Automated Opportunistic Osteoporosis Screening Using Low-Dose Chest CT among Individuals Undergoing Lung Cancer Screening in a Korean Population

Abstract

Opportunistic osteoporosis screening using deep learning (DL) analysis of low-dose chest CT (LDCT) scans is a potentially promising approach for the early diagnosis of this condition. We explored bone mineral density (BMD) profiles across all adult ages and prevalence of osteoporosis using LDCT with DL in a Korean population. This retrospective study included 1915 participants from two hospitals who underwent LDCT during general health checkups between 2018 and 2021. Trabecular volumetric BMD of L1-2 was automatically calculated using DL and categorized according to the American College of Radiology quantitative computed tomography diagnostic criteria. BMD decreased with age in both men and women. Women had a higher peak BMD in their twenties, but lower BMD than men after 50. Among adults aged 50 and older, the prevalence of osteoporosis and osteopenia was 26.3% and 42.0%, respectively. Osteoporosis prevalence was 18.0% in men and 34.9% in women, increasing with age. Compared to previous data obtained using dual-energy X-ray absorptiometry, the prevalence of osteoporosis, particularly in men, was more than double. The automated opportunistic BMD measurements using LDCT can effectively predict osteoporosis for opportunistic screening and identify high-risk patients. Patients undergoing lung cancer screening may especially profit from this procedure requiring no additional imaging or radiation exposure.

Keywords: bone mineral density; computed tomography; deep learning; opportunistic screening; osteoporosis; prevalence.

Figure 1

Flowchart of participants.

Figure 2

Overview of automatic volumetric bone mineral density (vBMD) measurement using deep learning comprising lumbar spine segmentation and region of interest (ROI) identification at L1 and L2. (a) The spine segmentation model was trained based on a U-Net architecture using only CT images paired with lumbar spine-labeled masks. We only trained the lumbar spine mask to obtain the binary mask in the lumbar range, which was used for detecting the L1 spine location to designate the vBMD ROI. (b) Using the lumbar spine segmentation-trained model, the first binary mask object of the output would be L1, which was the optimal location for measuring bone density. To define the specific ROI for vBMD, only the trabecular area of the spine should be covered, avoiding the basivertebral vein area. We used a morphological erosion algorithm for the elliptical ROI based on the segmented spine mask.

Figure 3

Age–related changes in deep learning-derived vertebral bone mineral density (DL vBMD, median and interquartile range) for men and women. Both sexes show a decline in DL vBMD with increasing age, with a more pronounced decrease in women from the age of 50–54 years.

Figure 4

Age–standardized prevalence of osteoporosis and osteopenia stratified by age group and sex. The prevalence of osteoporosis increases with age, with a notable rise in women compared to men, particularly in the older age groups.