Detection of Cervical Foraminal Stenosis from Oblique Radiograph Using Convolutional Neural Network Algorithm

Overview

Journal Yonsei Med J

Specialty General Medicine

Date 2024 Jun 24

PMID 38910301

Authors

Jihie Kim

Jae Jun Yang

Jaeha Song

SeongWoon Jo

Younghoon Kim

Jiho Park

Jin Bog Lee

Gun Woo Lee

Sehan Park

Affiliations

Soon will be listed here.

Abstract

Purpose: This study was conducted to develop a convolutional neural network (CNN) algorithm that can diagnose cervical foraminal stenosis using oblique radiographs and evaluate its accuracy.

Materials And Methods: A total of 997 patients who underwent cervical MRI and cervical oblique radiographs within a 3-month interval were included. Oblique radiographs were labeled as "foraminal stenosis" or "no foraminal stenosis" according to whether foraminal stenosis was present in the C2-T1 levels based on MRI evaluation as ground truth. The CNN model involved data augmentation, image preprocessing, and transfer learning using DenseNet161. Visualization of the location of the CNN model was performed using gradient-weight class activation mapping (Grad-CAM).

Results: The area under the curve (AUC) of the receiver operating characteristic curve based on DenseNet161 was 0.889 (95% confidence interval, 0.851-0.927). The F1 score, accuracy, precision, and recall were 88.5%, 84.6%, 88.1%, and 88.5%, respectively. The accuracy of the proposed CNN model was significantly higher than that of two orthopedic surgeons (64.0%, <0.001; 58.0%, <0.001). Grad-CAM analysis demonstrated that the CNN model most frequently focused on the foramen location for the determination of foraminal stenosis, although disc space was also frequently taken into consideration.

Conclusion: A CNN algorithm that can detect neural foraminal stenosis in cervical oblique radiographs was developed. The AUC, F1 score, and accuracy were 0.889, 88.5%, and 84.6%, respectively. With the current CNN model, cervical oblique radiography could be a more effective screening tool for neural foraminal stenosis.

References

Vaiyapuri T, Jothi A, Narayanasamy K, Kamatchi K, Kadry S, Kim J . Design of a Honey Badger Optimization Algorithm with a Deep Transfer Learning-Based Osteosarcoma Classification Model. Cancers (Basel). 2022; 14(24). PMC: 9776881. DOI: 10.3390/cancers14246066. View

Lee D, Lee H, Cho J, Hwang C, Yang J, Lee C . Foraminal Restenosis After Posterior Cervical Foraminotomy for the Treatment of Cervical Radiculopathy. Global Spine J. 2022; 13(8):2357-2366. PMC: 10538319. DOI: 10.1177/21925682221083268. View

Kang K, Lee H, Lee J . Cervical Radiculopathy Focus on Characteristics and Differential Diagnosis. Asian Spine J. 2020; 14(6):921-930. PMC: 7788378. DOI: 10.31616/asj.2020.0647. View

Kane P, Daniels A, Akelman E . Double Crush Syndrome. J Am Acad Orthop Surg. 2015; 23(9):558-62. DOI: 10.5435/JAAOS-D-14-00176. View

Lee K, Park H, Lee S, Kim M, Lee Y, Kim Y . Inter-observer reliability and clinical validity of the MRI grading system for cervical central stenosis based on sagittal T2-weighted image. Eur J Radiol. 2020; 127:108987. DOI: 10.1016/j.ejrad.2020.108987. View

Kim T, Kim Y, Park S, Lee J, Lee C, Hyun S . Diagnostic triage in patients with central lumbar spinal stenosis using a deep learning system of radiographs. J Neurosurg Spine. 2022; 37(1):104-111. DOI: 10.3171/2021.11.SPINE211136. View

Meacock J, Schramm M, Selvanathan S, Currie S, Stocken D, Jayne D . Systematic review of radiological cervical foraminal grading systems. Neuroradiology. 2021; 63(3):305-316. DOI: 10.1007/s00234-020-02596-5. View

