Ph.D., D.Sc. Michał Byra, a longtime researcher at the Ultrasound Department, has earned his habilitation in biomedical engineering at IPPT PAN. As part of this achievement, he presented a series of eight publications titled "Deep Learning Methods in Ultrasound Diagnosis of Breast Cancer."

His research focuses on machine learning, particularly neural networks for medical image analysis and processing, as well as computational methods in ultrasound imaging. He has also pioneered a new research direction in quantitative ultrasound (QUS), which involves analyzing raw ultrasound signals to determine the physical properties of tissues. By leveraging deep learning, neural networks trained on raw ultrasound signals enable automatic feature extraction for tissue characterization, overcoming the limitations of traditional QUS techniques. His work also explores transfer learning, allowing neural networks pre-trained on large non-medical datasets to be adapted for ultrasound image analysis. Additionally, his research includes pioneering studies demonstrating the first applications of neural networks in predicting breast cancer response to chemotherapy.

Zaproszenie na studia doktoranckie

Zakład Ultradźwięków

Instytut Podstawowych Problemów Techniki PAN