Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform

Abstract Background The histopathological analysis of minor salivary gland biopsies, particularly through the quantification of the Focus Score (FS), is pivotal in the diagnostic workflow for Sjögren's Syndrome (SS). AI-based image recognition using deep learning models has demonstrated potenti...

Full description

Saved in:
Bibliographic Details
Main Authors: Jorge Álvarez Troncoso, Elena Ruiz-Bravo, Clara Soto Abánades, Alexandre Dumusc, Álvaro López-Janeiro, Thomas Hügle
Format: Article
Language:English
Published: BMC 2024-11-01
Series:BMC Rheumatology
Subjects:
Online Access:https://doi.org/10.1186/s41927-024-00417-3
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850196137173057536
author Jorge Álvarez Troncoso
Elena Ruiz-Bravo
Clara Soto Abánades
Alexandre Dumusc
Álvaro López-Janeiro
Thomas Hügle
author_facet Jorge Álvarez Troncoso
Elena Ruiz-Bravo
Clara Soto Abánades
Alexandre Dumusc
Álvaro López-Janeiro
Thomas Hügle
author_sort Jorge Álvarez Troncoso
collection DOAJ
description Abstract Background The histopathological analysis of minor salivary gland biopsies, particularly through the quantification of the Focus Score (FS), is pivotal in the diagnostic workflow for Sjögren's Syndrome (SS). AI-based image recognition using deep learning models has demonstrated potential in enhancing diagnostic accuracy and efficiency in preclinical research. Objectives The primary aim of this investigation was to utilize an auto-machine learning (autoML) platform for the automated segmentation and quantification of FS on histopathological slides, aiming to augment diagnostic precision and speed in SS. Methods A cohort comprising 86 patients with sicca syndrome (37 diagnosed with SS based on the 2016 ACR/EULAR Classification Criteria and 49 non-SS) was selected for an in-depth histological examination. A repository of 172 slides (two per patient) was assembled, encompassing 74 slides meeting the classificatory thresholds for SS (FS ≥ 1, indicative of lymphocytic infiltration) and 98 slides showcasing normal salivary gland histology. The autoML platform utilized (Giotto, L2F, Lausanne Switzerland) employed a Convolutional Neural Network (CNN) architecture (ResNet-152) for the training and validation phases, using a dataset of 172 slides. Results The developed model exhibited a reliability score of 0.88, proficiently distinguishing SS cases, with a sensitivity of 89.47% (95% CI: 66.86% to 98.70%) and a specificity of 88.24% (95% CI: 63.56% to 98.54%). The model found histological slides of suboptimal quality (e.g., those compromised during fixation or staining processes) to be the most challenging for accurate classification. Conclusion AutoML platforms offer a rapid and flexible approach to developing machine learning models, even with smaller datasets, as demonstrated in this study for SS. These platforms hold significant potential for enhancing diagnostic precision and efficiency in both clinical and research settings. Multicentric studies with larger patient cohorts are essential for thorough evaluation and validation of this innovative diagnostic approach.
format Article
id doaj-art-bbb4bfa22e964a7fb2c007ecc05fc6a8
institution OA Journals
issn 2520-1026
language English
publishDate 2024-11-01
publisher BMC
record_format Article
series BMC Rheumatology
spelling doaj-art-bbb4bfa22e964a7fb2c007ecc05fc6a82025-08-20T02:13:32ZengBMCBMC Rheumatology2520-10262024-11-01811810.1186/s41927-024-00417-3Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platformJorge Álvarez Troncoso0Elena Ruiz-Bravo1Clara Soto Abánades2Alexandre Dumusc3Álvaro López-Janeiro4Thomas Hügle5Systemic Autoimmune Diseases Unit, Hospital Universitario La PazPathology Department, Hospital Universitario La PazSystemic Autoimmune Diseases Unit, Hospital Universitario La PazDepartment of Rheumatology, Lausanne University Hospital (CHUV)Pathology Department, Clínica Universidad de NavarraDepartment of Rheumatology, Lausanne University