Koo B, Lee J, Jung J, Kang C, Joo K, Kim T . A pilot study on deep learning-based grading of corners of vertebral bodies for assessment of radiographic progression in patients with ankylosing spondylitis. Ther Adv Musculoskelet Dis. 2022; 14:1759720X221114097. PMC: 9310199. DOI: 10.1177/1759720X221114097. View

Al-Rasheed A, Ksibi A, Ayadi M, Alzahrani A, Zakariah M, Ali Hakami N . An Ensemble of Transfer Learning Models for the Prediction of Skin Cancers with Conditional Generative Adversarial Networks. Diagnostics (Basel). 2022; 12(12). PMC: 9777332. DOI: 10.3390/diagnostics12123145. View

10.

Ogawa T, Yoshii T, Oyama J, Sugimura N, Akada T, Sugino T . Detecting ossification of the posterior longitudinal ligament on plain radiographs using a deep convolutional neural network: a pilot study. Spine J. 2022; 22(6):934-940. DOI: 10.1016/j.spinee.2022.01.004. View

11.

Kim G, Park W, Chang M, Lee G . Diagnostic Technology for Spine Pathology. Asian Spine J. 2022; 16(5):764-775. PMC: 9633243. DOI: 10.31616/asj.2022.0374. View

12.

Mandrekar J . Receiver operating characteristic curve in diagnostic test assessment. J Thorac Oncol. 2010; 5(9):1315-6. DOI: 10.1097/JTO.0b013e3181ec173d. View

13.

Yamada Y, Maki S, Kishida S, Nagai H, Arima J, Yamakawa N . Automated classification of hip fractures using deep convolutional neural networks with orthopedic surgeon-level accuracy: ensemble decision-making with antero-posterior and lateral radiographs. Acta Orthop. 2020; 91(6):699-704. PMC: 8023868. DOI: 10.1080/17453674.2020.1803664. View

14.

Maduri R, Cossu G, Aureli V, Plaza Wuthrich S, Bobinski L, Duff J . Transtubular Anterior Cervical Foraminotomy for the Treatment of Compressive Cervical Radiculopathy: Surgical Results and Complications in a Consecutive Series of Cases. Asian Spine J. 2020; 15(5):673-681. PMC: 8561162. DOI: 10.31616/asj.2020.0323. View

15.

Miura M, Maki S, Miura K, Takahashi H, Miyagi M, Inoue G . Automated detection of cervical ossification of the posterior longitudinal ligament in plain lateral radiographs of the cervical spine using a convolutional neural network. Sci Rep. 2021; 11(1):12702. PMC: 8208978. DOI: 10.1038/s41598-021-92160-9. View

16.

Subramani B, Veluchamy M . Fuzzy Gray Level Difference Histogram Equalization for Medical Image Enhancement. J Med Syst. 2020; 44(6):103. DOI: 10.1007/s10916-020-01568-9. View

17.

Meng M, Zhang M, Shen D, He G, Guo Y . Detection and classification of breast lesions with You Only Look Once version 5. Future Oncol. 2022; 18(39):4361-4370. DOI: 10.2217/fon-2022-0593. View

18.

Mehdizadeh M, Tafti K, Soltani P . Evaluation of histogram equalization and contrast limited adaptive histogram equalization effect on image quality and fractal dimensions of digital periapical radiographs. Oral Radiol. 2022; 39(2):418-424. DOI: 10.1007/s11282-022-00654-7. View

19.

Lee J, Park H, Lee S, Lee Y, Kim Y, Lee K . Interreader Reliability and Clinical Validity of a Magnetic Resonance Imaging Grading System for Cervical Foraminal Stenosis. J Comput Assist Tomogr. 2017; 41(6):926-930. DOI: 10.1097/RCT.0000000000000628. View

20.

Park S, Kim J, Chang M, Park J, Yang J, Lee G . Assessment of Fusion After Anterior Cervical Discectomy and Fusion Using Convolutional Neural Network Algorithm. Spine (Phila Pa 1976). 2022; 47(23):1645-1650. DOI: 10.1097/BRS.0000000000004439. View