Advancing Prostate Cancer Diagnosis: A Deep Learning Approach for Enhanced Detection in MRI Images

Area of Science:

• Medical Imaging
• Artificial Intelligence in Oncology
• Radiology

Background:

• Prostate cancer is a major global health concern for men.
• Accurate diagnostic methods are vital for effective treatment.
• Magnetic Resonance Imaging (MRI) is a key tool for prostate cancer detection.

Purpose of the Study:

• To enhance prostate cancer detection accuracy using deep learning on MRI.
• To evaluate the performance of an ensemble of deep learning models.
• To improve the robustness of diagnostic capabilities in prostate cancer imaging.

Main Methods:

• Utilized multiple deep learning models: 3D Convolutional Neural Network, Residual Network, and Inception Network.
• Employed an ensemble method with soft voting to combine model predictions.
• Applied meticulous pre-processing and data augmentation techniques to MRI data.

Main Results:

• The ensemble model achieved high performance metrics: 91.3% accuracy, 90.2% sensitivity, 92.1% specificity.
• Demonstrated strong precision (89.8%) and F1 score (90.0%) on the PROSTATEx dataset.
• The integrated deep learning approach showed state-of-the-art results.

Conclusions:

• Ensemble deep learning methods offer significant improvements for prostate cancer diagnostics.
• This approach provides a robust and precise tool for clinical applications.
• Advanced AI integration in MRI enhances diagnostic capabilities for prostate cancer.