Citation: Baldacci, J.; Calderisi, M.;
Fiorillo, C.; Santorelli, F.M.; Rubegni,
A. Automatic Recognition of Ragged
Red Fibers in Muscle Biopsy from
Patients with Mitochondrial
Disorders. Healthcare 2022, 10, 574.
https://doi.org/10.3390/
healthcare10030574
Academic Editors: Keun Ho Ryu and
Nipon Theera-Umpon
Received: 25 January 2022
Accepted: 17 March 2022
Published: 19 March 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Automatic Recognition of Ragged Red Fibers in Muscle Biopsy
from Patients with Mitochondrial Disorders
Jacopo Baldacci
1
, Marco Calderisi
1
, Chiara Fiorillo
2
, Filippo Maria Santorelli
3,
* and Anna Rubegni
3,
*
1
Kode Solutions, 56125 Pisa, Italy; j.baldacci@kode-solutions.net (J.B.); m.calderisi@kode-solutions.net (M.C.)
2
Paediatric Neurology and Muscular Diseases Unit, University of Genoa and G. Gaslini Institute,
16147 Genova, Italy; chi.fiorillo@gmail.com
3
Molecular Medicine for Neurodegenerative and Neuromuscular Diseases Unit, IRCCS Stella Maris
Foundation, 56128 Calambrone, Italy
* Correspondence: filippo3364@gmail.com (F.M.S.); anna.rubegni@fsm.unipi.it (A.R.);
Tel.: +39-050886275 (F.M.S.); Fax: +39-050886247 (F.M.S.)
Abstract:
Mitochondrial dysfunction is considered to be a major cause of primary mitochondrial
myopathy in children and adults, as reduced mitochondrial respiration and morphological changes
such as ragged red fibers (RRFs) are observed in muscle biopsies. However, it is also possible to
hypothesize the role of mitochondrial dysfunction in aging muscle or in secondary mitochondrial
dysfunctions. The recognition of true histological patterns of mitochondrial myopathy can avoid
unnecessary genetic investigations. The aim of our study was to develop and validate machine-
learning methods for RRF detection in light microscopy images of skeletal muscle tissue. We used
image sets of 489 color images captured from representative areas of Gomori’s trichrome-stained
tissue retrieved from light microscopy images at a 20
×
magnification. We compared the performance
of random forest, gradient boosting machine, and support vector machine classifiers. Our results
suggested that the advent of scanning technologies, combined with the development of machine-
learning models for image classification, make neuromuscular disorders’ automated diagnostic
systems a concrete possibility.
Keywords:
muscle biopsy; ragged red fibers; machine learning; image recognition; computer-
aided diagnosis
1. Introduction
Mitochondrial diseases represent a group of metabolic disorders with a common link
of impaired mitochondrial function producing a chronic state of energy failure.
With a prevalence of 1/12,000 individuals, mitochondrial diseases have an extremely
variable phenotype and can present at any age [
1
]. The most involved tissues are those with
a high metabolic demand, such as the central nervous system, skeletal muscle, and heart.
Further complexity arises as a result of the dual genomic expression of mitochondrial
proteins from both nuclear and mitochondrial DNAs [2].
A multidisciplinary approach to the diagnosis of mitochondrial disease is to integrate
information from clinical, histochemical and biochemical tests in order to target molecular
genetic screening [
1
]. Recently, the traditional diagnostic approach requiring histopatho-
logical investigations in muscle biopsy and the study of the oxidative phosphorylation
(OxPhos) enzyme in muscle samples (“biopsy first”) has been replaced by massive gene
testing-adopting methodologies of next-generation sequencing which target a few hundred
genes or the whole exome (“genetics first”) [
3
–
6
]. The latter approach attempts to disentan-
gle the associated clinical phenotypes based on the genotype (“reverse phenotyping”) [
7
]
but often lacks the direct view of the consequences in disease tissues necessary to clarify
uncertain cases [8].
Healthcare 2022, 10, 574. https://doi.org/10.3390/healthcare10030574 https://www.mdpi.com/journal/healthcare
