Development and Validation of a Non-Invasive AI Diagnostic Model for Prostate Cancer Using Multiparametric MRI and PSMA PET/CT

A Retrospective, Multicenter Study to Develop and Validate a Non-Invasive Artificial Intelligence Diagnostic Model for Prostate Cancer Using Multiparametric MRI and PSMA PET/CT, With Pathologically Confirmed Diagnosis as the Reference Standard

Prostate cancer is one of the most common malignancies in men. Currently, due to the limited diagnostic accuracy of existing imaging tests, there is a risk of missed diagnosis or unnecessary prostate biopsy. This study aims to develop and validate a non-invasive artificial intelligence (AI) diagnostic model using two advanced imaging techniques: multiparametric MRI (mpMRI) and PSMA PET/CT. By integrating information from both imaging modalities, the AI model is expected to improve the diagnostic accuracy of prostate cancer, reduce unnecessary biopsies, and assist physicians in making better clinical decisions. This is a retrospective, multicenter study that plans to collect imaging and pathology data from approximately 1,000 to 1,500 patients across six major hospitals in China. The diagnostic performance of the model will be evaluated, including its ability to identify clinically significant prostate cancer and its value in assisting diagnosis in patients with PSA levels in the gray zone (4-20 ng/mL).

Study Overview

• Status: Enrolling by invitation
• Conditions: Prostate Cancer (Diagnosis)

Detailed Description

Study Design:

This is a retrospective and prospective, multicenter, case-control study. The study aims to develop and validate a non-invasive artificial intelligence (AI) diagnostic model for prostate cancer by integrating multiparametric MRI (mpMRI) and PSMA PET/CT imaging.

Participants:

Patients who underwent mpMRI, PSMA PET/CT, and prostate biopsy or radical prostatectomy at participating centers will be retrospectively enrolled. Eligible participants include pathologically confirmed prostate cancer patients (cases) and benign prostatic hyperplasia (BPH) patients (controls). Inclusion criteria include age ≥18 years, ECOG performance status 0-2, life expectancy >6 months, and availability of complete clinical data (PSA, Gleason score, PI-RADS, SUVmax, prostate volume, etc.). Exclusion criteria include prior prostate cancer treatment (endocrine therapy or radiotherapy), previous prostate surgery (e.g., TURP), severe renal insufficiency, other malignancies. Informed consent is waived for retrospective patients, while signed informed consent is required for prospective patients.

Sample Size:

Approximately 1,000 to 1,500 participants will be enrolled from six hospitals in China: Xiangya Hospital of Central South University, Qilu Hospital of Shandong University, Chinese PLA General Hospital, The First Affiliated Hospital of Guangzhou Medical University, Beijing Hospital and Renji Hospital, Shanghai Jiao Tong University School of Medicine.

AI Model Development:

The AI model will be developed using deep learning and radiomics techniques. mpMRI sequences (including T2-weighted, DWI/ADC, and DCE) and PSMA PET/CT images will be preprocessed, coregistered, and fused. The model will be trained to distinguish clinically significant prostate cancer (csPCa) from non-csPCa or benign conditions. The reference standard is histopathology from prostate biopsy or radical prostatectomy.

Outcome Measures:

Primary outcome measures include the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity of the AI model for csPCa detection. Secondary outcome measures include positive predictive value (PPV), negative predictive value (NPV), overall accuracy, decision curve analysis (DCA) net benefit, diagnostic performance in the PSA gray zone (4-20 ng/mL) subgroup, and the proportion of patients who could avoid biopsy at 100% specificity threshold.

Statistical Analysis:

Model performance will be evaluated using internal cross-validation and external validation on data from different centers. Calibration curves and decision curve analysis will be used to assess clinical utility. Subgroup analyses will be performed for PSA gray zone patients and different Gleason grade groups.

• Study Type: Observational
• Enrollment (Estimated): 1500

Contacts and Locations

Study Locations

• China
  • Beijing Municipality
    • Beijing, Beijing Municipality, China, 100853
      • Chinese PLA General Hospital
    • Beijing, Beijing Municipality, China, 100730
      • Beijing Hospital
  • Guangdong
    • Guangzhou, Guangdong, China, 510120
      • The First Affiliated Hospital of Guangzhou Medical University
  • Shandong
    • Jinan, Shandong, China, 250012
      • Qilu Hospital of Shandong University
  • Shanghai Municipality
    • Shanghai, Shanghai Municipality, China, 200127
      • Renji Hospital, Shanghai Jiao Tong University School of Medicine
• United States
  • California
    • Los Angeles, California, United States, 90095
      • David Geffen School of Medicine University of California, Los Angeles(UCLA)

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

This study will enroll patients who underwent mpMRI, PSMA PET/CT, and prostate biopsy or radical prostatectomy at six major hospitals in China: Xiangya Hospital of Central South University, Qilu Hospital of Shandong University, Chinese PLA General Hospital, The First Affiliated Hospital of Guangzhou Medical University, Beijing Hospital and Renji Hospital, Shanghai Jiao Tong University School of Medicine. The study population includes two groups: patients with pathologically confirmed prostate cancer (cases) and patients with pathologically confirmed benign prostatic hyperplasia (BPH, controls). All participants are male, aged 18 years or older, with no prior prostate cancer treatment (endocrine therapy or radiotherapy) or prostate surgery. Patients with severe renal insufficiency or other malignancies are excluded. A total of approximately 1,000 to 1,500 participants will be enrolled retrospectively from the participating centers.

