Differentiating Spinal Pathologies by Deep Learning Approach.

Deep Learning Diagnosis MRI Machine Learning Spine

Journal

The spine journal : official journal of the North American Spine Society
ISSN: 1878-1632
Titre abrégé: Spine J
Pays: United States
ID NLM: 101130732

Informations de publication

Date de publication:
03 Oct 2023
Historique:
received: 03 07 2023
revised: 19 09 2023
accepted: 26 09 2023
medline: 6 10 2023
pubmed: 6 10 2023
entrez: 5 10 2023
Statut: aheadofprint

Résumé

Spinal pathologies are diverse in nature and, excluding trauma and degenerative diseases, includes infectious, neoplastic (either extradural or intradural) and inflammatory conditions. The preoperative diagnosis is made with clinical judgment incorporating lab findings and radiological studies. When the diagnosis is uncertain, a biopsy is almost always mandatory since the treatment is dictated by the type of pathology. This is an invasive, timely and costly process. The aim of this study was to develop a deep learning (DL) algorithm, based on preoperative MRI and post-operative pathological results, to differentiate between leading spinal pathologies. We retrospectively collected and analyzed clinical, radiological, and pathological data of patients who underwent spinal surgery or biopsy for various spinal pathologies between 2008-2022 at a tertiary center. The patients were stratified according to their pathological reports (the threshold for inclusion was set to 25 patients per diagnosis). Preoperative MRI, clinical data and pathological results were processed by a deep learning model built on the Fast.ai framework on top of the PyTorch environment. Two-hundred and thirty-one patients diagnosed with carcinoma (80), infection (57), meningioma (52) or schwannoma (42), were included in our model. The mean overall accuracy was 0.78±0.06 for the validation, and 0.93±0.03 for the test dataset. DL algorithm for differentiation between the aforementioned spinal pathologies, based solely on clinical MRI, proves as a feasible primary diagnostic modality. Larger studies should be performed to validate and improve this algorithm for clinical use. This study provides a proof-of-concept for predicting spinal pathologies solely by MRI based DL technology, allowing for a rapid, targeted and cost-effective work-up and subsequent treatment.

Sections du résumé

BACKGROUND CONTEXT BACKGROUND
Spinal pathologies are diverse in nature and, excluding trauma and degenerative diseases, includes infectious, neoplastic (either extradural or intradural) and inflammatory conditions. The preoperative diagnosis is made with clinical judgment incorporating lab findings and radiological studies. When the diagnosis is uncertain, a biopsy is almost always mandatory since the treatment is dictated by the type of pathology. This is an invasive, timely and costly process.
PURPOSE OBJECTIVE
The aim of this study was to develop a deep learning (DL) algorithm, based on preoperative MRI and post-operative pathological results, to differentiate between leading spinal pathologies.
STUDY DESIGN METHODS
We retrospectively collected and analyzed clinical, radiological, and pathological data of patients who underwent spinal surgery or biopsy for various spinal pathologies between 2008-2022 at a tertiary center. The patients were stratified according to their pathological reports (the threshold for inclusion was set to 25 patients per diagnosis).
METHODS METHODS
Preoperative MRI, clinical data and pathological results were processed by a deep learning model built on the Fast.ai framework on top of the PyTorch environment.
RESULTS RESULTS
Two-hundred and thirty-one patients diagnosed with carcinoma (80), infection (57), meningioma (52) or schwannoma (42), were included in our model. The mean overall accuracy was 0.78±0.06 for the validation, and 0.93±0.03 for the test dataset.
CONCLUSION CONCLUSIONS
DL algorithm for differentiation between the aforementioned spinal pathologies, based solely on clinical MRI, proves as a feasible primary diagnostic modality. Larger studies should be performed to validate and improve this algorithm for clinical use.
CLINICAL SIGNIFICANCE CONCLUSIONS
This study provides a proof-of-concept for predicting spinal pathologies solely by MRI based DL technology, allowing for a rapid, targeted and cost-effective work-up and subsequent treatment.

Identifiants

pubmed: 37797840
pii: S1529-9430(23)03425-3
doi: 10.1016/j.spinee.2023.09.019
pii:
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Informations de copyright

Copyright © 2023. Published by Elsevier Inc.

Déclaration de conflit d'intérêts

Declaration of Competing Interest The authors declare no competing interests.

Auteurs

Oz Haim (O)

Department of Neurosurgery, Tel Aviv Medical Center, Tel Aviv, Israel; Department of Neurosurgery, Baylor College of Medicine, Houston, Texas, USA.

Ariel Agur (A)

Department of Neurosurgery, Tel Aviv Medical Center, Tel Aviv, Israel.

Segev Gabay (S)

Department of Neurosurgery, Tel Aviv Medical Center, Tel Aviv, Israel.

Lee Azolai (L)

Department of Neurosurgery, Tel Aviv Medical Center, Tel Aviv, Israel.

Itay Shutan (I)

Department of Neurosurgery, Tel Aviv Medical Center, Tel Aviv, Israel.

May Chitayat (M)

The Iby and Aladar Fleischman Faculty of Engineering, Tel Aviv University, Tel Aviv 6997801, Israel; Sagol Brain Institute, Tel Aviv Medical Center, and the Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Tel-Aviv, Israel.

Michal Katirai (M)

The Iby and Aladar Fleischman Faculty of Engineering, Tel Aviv University, Tel Aviv 6997801, Israel; Sagol Brain Institute, Tel Aviv Medical Center, and the Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Tel-Aviv, Israel.

Sapir Sadon (S)

Department of Cardiology, Tel Aviv Medical Center, Tel Aviv, Israel.

Moran Artzi (M)

Sagol Brain Institute, Tel Aviv Medical Center, and the Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Tel-Aviv, Israel. Electronic address: artzimy@gmail.com.

Zvi Lidar (Z)

Department of Neurosurgery, Tel Aviv Medical Center, Tel Aviv, Israel.

Classifications MeSH