Daniel Kvak

Bone age is a radiological indicator of bone maturity that is routinely assessed in children and adolescents to evaluate growth and diagnose endocrine or chronic diseases. This retrospective study verifies the accuracy of an artificial intelligence algorithm (Carebot AI Bones, Bone Age function; Carebot s. r. o.) for automatically estimating bone maturity from dorsopalmary X-ray images. We analyzed 96 anonymized images (20–216 months; median 108) taken between January and June 2025. The reference standard was independently established by a radiologist and anthropologist according to the GP atlas, with consensus in case of disagreement. The index test was the algorithm’s prediction in months. The primary endpoint was the mean absolute error (MAE) compared to a pre-specified non-inferiority limit of 12 months. Secondary measures included RMSE, bias, Pearson’s r, Bland–Altman limits of agreement, and proportions within ±6/±12/±24 months. The algorithm showed a high correlation with the reference standard (r = 0.981; 95% CI 0.970–0.989). MAE was 5.97 months (95% CI 4.76–7.28), RMSE was 8.70, and bias was −0.27 with LoA −17.40 to +16.86. Predictions were within ±6/±12/±24 months in 66.7%/82.3%/96.9% of cases. Non-inferiority was met (t=−9.29; p <0.001). By gender, the MAE was 5.04 months for men (bias +2.79) and 6.82 months for women (bias −3.09). The lowest error was ≤60 months (MAE 3.40), with a slight underestimation occurring at 121–180 months (MAE 7.15; bias −3.47). The results show that the AI algorithm achieves an average error of less than 1 year across the entire pediatric age spectrum and meets the criteria for clinical acceptability, supporting its use as a tool to aid radiological decision-making.

A retrospective, single-center, multi-reader, multi-case study evaluated the diagnostic performance of two artificial intelligence systems (AI 1, AI 2) in detecting primary fractures on musculoskeletal X-ray images and compared it with the performance of three clinically active radiologists (RAD 1–3). A total of 1,137 anonymized images taken between March 5 and 11, 2025, were selected from the PACS archive of the Olomouc University Hospital; the prevalence of fractures was 10.4%. The reference standard was established by majority vote consensus of three independent radiologists, with substantial agreement confirmed by Fleiss‘ κ = 0.739. The power analysis required a minimum of 110 fractures and 900 negative cases; the actual sample comfortably exceeded these limits. The statistical evaluation was based on four-field contingency tables with Wilson’s 95% confidence intervals (CI); differences between readers were tested using McNemar’s test with Holm’s correction. AI 1 achieved a sensitivity of 0.941 (95% CI 0.883–0.971) and a specificity of 0.930 (0.913–0.944), AI 2 achieved a sensitivity of 0.932 (0.872–0.965) and specificity of 0.935 (0.918–0.949). The sensitivity of both algorithms did not differ statistically (p = 0.79); However, AI 2 showed higher specificity than AI 1 (p < 0.05). Radiologists achieved sensitivity of 0.915–0.941 and specificity of 0.944–0.977. Their sensitivity was comparable to both AIs, while their specificity exceeded AI 1 and was similar to or slightly higher than AI 2. The negative predictive value of all evaluated entities exceeded 0.990. Subanalyses of six anatomical regions and four age cohorts confirmed the consistently high sensitivity of both AIs, with the highest accuracy in the shoulder/clavicle region and the greatest decrease in specificity of AI 1 in the hand/wrist region. The study shows that both systems provide fracture detection with sensitivity comparable to that of the radiologists involved.

Scoliosis affects roughly 2 to 4 percent of adolescents, and treatment decisions depend on precise Cobb angle measurement. Manual assessment is time consuming and subject to inter observer variation. We conducted a retrospective, multi centre evaluation of a fully automated deep learning software (Carebot AI Bones, Spine Measurement functionality; Carebot s.r.o.) on 103 standing anteroposterior whole spine radiographs collected from ten hospitals. Two musculoskeletal radiologists independently measured each study and served as reference readers. Agreement between the AI and each radiologist was assessed with Bland Altman analysis, mean absolute error (MAE), root mean squared error (RMSE), Pearson correlation coefficient, and Cohen kappa for four grade severity classification. Against Radiologist 1 the AI achieved an MAE of 3.89 degrees (RMSE 4.77 degrees) with a bias of 0.70 degrees and limits of agreement from minus 8.59 to plus 9.99 degrees. Against Radiologist 2 the AI achieved an MAE of 3.90 degrees (RMSE 5.68 degrees) with a bias of 2.14 degrees and limits from minus 8.23 to plus 12.50 degrees. Pearson correlations were r equals 0.906 and r equals 0.880 (inter reader r equals 0.928), while Cohen kappa for severity grading reached 0.51 and 0.64 (inter reader kappa 0.59). These results demonstrate that the proposed software reproduces expert level Cobb angle measurements and categorical grading across multiple centres, suggesting its utility for streamlining scoliosis reporting and triage in clinical workflows.

This study evaluates the economic and operational impact of implementing AIdriven solutions in radiology, focusing on workflow optimization and cost reduction in the evaluation of low-risk chest X-rays. Building on previous findings that demonstrated a 62.82% reduction in reporting time with AI-generated reports compared to manual methods, this study expands the analysis to quantify the potential financial benefits. A costminimization analysis indicates that AI adoption could result in annual savings of up to 127,228 CZK per radiologist. For an average facility with 5.45 radiologists, the estimated total annual savings could reach 693,396 CZK. The study highlights AI's role in alleviating radiologists' workload by automating routine evaluations, enabling them to focus on complex diagnostic tasks. Furthermore, the financial feasibility of AI implementation was assessed across various salary levels and facility sizes, demonstrating its scalability and adaptability in radiological workflows. Future research should explore the long-term impacts on diagnostic quality, resource allocation, and healthcare system sustainability.