AutoDXA, a Custom-Built Dual-Energy X-ray Absorptiometry Reporting Software, Improves Reporting Efficiency and...

Dual-energy x-ray absorptiometry (DXA) technology evaluates bone mineral density and diagnoses osteoporosis. Traditional DXA reporting with manual transcription of numeric data is time consuming and can lead to errors. In this study, we evaluate the efficiency and reporting accuracy of a semi-automated DXA reporting software (AutoDXA). A single institution retrospective analysis from January 2015 to August 2025 was performed. Three methods of DXA reporting, with varying degrees of automation, were compared in terms of report turnaround time. Method 1 (1056 reports) involved only manual transcription of data; Method 2 (1027 reports) involved automated extraction of numeric data for the current study; and Method 3 (615 reports) leveraged the AutoDXA software, which enabled automated extraction of numeric data from the current and previous studies to determine the diagnostic category and 10-year fracture risk. Subgroup analyses evaluating the error rate and comparing the turnaround time when reporting baseline versus comparison studies were also performed. The mean turnaround time with the AutoDXA software (Method 3) was 2.6 × faster than Method 1 [100 vs 261 s, p < 0.0001], and 1.6 × faster than Method 2 [100 vs 155 s, p < 0.0001]. This difference was most evident in the comparison subgroup analysis, where the mean turnaround time with Method 3 was 2.9 × faster than Method 1 [108 vs 317 s, p < 0.0001]. The total error rate with Method 1 was double that with Method 2 [0.8% vs 0.4%]. There were no errors with Method 3. AutoDXA could improve the efficiency of DXA reporting, while maintaining high reporting accuracy.