Three-dimensional conditional generative adversarial network-based virtual thin-slice technique for the morphological evaluation of the spine

Abstract

Virtual thin-slice (VTS) technique is a generative adversarial network-based algorithm that can generate virtual 1-mm-thick CT images from images of 3-10-mm thickness. We evaluated the performance of VTS technique for assessment of the spine. VTS was applied to 4-mm-thick CT images of 73 patients, and the visibility of intervertebral spaces was evaluated on the 4-mm-thick and VTS images. The heights of vertebrae measured on sagittal images reconstructed from the 4-mm-thick images and VTS images were compared with those measured on images reconstructed from 1-mm-thick images. Diagnostic performance for the detection of compression fractures was also compared. The intervertebral spaces were significantly more visible on the VTS images than on the 4-mm-thick images (P < 0.001). The absolute value of the measured difference in mean vertebral height between the VTS and 1-mm-thick images was smaller than that between the 4-mm-thick and 1-mm-thick images (P < 0.01-0.54). The diagnostic performance of the VTS images for detecting compression fracture was significantly lower than that of the 4-mm-thick images for one reader (P = 0.02). VTS technique enabled the identification of each vertebral body, and enabled accurate measurement of vertebral height. However, this technique is not suitable for diagnosing compression fractures.

Figure 1

Adversarial training framework for thick–thin slice translation of CT images.

Figure 2

Sagittal reformatted images reconstructed from 1-mm-thick images (a), 4-mm-thick images (b), and virtual thin-slice images (c). The intervertebral spaces of the cervical and upper thoracic spine are not clearly depicted on the reconstruction from 4-mm-thick images. On the reconstruction of virtual thin-slice images, the intervertebral spaces are more clearly depicted and it is easier to recognize the shapes of the vertebral bodies.

Figure 3

Sagittal reformatted images reconstructed from 1-mm-thick images (a), 4-mm-thick images (b), and virtual thin-slice images (c). A compression fracture of the 8th thoracic vertebra is seen on the reconstruction from 1-mm-thick images (arrow), but is not depicted on that from virtual thin-slice images.

Figure 4

Sagittal reformatted images reconstructed from 1-mm-thick images (a), 4-mm-thick images (b), and virtual thin-slice images (c). Multiple compression fractures are seen in the thoracic and lumbar spine. The fracture of the 10th thoracic vertebra can be seen on the reconstruction from 1-mm-thick images (arrow), and is also identifiable on that from the 4-mm-thick images. However, it is barely visible on that from the virtual thin-slice image, and is therefore difficult to diagnose.

Figure 5

JAFROC curves for Reader 1 (a) and Reader 2 (b). The figure of merit was significantly higher for thick-slice images than for virtual thin-slice images for Reader 1 (P = 0.02).