The Application of Radiomics and AI to Molecular Imaging for Prostate Cancer.

Area of Science:

• Oncology
• Radiology
• Medical Imaging
• Artificial Intelligence

Background:

• Molecular imaging, particularly Prostate-Specific Membrane Antigen-based (PSMA-based) positron emission tomography/computed tomography (PET/CT), is crucial for prostate cancer (PCa) management.
• Machine learning (ML) and artificial intelligence (AI) are emerging technologies with growing applications in medical imaging.
• While Magnetic Resonance (MR) imaging reviews are available, there's a need for focused reviews on AI in PCa PET/CT.

Purpose of the Study:

• To introduce available AI technologies relevant to molecular imaging.
• To provide a comprehensive overview of AI applications in PSMA PET/CT for prostate cancer.
• To highlight the clinical utility of AI in PCa diagnosis, staging, and treatment.

Main Methods:

• Review of current literature on AI and ML in prostate cancer molecular imaging.
• Discussion of AI techniques including radiomics, convolutional neural networks (CNNs), and generative adversarial networks (GANs).
• Explanation of supervised, unsupervised, and semi-supervised learning methods in the context of PCa imaging.

Main Results:

• AI and ML offer advanced tools for analyzing complex molecular imaging data in PCa.
• AI facilitates improved accuracy in diagnosis, staging, and target volume definition for radiation therapy planning.
• AI aids in predicting and monitoring treatment outcomes for prostate cancer patients.

Conclusions:

• AI is transforming prostate cancer molecular imaging, particularly with PSMA PET/CT.
• AI-driven insights enhance clinical decision-making across the PCa patient journey.
• Further integration of AI in PCa imaging promises more personalized and effective cancer care.