Detection of Critical Spinal Epidural Lesions on CT Using Machine Learning.
Journal
Spine
ISSN: 1528-1159
Titre abrégé: Spine (Phila Pa 1976)
Pays: United States
ID NLM: 7610646
Informations de publication
Date de publication:
01 Jan 2023
01 Jan 2023
Historique:
received:
27
05
2022
accepted:
12
07
2022
pubmed:
30
7
2022
medline:
15
12
2022
entrez:
29
7
2022
Statut:
ppublish
Résumé
Critical spinal epidural pathologies can cause paralysis or death if untreated. Although magnetic resonance imaging is the preferred modality for visualizing these pathologies, computed tomography (CT) occurs far more commonly than magnetic resonance imaging in the clinical setting. A machine learning model was developed to screen for critical epidural lesions on CT images at a large-scale teleradiology practice. This model has utility for both worklist prioritization of emergent studies and identifying missed findings. There were 153 studies with epidural lesions available for training. These lesions were segmented and used to train a machine learning model. A test data set was also created using previously missed epidural lesions. The trained model was then integrated into a teleradiology workflow for 90 days. Studies were sent to secondary manual review if the model detected an epidural lesion but none was mentioned in the clinical report. The model correctly identified 50.0% of epidural lesions in the test data set with 99.0% specificity. For prospective data, the model correctly prioritized 66.7% of the 18 epidural lesions diagnosed on the initial read with 98.9% specificity. There were 2.0 studies flagged for potential missed findings per day, and 17 missed epidural lesions were found during a 90-day time period. These results suggest almost half of critical spinal epidural lesions visible on CT imaging are being missed on initial diagnosis. A machine learning model for identifying spinal epidural hematomas and abscesses on CT can be implemented in a clinical workflow.
Sections du résumé
BACKGROUND
BACKGROUND
Critical spinal epidural pathologies can cause paralysis or death if untreated. Although magnetic resonance imaging is the preferred modality for visualizing these pathologies, computed tomography (CT) occurs far more commonly than magnetic resonance imaging in the clinical setting.
OBJECTIVE
OBJECTIVE
A machine learning model was developed to screen for critical epidural lesions on CT images at a large-scale teleradiology practice. This model has utility for both worklist prioritization of emergent studies and identifying missed findings.
MATERIALS AND METHODS
METHODS
There were 153 studies with epidural lesions available for training. These lesions were segmented and used to train a machine learning model. A test data set was also created using previously missed epidural lesions. The trained model was then integrated into a teleradiology workflow for 90 days. Studies were sent to secondary manual review if the model detected an epidural lesion but none was mentioned in the clinical report.
RESULTS
RESULTS
The model correctly identified 50.0% of epidural lesions in the test data set with 99.0% specificity. For prospective data, the model correctly prioritized 66.7% of the 18 epidural lesions diagnosed on the initial read with 98.9% specificity. There were 2.0 studies flagged for potential missed findings per day, and 17 missed epidural lesions were found during a 90-day time period. These results suggest almost half of critical spinal epidural lesions visible on CT imaging are being missed on initial diagnosis.
CONCLUSION
CONCLUSIONS
A machine learning model for identifying spinal epidural hematomas and abscesses on CT can be implemented in a clinical workflow.
Identifiants
pubmed: 35905328
doi: 10.1097/BRS.0000000000004438
pii: 00007632-202301010-00001
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
1-7Informations de copyright
Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.
Déclaration de conflit d'intérêts
The authors report no conflicts of interest.
Références
Al-Mutair A, Bednar DA. Spinal epidural hematoma. J Am Acad Orthop Surg. 2010;18:494–502.
Schwab JH, Shah AA. Spinal epidural abscess: diagnosis, management, and outcomes. J Am Acad Orthop Surg. 2020;28:929–38.
Pierce JL, Donahue JH, Nacey NC, Quirk CR, Perry MT, Faulconer N, et al. Spinal hematomas: what a radiologist needs to know. Radiographics. 2018;38:1516–35.
Ortiz AO, Levitt A, Shah LM, Parsons MS, Agarwal V, Baldwin K, et al. ACR Appropriateness Criteria® Suspected Spine Infection. J Am Coll Radiol. 2021;18:488–501.
Harris RJ, Kim S, Lohr J, Towey S, Velichkovich Z, Kabachenko T, et al. Classification of aortic dissection and rupture on post-contrast CT images using a Convolutional Neural Network. J Digit Imaging. 2019;32:939–46.
Harris RJ, Baginski SG, Bronstein Y, Kim S, Lohr J, Towey S, et al. Measurement of endotracheal tube positioning on chest X-ray using object detection. J Digit Imaging. 2021;34:846–52.
Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: machine learning in Python. J Mach Learn Res. 2011;12:2825–30.
Redmon J, Farhadi A YOLO v.3: An Incremental Improvement. ArXiv abs/1804.02767. 2018.
Kingma DP, Ba JL Adam: a method for stochastic optimization. In: 3rd International Conference on Learning Representations, ICLR 2015—Conference Track Proceedings. 2015.
Small JE, Osler P, Paul AB, Kunst M. Ct cervical spine fracture detection using a convolutional neural network. Am J Neuroradiol. 2021;42:1341–47.
Tomita N, Cheung YY, Hassanpour S. Deep neural networks for automatic detection of osteoporotic vertebral fractures on CT scans. Comput Biol Med. 2018;98:8–15.
Ong T, Copeland R, Thiam CN, Cerda Mas G, Marshall L, Sahota O. Integration of a vertebral fracture identification service into a fracture liaison service: a quality improvement project. Osteoporos Int. 2021;32:921–26.
Chang PD, Kuoy E, Grinband J, Weinberg BD, Thompson M, Homo R, et al. Hybrid 3D/2D convolutional neural network for hemorrhage evaluation on head CT. Am J Neuroradiol. 2018;39:1609–1616.
Ye H, Gao F, Yin Y, Guo D, Zhao P, Lu Y, et al. Precise diagnosis of intracranial hemorrhage and subtypes using a three-dimensional joint convolutional and recurrent neural network. Eur Radiol. 2019;29:6191–201.
Kuo W, Häne C, Mukherjee P, Malik J, Yuh EL. Expert-level detection of acute intracranial hemorrhage on head computed tomography using deep learning. Proc Natl Acad Sci USA. 2019;116:22735–22737.
Holtås S, Heiling M, Lönntoft M. Spontaneous spinal epidural hematoma: findings at MR imaging and clinical correlation. Radiology. 1996;199:409–13.
Amiri AR, Fouyas IP, Cro S, Casey ATH. Postoperative spinal epidural hematoma (SEH): incidence, risk factors, onset, and management. Spine J. 2013;13:134–40.
Vakili M, Crum-Cianflone NF. Spinal epidural abscess: a series of 101 cases. Am J Med. 2017;130:1458–63.
Lamoureux C, Hanna TN, Sprecher D, Weber S, Callaway E. Radiologist errors by modality, anatomic region, and pathology for 1.6 million exams: what we have learned. Emerg Radiol. 2021;28:1135–41.
Hanna TN, Lamoureux C, Krupinski EA, Weber S, Johnson JO. Effect of shift, schedule, and volume on interpretive accuracy: a retrospective analysis of 2.9 million radiologic examinations. Radiology. 2018;287:205–12.
Goldberg-Stein S, Frigini LA, Long S, Metwalli Z, Nguyen XV, Parker M, et al. ACR RADPEER Committee White Paper with 2016 Updates: Revised Scoring System, New Classifications, Self-Review, and Subspecialized Reports. J Am Coll Radiol. 2017;14:1080–6.