Hospital (CHUV)Abstract Background The histopathological analysis of minor salivary gland biopsies, particularly through the quantification of the Focus Score (FS), is pivotal in the diagnostic workflow for Sjögren's Syndrome (SS). AI-based image recognition using deep learning models has demonstrated potential in enhancing diagnostic accuracy and efficiency in preclinical research. Objectives The primary aim of this investigation was to utilize an auto-machine learning (autoML) platform for the automated segmentation and quantification of FS on histopathological slides, aiming to augment diagnostic precision and speed in SS. Methods A cohort comprising 86 patients with sicca syndrome (37 diagnosed with SS based on the 2016 ACR/EULAR Classification Criteria and 49 non-SS) was selected for an in-depth histological examination. A repository of 172 slides (two per patient) was assembled, encompassing 74 slides meeting the classificatory thresholds for SS (FS ≥ 1, indicative of lymphocytic infiltration) and 98 slides showcasing normal salivary gland histology. The autoML platform utilized (Giotto, L2F, Lausanne Switzerland) employed a Convolutional Neural Network (CNN) architecture (ResNet-152) for the training and validation phases, using a dataset of 172 slides. Results The developed model exhibited a reliability score of 0.88, proficiently distinguishing SS cases, with a sensitivity of 89.47% (95% CI: 66.86% to 98.70%) and a specificity of 88.24% (95% CI: 63.56% to 98.54%). The model found histological slides of suboptimal quality (e.g., those compromised during fixation or staining processes) to be the most challenging for accurate classification. Conclusion AutoML platforms offer a rapid and flexible approach to developing machine learning models, even with smaller datasets, as demonstrated in this study for SS. These platforms hold significant potential for enhancing diagnostic precision and efficiency in both clinical and research settings. Multicentric studies with larger patient cohorts are essential for thorough evaluation and validation of this innovative diagnostic approach.https://doi.org/10.1186/s41927-024-00417-3Artificial intelligenceSalivary gland biopsySjögrenSicca
spellingShingle Jorge Álvarez Troncoso
Elena Ruiz-Bravo
Clara Soto Abánades
Alexandre Dumusc
Álvaro López-Janeiro
Thomas Hügle
Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform
BMC Rheumatology
Artificial intelligence
Salivary gland biopsy
Sjögren
Sicca
title Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform
title_full Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform
title_fullStr Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform
title_full_unstemmed Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform
title_short Classification of salivary gland biopsies in Sjögren’s syndrome by a convolutional neural network using an auto-machine learning platform
title_sort classification of salivary gland biopsies in sjogren s syndrome by a convolutional neural network using an auto machine learning platform
topic Artificial intelligence
Salivary gland biopsy
Sjögren
Sicca
url https://doi.org/10.1186/s41927-024-00417-3
work_keys_str_mv AT jorgealvareztroncoso classificationofsalivaryglandbiopsiesinsjogrenssyndromebyaconvolutionalneuralnetworkusinganautomachinelearningplatform
AT elenaruizbravo classificationofsalivaryglandbiopsiesinsjogrenssyndromebyaconvolutionalneuralnetworkusinganautomachinelearningplatform
AT clarasotoabanades classificationofsalivaryglandbiopsiesinsjogrenssyndromebyaconvolutionalneuralnetworkusinganautomachinelearningplatform
AT alexandredumusc classificationofsalivaryglandbiopsiesinsjogrenssyndromebyaconvolutionalneuralnetworkusinganautomachinelearningplatform
AT alvarolopezjaneiro classificationofsalivaryglandbiopsiesinsjogrenssyndromebyaconvolutionalneuralnetworkusinganautomachinelearningplatform
AT thomashugle classificationofsalivaryglandbiopsiesinsjogrenssyndromebyaconvolutionalneuralnetworkusinganautomachinelearningplatform