Cobb Angle Measurement of Spine from X-Ray Images Using Convolutional Neural Network

Abstract

Scoliosis is a common spinal condition where the spine curves to the side and thus deforms the spine. Curvature estimation provides a powerful index to evaluate the deformation severity of scoliosis. In current clinical diagnosis, the standard curvature estimation method for assessing the curvature quantitatively is done by measuring the Cobb angle, which is the angle between two lines, drawn perpendicular to the upper endplate of the uppermost vertebra involved and the lower endplate of the lowest vertebra involved. However, manual measurement of spine curvature requires considerable time and effort, along with associated problems such as interobserver and intraobserver variations. In this article, we propose an automatic system for measuring spine curvature using the anterior-posterior (AP) view spinal X-ray images. Due to the characteristic of AP view images, we first reduced the image size and then used horizontal and vertical intensity projection histograms to define the region of interest of the spine which is then cropped for sequential processing. Next, the boundaries of the spine, the central spinal curve line, and the spine foreground are detected by using intensity and gradient information of the region of interest, and a progressive thresholding approach is then employed to detect the locations of the vertebrae. In order to reduce the influences of inconsistent intensity distribution of vertebrae in the spine AP image, we applied the deep learning convolutional neural network (CNN) approaches which include the U-Net, the Dense U-Net, and Residual U-Net, to segment the vertebrae. Finally, the segmentation results of the vertebrae are reconstructed into a complete segmented spine image, and the spine curvature is calculated based on the Cobb angle criterion. In the experiments, we showed the results for spine segmentation and spine curvature; the results were then compared to manual measurements by specialists. The segmentation results of the Residual U-Net were superior to the other two convolutional neural networks. The one-way ANOVA test also demonstrated that the three measurements including the manual records of two different physicians and our proposed measured record were not significantly different in terms of spine curvature measurement. Looking forward, the proposed system can be applied in clinical diagnosis to assist doctors for a better understanding of scoliosis severity and for clinical treatments.

Automation analysis X-ray of the spine to objectify the assessment of the severity of scoliotic deformity in idiopathic scoliosis: a preliminary report

“…В качестве одной из ключевых задач была выбрана подготовка данных, и значительное внимание уделено созданию датасета рентгенограмм для обучения нейросети. Если большинство результатов исследований базировалось на анализе 14-49 рентгенограмм, то мы провели обучение сети на 300 цифровых рентгенограммах позвоночника больных идиопатическим сколиозом, предварительно расчерченных врачомрентгенологом [9][10][11][12][13][14]25]. Сопоставимый объем исследований прослежен только в работах J. Wang et al и Y. Pan et al [11,13].…”

“…Анализ литературы показал, что за последние 10 лет выполнено большое количество работ, посвященных цифровым программным решениям при вычислении величины угла Кобба на рентгенограммах, причем увеличение количества таких исследований в первую очередь обусловлено развитием цифровой рентгенографии [9][10][11][12][13]. В то же время предлагаемые решения фактически являются либо полуавтоматическими в ситуациях, когда врач определяет начальный и конечный позвонки дуг, либо полностью ручными, но с использованием программного обеспечения, позволяющего увеличить изображение и более точно провести касательные к проекции тела позвонка [14].…”

“…Выпуск 3. 2020 при условии, что с помощью детерминированных алгоритмов выделен позвоночный столб и на нем сегментированы позвонки [12]. Для применения этих алгоритмов необходима определенная структура снимков (к примеру, наличие черепа и шейного отдела позвоночника).…”

The SpinalMeter biometrical assessment to improve posture diagnosis in school-age girls: a validation study

“…However, these studies need more clarification. 19 It should be noted that previous recommendations showed that it should be a routine to screen children for scoliosis, because there is a large number of school children suffering from adolescent idiopathic scoliosis (AIS). However, the specific methods should be used because of the controversy shown in the context of school screening programs.…”

Nusinersen Administration in Spinal Muscular Atrophy Patients with Severe Scoliosis: Interlaminar Approaches at the Lumbar Level

“…Scoliosis is usually considered severe when the spinal curve measured with the Cobb method exceeds 40° to 50°. Progressed scoliosis requires surgery, mostly posterior corrective spinal fusion surgery in adolescence [6,7].…”