Basar, Yeliz https://orcid.org/0000-0002-1321-3617
Seker, Mustafa Ege https://orcid.org/0000-0001-7664-5786
Kirova-Nedyalkova, Galina Ivanova https://orcid.org/0000-0002-0158-8972
Milkovska, Elina Plamenova
Manov, Boyan Emilov
Tamturk, Atahan
Tasci, Ilke https://orcid.org/0009-0000-8143-6583
Karadag, Melih
Sarac, Nuri
Duzgun, Selin Ardali https://orcid.org/0000-0002-2623-2542
Karakoc, Erencan https://orcid.org/0000-0002-8945-5364
Karabulut, Ahmet
Kavi, Cemre https://orcid.org/0000-0003-0181-1393
Buyukkirli, Kaan
Onal, Mehmet Onur https://orcid.org/0009-0001-9300-2907
Alis, Deniz https://orcid.org/0000-0002-7045-1793
Savas, Recep https://orcid.org/0000-0002-7520-760X
Karaarslan, Ercan https://orcid.org/0000-0002-4581-4273
Durhan, Gamze https://orcid.org/0000-0002-6281-9287
Article History
Received: 10 March 2026
Accepted: 25 May 2026
First Online: 25 July 2026
Declarations
Competing interests: Deniz Alis is the CEO and co-founder of Hevi AI Health Tech, the company that developed the evaluated hChestXR system. The remaining authors declare no competing interests. To mitigate potential bias, the evaluated model version and manufacturer’s default threshold were locked before analysis; the study used deidentified retrospective data; reference labeling and adjudication were performed by academic radiologists; and statistical analysis and table generation were performed independently of the company. Statistical analysis was performed independently by M.E.S. Non-company academic authors had access to the deidentified data from their respective centers. The company-affiliated author had no role in reference labeling, adjudication, statistical analysis, table generation, interpretation of results, preparation of the revised manuscript, or the decision to submit the manuscript. The company-affiliated author’s role was limited to conceptualization and provision of technical information/resources regarding the evaluated AI system.
Consent for publication: Written informed consent was not required because it was related to identification imaging only, and patient anonymity was maintained.
Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.The local ethics committee approved this retrospective study and waived the need for informed consent for the retrospective evaluation of anonymized medical data (Acibadem University and Acibadem Healthcare Institutions Medical Research Ethics Committee, Date: 21/11/2024, Number: 2024-18/676).