Classification of endplate lesions in the lumbar spine and association with risk factors, biochemistry, and genetics

Automatic classification of the vertebral endplate lesions in magnetic resonance imaging by deep learning model

Introduction

A novel classification scheme for endplate lesions, based on T2-weighted images from magnetic resonance imaging (MRI) scan, has been recently introduced and validated. The scheme categorizes intervertebral spaces as "normal," "wavy/irregular," "notched," and "Schmorl's node." These lesions have been associated with spinal pathologies, including disc degeneration and low back pain. The exploitation of an automatic tool for the detection of the lesions would facilitate clinical practice by reducing the workload and the diagnosis time. The present work exploits a deep learning application based on convolutional neural networks to automatically classify the type of lesion.

Methods

T2-weighted MRI scans of the sagittal lumbosacral spine of consecutive patients were retrospectively collected. The middle slice of each scan was manually processed to identify the intervertebral spaces from L1L2 to L5S1, and the corresponding lesion type was labeled. A total of 1,559 gradable discs were obtained, with the following types of distribution: "normal" (567 discs), "wavy/irregular" (485), "notched" (362), and "Schmorl's node" (145). The dataset was divided randomly into a training set and a validation set while preserving the original distribution of lesion types in each set. A pretrained network for image classification was utilized, and fine-tuning was performed using the training set. The retrained net was then applied to the validation set to evaluate the overall accuracy and accuracy for each specific lesion type.

Results

The overall rate of accuracy was found equal to 88%. The accuracy for the specific lesion type was found as follows: 91% (normal), 82% (wavy/irregular), 93% (notched), and 83% (Schmorl's node).

Discussion

The results indicate that the deep learning approach achieved high accuracy for both overall classification and individual lesion types. In clinical applications, this implementation could be employed as part of an automatic detection tool for pathological conditions characterized by the presence of endplate lesions, such as spinal osteochondrosis.

The lumbar spinal endplate lesions grades and association with lumbar disc disorders, and lumbar bone mineral density in a middle-young general Chinese population

BACKGROUND

Lumbar vertebral endplates lesions (LEPLs), one of the etiologies of low back pain (LBP), are one of the most prevalent causes of health-care costs. Despite progressively becoming the focus in recent years, almost all studies have concentrated on symptomatic patients rather than general populations. As a result, our study was designed to determine the prevalence and distribution patterns of LEPLs in a middle-young general population, as well as their associations with lumbar disc herniation (LDH), lumbar disc degeneration (LDD), and lumbar vertebral volumetric bone mineral density (vBMD).

METHODS

Seven hundred fifty-four participants aged 20-60 years were recruited from the subjects enrolled in a 10-year longitudinal study of degeneration of the spine and knee being conducted at the Beijing Jishuitan Hospital and 4 of them were excluded due to the missing of MRIs. In this observational study, a lumbar quantitative computed tomography (QCT) and MRI scan were performed among participants within 48 h. T2-weighted sagittal lumbar MRI images for all included subjects were identified for LEPLs by two independent observers based on morphological and local characteristics. Lumbar vertebral vBMD was measured with QCT. The age, BMI, waistline, hipline, lumbar vBMD, LDD, and LDH were measured to investigate their associations with LEPLs.

RESULTS

The prevalence of LEPLs was higher among the male subjects. 80% of endplates were recognition as no lesions with a substantial disparity between female (75.6%) and male subjects (83.4%) (p < 0.001). The most common lesions were "wavy/irregular" and "notched", and "fracture" is most involved in L3-4 inferior endplate both in two genders. LEPLs were found to be associated with LDH (≥ 2 levels: OR = 6.859, P < 0.001; 1 level: OR = 2.328, P = 0.002 in men. OR = 5.004, P < 0.001; OR = 1.805, P = 0.014 in women) reference for non-LDH, and hipline in men (OR = 1.123, P < 0.001).

CONCLUSIONS

LEPLs are the common findings on lumbar MRIs in general population, particularly in men. The presence of these lesions and advance from slightly to severely could be mainly attributed to LDH and men's higher hipline.