AI Assistance Reduces Missed Fractures on Musculoskeletal
Radiographs: A Multicentre Multi-Reader Crossover Study

J 2026

AI Assistance Reduces Missed Fractures on Musculoskeletal Radiographs: A Multicentre Multi-Reader Crossover Study

KLIČNÍK, Simon; Karel NĚMEC; Marika DAVIDOVÁ; Samuel BILANIN; Veronika JIRKŮ et al.

Základní údaje

Originální název

AI Assistance Reduces Missed Fractures on Musculoskeletal Radiographs: A Multicentre Multi-Reader Crossover Study

Autoři

KLIČNÍK, Simon; Karel NĚMEC; Marika DAVIDOVÁ; Samuel BILANIN; Veronika JIRKŮ; Šimon KUBOV; Lukáš KORBEL; David NOHEJL; Petra OVESNÁ ; Karolína KVAKOVÁ; Jakub DANDÁR a Daniel KVAK

Vydání

European Journal of Radiology Artificial Intelligence, Amsterdam, Netherlands, Elsevier, 2026, 3050-5771

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

30200 3.2 Clinical medicine

Stát vydavatele

Nizozemské království

Utajení

není předmětem státního či obchodního tajemství

Označené pro přenos do RIV

Organizační jednotka

Lékařská fakulta

DOI

https://doi.org/10.1016/j.ejrai.2026.100106

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 13. 7. 2026 14:09, Mgr. Daniel Kvak

Anotace

V originále

Purpose: To quantify the impact of AI software on fracture detection on musculoskeletal radiographs in a multicentre multi-reader crossover study. Methods: In this retrospective multicentre study, 726 radiographs from five Czech hospitals were interpreted twice by the seven clinicians: first without AI and, after a 30-day washout, with AI decision support. The endpoint was binary fracture detection. The reference standard was defined by independent review by one trauma surgeon and one radiologist; cases with discordant expert assessments were excluded. Paired analyses were performed on the expert-agreement set. Sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR) were calculated per reader, and overall * Corresponding author Email address: daniel.kvak@carebot.com (Daniel Kvak) performance was summarised using a nonparametric case-level bootstrap. Results: The primary analysis set included 630 radiographs, of which 197 were fracture-positive and 433 were fracture-negative. Overall bootstrap-estimated sensitivity was 86.26% (80.82;90.41) without AI and 94.56% (90.54; 97.00) with AI (p < 0.001). Overall specificity was 93.79% (91.14;95.72) without AI and 90.21% (87.08;92.69) with AI (p = 0.035). Overall NLR improved from 0.15 (0.10;0.21) to 0.06 (0.03;0.11) (p = 0.008), while overall PLR decreased from 14.24 (9.94;20.84) to 9.75 (7.36;12.97) (p = 0.134). Sensitivity increased in six of seven readers, and the largest gain occurred in the reader with the lowest baseline sensitivity. Conclusion: AI assistance reduced missed fractures across readers, with a modest and reader-dependent increase in false-positive fracture calls. These findings suggest that AI-assisted interpretation may function as a useful second-reader aid in musculoskeletal radiography.

Citovat

KLIČNÍK, Simon; Karel NĚMEC; Marika DAVIDOVÁ; Samuel BILANIN; Veronika JIRKŮ; Šimon KUBOV; Lukáš KORBEL; David NOHEJL; Petra OVESNÁ; Karolína KVAKOVÁ; Jakub DANDÁR a Daniel KVAK. AI Assistance Reduces Missed Fractures on Musculoskeletal Radiographs: A Multicentre Multi-Reader Crossover Study. European Journal of Radiology Artificial Intelligence. Amsterdam, Netherlands: Elsevier, 2026, roč. 7, s. 1-9. ISSN 3050-5771. Dostupné z: https://doi.org/10.1016/j.ejrai.2026.100106.

@article{2582682,
   author = {Kličník, Simon and Němec, Karel and Davidová, Marika and Bilanin, Samuel and Jirků, Veronika and Kubov, Šimon and Korbel, Lukáš and Nohejl, David and Ovesná, Petra and Kvaková, Karolína and Dandár, Jakub and Kvak, Daniel},
   article_location = {Amsterdam, Netherlands},
   doi = {https://doi.org/10.1016/j.ejrai.2026.100106},
   language = {eng},
   issn = {3050-5771},
   journal = {European Journal of Radiology Artificial Intelligence},
   title = {AI Assistance Reduces Missed Fractures on Musculoskeletal Radiographs: A Multicentre Multi-Reader Crossover Study},
   volume = {7},
   year = {2026}
}

TY  - JOUR
ID  - 2582682
AU  - Kličník, Simon - Němec, Karel - Davidová, Marika - Bilanin, Samuel - Jirků, Veronika - Kubov, Šimon - Korbel, Lukáš - Nohejl, David - Ovesná, Petra - Kvaková, Karolína - Dandár, Jakub - Kvak, Daniel
PY  - 2026
TI  - AI Assistance Reduces Missed Fractures on Musculoskeletal Radiographs: A Multicentre Multi-Reader Crossover Study
JF  - European Journal of Radiology Artificial Intelligence
VL  - 7
SP  - 1-9
EP  - 1-9
PB  - Elsevier
SN  - 30505771
N2  - Purpose: To quantify the impact of AI software on fracture detection on musculoskeletal radiographs in a multicentre multi-reader crossover study. Methods: In this retrospective multicentre study, 726 radiographs from five Czech hospitals were interpreted twice by the seven clinicians: first without AI and, after a 30-day washout, with AI decision support. The endpoint was binary fracture detection. The reference standard was defined by independent review by one trauma surgeon and one radiologist; cases with discordant expert assessments were excluded. Paired analyses were performed on the expert-agreement set. Sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR) were calculated per reader, and overall * Corresponding author Email address: daniel.kvak@carebot.com (Daniel Kvak) performance was summarised using a nonparametric case-level bootstrap. Results: The primary analysis set included 630 radiographs, of which 197 were fracture-positive and 433 were fracture-negative. Overall bootstrap-estimated sensitivity was 86.26% (80.82;90.41) without AI and 94.56% (90.54; 97.00) with AI (p < 0.001). Overall specificity was 93.79% (91.14;95.72) without AI and 90.21% (87.08;92.69) with AI (p = 0.035). Overall NLR improved from 0.15 (0.10;0.21) to 0.06 (0.03;0.11) (p = 0.008), while overall PLR decreased from 14.24 (9.94;20.84) to 9.75 (7.36;12.97) (p = 0.134). Sensitivity increased in six of seven readers, and the largest gain occurred in the reader with the lowest baseline sensitivity. Conclusion: AI assistance reduced missed fractures across readers, with a modest and reader-dependent increase in false-positive fracture calls. These findings suggest that AI-assisted interpretation may function as a useful second-reader aid in musculoskeletal radiography.
ER  -

KLIČNÍK, Simon; Karel NĚMEC; Marika DAVIDOVÁ; Samuel BILANIN; Veronika JIRKŮ; Šimon KUBOV; Lukáš KORBEL; David NOHEJL; Petra OVESNÁ; Karolína KVAKOVÁ; Jakub DANDÁR a Daniel KVAK. AI Assistance Reduces Missed Fractures on Musculoskeletal Radiographs: A Multicentre Multi-Reader Crossover Study. \textit{European Journal of Radiology Artificial Intelligence}. Amsterdam, Netherlands: Elsevier, 2026, roč.~7, s.~1-9. ISSN~3050-5771. Dostupné z: https://doi.org/10.1016/j.ejrai.2026.100106.

Přehled o publikaci