Image Quality Control in Lumbar Spine Radiography Using Enhanced U-Net Neural Networks

Abstract

Purpose: To standardize the radiography imaging procedure, an image quality control framework using the deep learning technique was developed to segment and evaluate lumbar spine x-ray images according to a defined quality control standard.

Materials and methods: A dataset comprising anteroposterior, lateral, and oblique position lumbar spine x-ray images from 1,389 patients was analyzed in this study. The training set consisted of digital radiography images of 1,070 patients (800, 798, and 623 images of the anteroposterior, lateral, and oblique position, respectively) and the validation set included 319 patients (200, 205, and 156 images of the anteroposterior, lateral, and oblique position, respectively). The quality control standard for lumbar spine x-ray radiography in this study was defined using textbook guidelines of as a reference. An enhanced encoder-decoder fully convolutional network with U-net as the backbone was implemented to segment the anatomical structures in the x-ray images. The segmentations were used to build an automatic assessment method to detect unqualified images. The dice similarity coefficient was used to evaluate segmentation performance.

Results: The dice similarity coefficient of the anteroposterior position images ranged from 0.82 to 0.96 (mean 0.91 ± 0.06); the dice similarity coefficient of the lateral position images ranged from 0.71 to 0.95 (mean 0.87 ± 0.10); the dice similarity coefficient of the oblique position images ranged from 0.66 to 0.93 (mean 0.80 ± 0.14). The accuracy, sensitivity, and specificity of the assessment method on the validation set were 0.971-0.990 (mean 0.98 ± 0.10), 0.714-0.933 (mean 0.86 ± 0.13), and 0.995-1.000 (mean 0.99 ± 0.12) for the three positions, respectively.

Conclusion: This deep learning-based algorithm achieves accurate segmentation of lumbar spine x-ray images. It provides a reliable and efficient method to identify the shape of the lumbar spine while automatically determining the radiographic image quality.

Keywords: U-net; deep learning; image segmentation; medical imaging; quality control; radiography.

Figure 1

A method for jugement of “Dog” sign. (A) The dog's mouth is for ipsilateral transverse process. The dog's eye is for pedicle. The dog's ear is for superior articular process. The dog's neck is for interarticularis. The dog's body is for lamina. The dog's front leg is for inferior articular process. The dog's tail is for contralateral transverse process. (B) The inferior articular processes were connected in blue line.

Figure 2

(A1,A2) Shows images of anteroposterior position. (A1) Shows qualified image. (A2) Shows unqualified image (1. Too many thoracics vertebrae; 2. Not centered and bent). (B1,B2) Shows images of lateral position. (B1) Shows qualified image. (B2) Shows unqualified image (1. Not clear; 2. Double shadow; 3. The left and right edges do not overlap). (C1,C2) Shows images of oblique position. (C1) Shows qualified image. (C2) Shows unqualified image (1. Excessive and foreign bodies in the chest; 2. Insufficient angle; 3. Less at the bottom).

Figure 3

The example of manual segmentation and AI segmentation for three positions, anteroposterior (A1,B1), lateral (A2,B2) and oblique view (A3,B3). (A1,A2,A3) Ground truth of segmentation by manual marking. (B1,B2,B3) AI segmentation results.

Figure 4

Architecture of the spatial information and channel Squeeze & Excitation “U-net.” The input of the network is the normalized image and the output is the probability map of the segmentation result. (A) SE blocks in U-net. (B) Spatial and channel SE block.

Figure 5

AI segmentation and automatic assessment by Quality Control Model. The unqualified cases (A–C). The qualified case in (D).

Figure 6

The application of Lumbar Spine X-ray radiography quality control model.