Related Experiment Video
Updated: Jun 27, 2025
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
PSMA-positive prostatic volume prediction with deep learning based on T2-weighted MRI
Riccardo Laudicella1,2,3, Albert Comelli4, Moritz Schwyzer5
1Department of Nuclear Medicine, University Hospital Zürich, University of Zurich, Zurich, Switzerland. riclaudi@hotmail.it.
Purpose:
High PSMA expression might be correlated with structural characteristics such as growth patterns on histopathology, not recognized by the human eye on MRI images. Deep structural image analysis might be able to detect such differences and therefore predict if a lesion would be PSMA positive. Therefore, we aimed to train a neural network based on PSMA PET/MRI scans to predict increased prostatic PSMA uptake based on the axial T2-weighted sequence alone.
Material And Methods:
All patients undergoing simultaneous PSMA PET/MRI for PCa staging or biopsy guidance between April 2016 and December 2020 at our institution were selected. To increase the specificity of our model, the prostatic beds on PSMA PET scans were dichotomized in positive and negative regions using an SUV threshold greater than 4 to generate a PSMA PET map. Then, a C-ENet was trained on the T2 images of the training cohort to generate a predictive prostatic PSMA PET map.
Results:
One hundred and fifty-four PSMA PET/MRI scans were available (133 [68Ga]Ga-PSMA-11 and 21 [18F]PSMA-1007). Significant cancer was present in 127 of them. The whole dataset was divided into a training cohort (n = 124) and a test cohort (n = 30). The C-ENet was able to predict the PSMA PET map with a dice similarity coefficient of 69.5 ± 15.6%.
Conclusion:
Increased prostatic PSMA uptake on PET might be estimated based on T2 MRI alone. Further investigation with larger cohorts and external validation is needed to assess whether PSMA uptake can be predicted accurately enough to help in the interpretation of mpMRI.
Insights
This study shows that artificial intelligence can predict prostate-specific membrane antigen (PSMA) levels using only T2 MRI scans. This deep learning approach may help identify aggressive prostate cancer (PCa) without needing PET imaging.
Area of Science:
- Radiology
- Artificial Intelligence
- Oncology
Background:
- Prostate-specific membrane antigen (PSMA) expression in prostate cancer (PCa) may correlate with histopathological features not visible on standard MRI.
- Deep learning analysis of structural MRI could potentially predict PSMA positivity.
- Predicting PSMA uptake from MRI alone could enhance PCa staging and treatment planning.
Purpose of the Study:
- To train a neural network using PSMA PET/MRI scans to predict increased prostatic PSMA uptake solely from axial T2-weighted MRI sequences.
- To assess the feasibility of deep structural image analysis for predicting PSMA expression in PCa.
- To develop a predictive model for PSMA uptake based on MRI findings.
Main Methods:
- A cohort of 154 patients undergoing simultaneous PSMA PET/MRI for PCa staging or biopsy guidance was analyzed.
- Prostatic beds on PSMA PET scans were dichotomized into positive and negative regions using an SUV threshold > 4 to create a PSMA PET map.
- A Convolutional Encoder-Decoder Network (C-ENet) was trained on T2-weighted MRI images to generate a predictive PSMA PET map.
Main Results:
- The study included 154 PSMA PET/MRI scans, with significant cancer detected in 127 patients.
- The dataset was split into a training cohort (n=124) and a test cohort (n=30).
- The C-ENet model achieved a Dice similarity coefficient of 69.5% ± 15.6% in predicting the PSMA PET map from T2 MRI.
Conclusions:
- Prostatic PSMA uptake on PET imaging may be predictable using only T2-weighted MRI.
- This AI-driven approach shows potential for estimating PSMA uptake, aiding mpMRI interpretation.
- Further validation with larger cohorts is necessary to confirm the accuracy and clinical utility of this predictive method.
