Sushentsev, Nikita https://orcid.org/0000-0003-4500-9714
Arya, Zobair
Budd, Jobie
Frary, Amy
Moreira da Silva, Nadia
Ferrer Rodriguez, Mirjana
Burn, Paul
Hindley, Richard
Vasdev, Nikhil
Ibrahim, Mohamed
Bradley, Alison
Andreou, Adrian
Liyanage, Sidath
Persad, Raj
Aning, Jonathan
Ng, Alexander B. C. D.
Asif, Aqua
Kasivisvanathan, Veeru
Barrett, Tristan
Hinton, Mark
Padhani, Anwar Roshanali
Shah, Aarti
Davies, Lucy
Rix, Antony
Sala, Evis
Article History
Received: 9 September 2025
Revised: 19 December 2025
Accepted: 10 January 2026
First Online: 20 February 2026
Compliance with ethical standards
:
: The scientific guarantor of this publication is Evis Sala.
: The following authors declare relationships with the following companies: N.S.: consultant, Lucida Medical; Z.A., J.B., A.F., M.F.R., N.M.d.S., M.H., L.D., A.R., E.S.: employee/consultant/stockholder, Lucida Medical; J.A., R.P., R.H.: stockholder, Lucida Medical; N.V., A.S.: travel stipends, Lucida Medical; A.R.P.: consultant and stockholder, Lucida Medical, speakers bureau and research support, Siemens Healthineers. A.B.C.D.N. receives grants from Prostate Cancer UK outside the submitted work; V.K. receives grants from the John Black Charitable Foundation and Prostate Cancer UK, European Association of Urology Research Foundation, and Wolfgang Dieckmann Foundation and consultant to Telix Pharmaceuticals. The remaining authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
: One of the authors has significant statistical expertise.
: Written informed consent was obtained from all subjects (patients) in this study.
: Data for this study came from two studies. For the PRIME study, ethical approval was obtained from the National Research Ethics Committee West Midlands, Nottingham 21/WM/0091 on 28th June 2021. PAIR-1 had approval from the NHS Health Research Authority (IRAS #278640). No research ethics committee (REC) approval was required due to the data being anonymised and retrospective.
: Study subjects in this study came from the previously reported PAIR-1 and PRIME studies. The external validation set was the same as the PAIR-1 study, while a subset of the PRIME cohort was used for training the model here. This study reports a fundamentally different model (AI integrating clinical features as well as imaging), compared to the original PAIR-1 and PRIME studies.
: