New Blood Test Models Reduce Prostate Biopsies
Researchers developed machine-learning diagnostic tools to lower unnecessary prostate procedures for men with specific PSA levels.
Updated on Oct. 27, 2026 in Men’s Health
Live Poll
Would you trust a machine-learning model to decide if you need a biopsy for prostate cancer?
Scientists have created new diagnostic models using routine blood tests to improve prostate cancer detection in men with PSA levels between 4 and 10 ng/mL. The approach aims to significantly reduce the number of unnecessary biopsies currently ordered based on standard screening results.
Why it matters
Prostate cancer screening often leads to invasive biopsies for patients whose PSA levels fall into the ambiguous 4-10 ng/mL range. Improving the specificity of these tests helps physicians distinguish between low-risk cases and those requiring urgent medical intervention.
A diagnostic model validation study evaluated 1,758 patients in a development cohort and 9,524 in a validation cohort. The machine-learning model achieved an internal validation AUC of 0.825 and reached 39.5% specificity at 95% sensitivity.
The details
The research team integrated routine hematologic and biochemical variables with standard PSA data to build their models. Performance was verified through 200 bootstrap resamples and an external validation study conducted using a patient database in Japan.
Timeline
Data collection for development and validation cohorts occurred from 2020 to 2024.
Findings were presented at the ESMO Annual Congress held October 23-27, 2026.
Health Landscape
This study follows research presented at the European Society for Medical Oncology (ESMO) annual clinical research programs regarding new diagnostic protocols. The findings signal a broader transition toward integrating machine learning and routine electronic health data to refine cancer screening precision.
Patients with PSA levels between 4 and 10 ng/mL may eventually see fewer invasive procedures as these models move toward clinical adoption. The method utilizes routine testing, which could simplify the diagnostic journey without requiring specialized, high-cost equipment.
The takeaway
These diagnostic models highlight how existing electronic medical records can be leveraged to increase the accuracy of routine cancer screening. Patients should discuss the implications of borderline PSA levels with their urologist to understand the current biopsy standard of care.
Further reading
Learn more about the latest innovations in Men’s Health.
Live Poll
Would you trust a machine-learning model to decide if you need a biopsy for prostate cancer?
