Related Experiment Video
Updated: Jun 24, 2025
07:15
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
6.8K
Risk score model to automatically detect prostate cancer patients by integrating diagnostic parameters
Rossana Castaldo1, Valentina Brancato1, Carlo Cavaliere1
1Bioinformatics and Biostatistics Lab, IRCCS SYNLAB SDN, Naples, Italy.
Frontiers in Oncology
|June 14, 2024
Summary
This study developed a risk model using radiomics, Prostate-Specific Antigen levels, and age to accurately detect prostate cancer (PCa). The model achieved 91% AUC, aiding in PCa diagnosis and management.
Area of Science:
- Oncology
- Medical Imaging
- Data Science
Background:
- Prostate cancer (PCa) is a leading cancer in men.
- Multiparametric MRI is crucial for PCa localization and staging.
- Radiomics offers potential for PCa detection, aggression characterization, and recurrence monitoring, potentially reducing unnecessary biopsies.
Purpose of the Study:
- To develop a risk model score for automatic PCa detection.
- Integrate radiomics, Prostate-Specific Antigen (PSA) levels, and patient age.
- Enhance PCa understanding and management while reducing invasive procedures.
Main Methods:
- Utilized a dataset of 189 PCa patients who underwent bi-parametric MRI.
- Employed an Elastic-Net Regularized Generalized Linear Model.
- Integrated non-invasive diagnostic parameters: radiomics, PSA levels, and age.
Main Results:
- The developed model achieved an Area Under the Curve (AUC) of 91% for automatic PCa detection.
- The risk score was effective in assessing equivocal PCa cases at biopsy.
- Performance was compared against the bi-parametric PI-RADS v2 standard.
Conclusions:
- A risk model combining radiomics, PSA, and age provides objective and accurate PCa risk stratification.
- This integrated approach supports clinical decision-making in PCa follow-up.
- The model shows promise in improving the diagnostic pathway for prostate cancer.
