Study Compares Deep Learning Methods for Distinguishing Asthma from COPD Using Lung Sound Analysis
Researchers developed and evaluated CNN- and GRU-based deep learning models to differentiate asthma from COPD using 2D representations of pulmonary sounds, achieving a best cycle-based F1-score of 0.877. The study compared mel-frequency cepstral coefficient (MFCC) matrices, log-mel spectrograms, and a VAR model, finding MFCC with adaptive-length windowing to be the top-performing input representation. The findings suggest that automated respiratory sound classification could support differential diagnosis of these two commonly confused chronic lung diseases.
A preprint study posted to arXiv investigated how different 2D audio representations and feature fusion strategies affect the ability of deep learning models to distinguish asthma from COPD using respiratory cycle recordings. The researchers compared MFCC matrices, log-mel spectrograms, and a vector autoregressive (VAR) model as input representations, and addressed a key challenge in the field — inconsistent temporal dimensions caused by varying respiratory cycle lengths — by introducing an adaptive-length windowing technique alongside traditional trimming and zero-padding. Multiple CNN architectures were used to extract features from sub-phases of each respiratory cycle, and those features were then combined using direct concatenation, gated recurrent unit (GRU) networks, or GRU with attention mechanisms. The best cycle-based F1-score of 0.877 was achieved with 13-coefficient MFCC matrices at 64-point time resolution using direct concatenation, while the best subject-based F1-score of 0.855 used 256-point full-cycle MFCC representations. Notably, more sophisticated fusion strategies such as GRU with attention did not outperform simple concatenation, and data augmentation techniques — including mixup, the best-performing augmentation method tested — generally degraded model performance, underscoring the importance of authentic, high-quality training data in pulmonary sound research.
What's missing
It is unclear whether the models were validated on external datasets or only on internal splits, limiting conclusions about real-world clinical applicability. The paper also does not address how the models perform in the presence of comorbidities, background noise, or recordings from different devices — common challenges in clinical deployment.
What different sources said
- arXiv cs.AICenter
Optimizing 2D Input Representations and Sub-phase Fusion Strategies for Differential Diagnosis of Asthma and COPD Using CNN- and GRU-Based Networks
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.