Prostate cancer is the most common cancer in men in the United States, characterized by uncontrolled cell growth in the prostate gland. Deep learning and medical image analysis significantly enhance its diagnosis and treatment by analyzing MRI, CT, and ultrasound data to detect abnormalities and assess tumor characteristics. These advanced algorithms improve diagnostic accuracy by identifying patterns that may be missed by human observers. Additionally, they aid in treatment planning by delineating tumor boundaries, predicting treatment response, and monitoring disease progression. Integrating these technologies enhances precision in prostate cancer management, leading to improved patient outcomes.
Prostate Segmentation on Micro Ultrasound Images
MicroSegNet is a clinically practical segmentation tool which has the potential to assist urologists in prostate biopsy and cancer diagnosis, offering valuable support in clinical decision-making.
MRI-Histopathology Image Registration
RAPHIA pipeline automates prostate segmentation using deep learning approaches to refine labor-intensive and time-consuming protocols we currently employ.
Micro-Ultrasound-Histopathology Image Registration
First deep-learning-based semi-automated approach for registering in-vivo micro-US and ex-vivo pseudo-whole mount histopathology images.
ProsRegNet: A Deep Learning Registration Model
First application of deep learning for MRI-histopathology image registration. Approximately 20-times faster than conventional approaches.
