Identification of cervical spine fracture using deep learning

ABSTRACT

The first seven vertebrae of the spine are referred to as the cervical spine. X-rays, Computed Tomography scans, Magnetic Resonance Imaging, and physical examinations methods are used to identify spine fractures which are complex and non-interpretable. To overcome these drawbacks, automated systems are required. In this paper, the fifth version of You Only Look Once (YOLO) and deep neural network have combined to detect dislocations in vertebral columns. YOLO v5 is used to detect major and minute fractures of C1 to C7 vertebrae and deep neural network is used to classify normal and fractured vertebrate. Training dataset provided by ASNR and ASSR spine radiology specialists is used to train model. Experimental results show that the proposed model offers classification accuracy of 97.7% and 89% on training and validation respectively. The proposed system is able to show exact location of fracture in Cervical Spine and perform better than existing systems.

KEYWORDS:

• Cervical spine: fracture
• vertebral column
• X-ray
• tensorFlow
• deep learning

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

D P Gaikwad

D P Gaikwad has completed his BE (Computer Science and Engineering) from SGGS College of Engineering, Nanded and M. Tech. (Computer Engineering) in 2006 from College of Engineering, Pune. He has complete his Ph. D in Computer Science and Engineering from SGGSIOET, Nanded He has published near about 70 paper in International journals and conferences. He is a reviewer of many international journals and conferences.

Ahire Sejal

Sejal Ahire has completed her BE (Computer Engineering) in 2022 from AISSMS College of Engineering, SPPU, Pune.

Swarupa Bagade

Swarupa Bagade has completed his BE (Computer Engineering) in 2022 from AISSMS College of Engineering, SPPU Pune.

Netra Ghodekar

Netra Ghodekar has completed her BE (Computer Engineering) in 2022 from AISSMS College of Engineering, SPPU, Pune.

Srushti Labade

Srushti Labade has completed her BE (Computer Engineering) in 2022 from AISSMS College of Engineering, SPPU, Pune.