(UroToday.com) The 2023 EAU annual meeting included a session on the role of imaging and PSA density for biopsy indication, and tumor staging, featuring a presentation by Dr. Karsten Günzel discussing a prospective evaluation of artificial intelligence to improve cancer detection in transperineal MRI/ultrasound fusion biopsies of the prostate. Due to the complexity of the evaluation, there are qualitative differences in the interpretation of mpMRI of the prostate:
The development of computer-aided diagnosis systems could improve the quality of MRI reporting. The artificial intelligence assisted software Quantib®Prostate was developed to facilitate the detection of prostate cancer suspicious lesions using T2W-, DWI- and DCE-sequences:
The aim of this study was to evaluate the diagnostic performance of an artificial intelligence software in comparison to PI-RADS reporting for prostate cancer detection, in men undergoing transperineal MRI/ultrasound fusion biopsies of the prostate (MRI-TRUS-transperineal biopsy).
There were 201 patients prospectively enrolled from January 2022 to May 2022. The patient inclusion flow chart is as follows:
Targeted biopsies were performed based on radiological findings (reports of 23 different institutes) and systematic biopsies were adapted based on suspicious lesions (ROIs) detected by the artificial intelligence-software. The primary endpoint was prostate cancer detection, and secondary endpoints included (i) the concordance of the localization of index lesions in the MRI reports with the detected ROIs by the artificial intelligence software, and (ii) prostate cancer detection in additional ROIs by the artificial intelligence.
The median patient age was 68 years (IQR 61-73), median PSA was 6.27 ng/ml (IQR 4.63-10.25), and median prostate volume was 43 ml (IQR 33-62). A median of 10 (IQR 8-10) biopsy cores were taken per patient:
There were 265 suspicious lesions detected by radiological reporting and 404 by artificial intelligence. Prostate cancer detection rates according to PI-RADS 3, 4, 5 were 29%, 76%, 92% for radiological reporting and 36%, 77%, 92% for artificial intelligence-assisted analysis, respectively. The prostate cancer detection rate summary is as follows:
Artificial intelligence detected all clinically relevant PI-RADS index lesions. There were 10 (5%) prostate cancers detected by the artificial intelligence software only, and in 43 patients (21%) additional prostate cancer-positive ROIs were found by artificial intelligence. Furthermore, 174/190 (92%) of PI-RADS index lesions on mpMRI were detected by artificial intelligence software:
The false positive rate of described suspicious lesions by radiological findings was 43% and by artificial intelligence 52%.
Dr. Günzel concluded this presentation by discussing a prospective evaluation of artificial intelligence to improve cancer detection in transperineal MRI/ultrasound fusion biopsies of the prostate with the following take-home messages:
- All clinically relevant PI-RADS index lesions in the MRI were detected by artificial intelligence software
- Using artificial intelligence with MRI/ultrasound fusion biopsies could increase prostate cancer detection and minimize radiological misdiagnosis
- The high false positive rate does not help to avoid biopsies
- Radiological expertise is required
Presented by: Karsten Günzel, Medizinische Hochschule Brandenburg - Theodor Fontane, Dept. of Urology, Berlin, Germany
Co-Authors: Hamm C.2, Heinrich S.3, Schlegel J.3, Schöndorf D.3, Ozimek T.3, Busch J.3, Magheli A.3, Hinz S.3
Affiliations: 1Medizinische Hochschule Brandenburg - Theodor Fontane, Dept. of Urology, Berlin, Germany, 2Charité - Universitätsmedizin Berlin, Dept. of Radiology, Berlin, Germany, 3Vivantes Klinikum Am Urban, Dept. of urology, Berlin, Germany
Written by: Zachary Klaassen, MD, MSc – Urologic Oncologist, Assistant Professor of Urology, Georgia Cancer Center, Augusta University/Medical College of Georgia, @zklaassen_md on Twitter during the 2023 European Association of Urology (EAU) 38th annual congress held in Milan, Italy between March 10-13, 2023