Reports from University of Sherbrooke Advance Knowledge in Engineering Science (Improving Deep Learning U-net Plus Plus By Discrete Wavelet and Attention Gate Mechanisms for Effective Pathological Lung Segmentation In Chest X-ray Imaging).

A report from the University of Sherbrooke in Canada discusses advancements in Engineering Science, specifically in improving the U-net architecture for deep learning in pathological lung segmentation from chest X-ray images. The new approach incorporates discrete wavelet transform and attention gate mechanisms, showing high efficacy with an accuracy of 99.1% on the Japanese Society of Radiological Technology dataset. The research highlights the potential for broad applications in medical imaging diagnostics, with consistent effectiveness across different datasets, including those related to COVID-19. [Extracted from the article]