Brief

Two new research papers propose advanced AI techniques for classifying respiratory sounds. One paper introduces QLung, a quality-adaptive framework that adjusts learning margins based on audio recording quality, improving performance on the ICBHI and SPRSound datasets. The other paper, Lung-SRAD, explores State Space Models as an alternative to Transformers for this task, incorporating spectral-aware regularization and contrastive learning to achieve a 5% improvement over baseline methods on the ICBHI benchmark. AI

Researchers have developed a meta-ensemble learning approach to improve the accuracy of respiratory sound classification models. This method trains base models on varied data splits to enhance prediction diversity, which is then combined by a meta-model for better generalization. The technique achieved state-of-the-art results on the ICBHI benchmark, demonstrating improved performance and applicability to real-world clinical data. AI