The DL-based algorithm was capable of accurately labeling vertebrae and screening opportunistic VCF on real-world data. The device shows promising results in the automatic detection of VCF on scans acquired for other medical reasons. Indeed, the results matched with the expert visual assessments, considered as the GT. 

Such automatic diagnosis has many benefits. For instance, radiologists no longer need to perform the tedious task of screening for incidental findings. Saved time can be used to improve early diagnosis of osteoporosis, initiate treatment and predict future fragility fractures [10,11]. Furthermore, these automatic diagnostic tools can address the lack of access to expert radiologists in rural, small, or poor communities[11]. As a result, a successful osteoporotic VCF detection system could potentially decrease the socio-economic burden of osteoporosis[10].

This robust DL tool may impact the early analysis of musculoskeletal disease, reducing the rate of undiagnosed scans and optimizing the vertebral labeling. Most importantly, this reliable and accurate automated software may accelerate the diagnostic workflow and assist physicians in the early diagnosis without exposing the patient to additional examination and radiation.

Future prospective studies are needed to confirm the generalized use of the software in larger cohorts. Furthermore, evaluating the algorithm's direct clinical influence on patient outcomes could offer a deeper understanding of the benefits associated with this software.