Description

Inclusion Criteria:

1. Age ≥ 18 years.
2. ECOG performance status 0-2.
3. Life expectancy > 6 months.
4. Underwent mpMRI and PSMA PET/CT before systemic treatment or radical prostatectomy, with original DICOM data available for export.
5. Has pathological diagnosis from prostate biopsy or radical prostatectomy as the gold standard.
6. Complete clinical data available, including pre-treatment PSA (tPSA, fPSA), TNM stage, Gleason score, PI-RADS score, SUVmax, prostate volume (from MRI/PSMA PET), age, and BMI.
7. Informed consent for data use for research purposes according to each center's ethics requirements.

Exclusion Criteria:

1. History of other malignant tumors
2. Previous prostate surgery (e.g., TURP)
3. Prior endocrine therapy or radiotherapy
4. Severe renal insufficiency
5. Major organ dysfunction or life expectancy < 1 year

Study Plan

How is the study designed?

Number of groups / cohorts

2

Cohorts and Interventions

Group / Cohort
Prostate Cancer Group
Non-cancer (BPH) Control

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Area Under the Curve (AUC) of the AI Model for Detecting Clinically Significant Prostate Cancer	The AUC (Area Under the Receiver Operating Characteristic Curve) will be calculated to evaluate the overall diagnostic performance of the AI model in distinguishing clinically significant prostate cancer (csPCa) from non-csPCa or benign conditions. The gold standard is histopathology from prostate biopsy or radical prostatectomy.	At histopathological diagnosis by prostate biopsy or radical prostatectomy
Specificity of the AI Model for Detecting Clinically Significant Prostate Cancer	Specificity (true negative rate) will be calculated to evaluate the model's ability to correctly identify patients without clinically significant prostate cancer. High specificity is a primary goal to reduce unnecessary prostate biopsies.	At histopathological diagnosis by prostate biopsy or radical prostatectomy
Sensitivity of the AI Model for Detecting Clinically Significant Prostate Cancer	Sensitivity (true positive rate) will be calculated to evaluate the model's ability to correctly identify patients with clinically significant prostate cancer.	At histopathological diagnosis by prostate biopsy or radical prostatectomy

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Overall Diagnostic Accuracy	Overall accuracy (proportion of true results among all cases) will be calculated for the AI model in detecting clinically significant prostate cancer.	At histopathological diagnosis by prostate biopsy or radical prostatectomy
Net Benefit of the AI Model for Detecting Clinically Significant Prostate Cancer	Net benefit will be calculated across a range of threshold probabilities to evaluate the clinical utility of the AI model for detecting clinically significant prostate cancer, using histopathological diagnosis from prostate biopsy or radical prostatectomy as the reference standard. Results will be compared with default strategies of biopsying all patients or no patients.	At histopathological diagnosis by prostate biopsy or radical prostatectomy
Proportion of Patients Who Could Avoid Biopsy at 100% Specificity Threshold	At a specificity of 100% (no false positives), the proportion of patients who would be correctly identified as non-csPCa and thus could safely avoid prostate biopsy will be calculated. Sensitivity at this threshold will also be reported.	At histopathological diagnosis by prostate biopsy or radical prostatectomy
AUC in PSA Gray Zone (4-20 ng/mL)	The area under the receiver operating characteristic curve (AUC) of the AI model will be evaluated in patients with PSA levels between 4 and 20 ng/mL to assess discrimination between clinically significant prostate cancer and non-clinically significant or benign disease, using histopathological diagnosis as the reference standard.	At histopathological diagnosis by prostate biopsy or radical prostatectomy
Specificity in PSA Gray Zone (4-20 ng/mL)	Specificity of the AI model will be evaluated in patients with PSA levels between 4 and 20 ng/mL using histopathological diagnosis as the reference standard.	At histopathological diagnosis (prostate biopsy or radical prostatectomy)

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Sensitivity in PSA Gray Zone (4-20 ng/mL)	Sensitivity of the AI model will be evaluated in patients with PSA levels between 4 and 20 ng/mL, using histopathological diagnosis as the reference standard.	At histopathological diagnosis (prostate biopsy or radical prostatectomy)

Collaborators and Investigators

• Sponsor: Xiangya Hospital of Central South University
• Collaborators: Chinese PLA General Hospital; RenJi Hospital; Qilu Hospital of Shandong University; The First Affiliated Hospital of Guangzhou Medical University

Study record dates

Study Major Dates

• Study Start (Actual): January 13, 2026
• Primary Completion (Estimated): September 30, 2026
• Study Completion (Estimated): December 31, 2026

Study Registration Dates

• First Submitted: June 16, 2026
• First Submitted That Met QC Criteria: July 1, 2026
• First Posted (Actual): July 8, 2026

Study Record Updates

• Last Update Posted (Actual): September 15, 2026
• Last Update Submitted That Met QC Criteria: September 14, 2026
• Last Verified: May 1, 2026

More Information

Terms related to this study

• Keywords: Prostate Cancer; Artificial Intelligence; Multiparametric MRI; PSMA PET/CT; Non-invasive Diagnosis
• Additional Relevant MeSH Terms: Urogenital Diseases; Genital Diseases; Pathologic Processes; Genital Neoplasms, Male; Urogenital Neoplasms; Neoplasms by Site; Neoplasms; Genital Diseases, Male; Prostatic Diseases; Male Urogenital Diseases; Pathological Conditions, Signs and Symptoms; Prostatic Neoplasms; Disease
• Other Study ID Numbers: 2026010083_2

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No