologies for the purpose of hypothesis generation.

In this retrospective study we used 1810 electron micrographs from human skeletal muscle capillaries derived from 70 study participants which were classified as “healthy” controls or “patients“ in dependence of the absence or presence of a documented history of diabetes mellitus, arterial hypertension or peripheral arterial disease. Using these micrographs, a pre-trained open-access CNN (ResNet101) was trained to discriminate between micrographs of capillaries of the two groups. The CNN with the highest diagnostic accuracies during training were subsequently compared with manual quantitative analysis of the capillary ultrastructure to distinguish between “healthy” controls and patients.

Using classification into controls or patients as allocation reference, receiver-operating-characteristics (ROC)-analysis of manually obtained BM thickness showed the best diagnostic accuracy of all morphometric indicators (area under the ROC-curve (AUC): 0.657 ± 0.050). The best performing CNN demonstrated a diagnostic accuracy of 79% (sensitivity 93%, specificity 92%). DeLong-Test of the ROC-curves showed a significant difference (p < 0.001) between the AUC of the best performing CNN and the BM thickness. The underlying morphology responsible for the network prediction focuses mainly on debridement of pericytes.

The hypothesis-generating approach using pretrained CNN distinguishes between capillaries depicted on electron micrographs of “healthy” controls and participants with a systemic pathology more accurately than by commonly used morphometric analysis.

Figures