Following the COVID-19 pandemic, public interest in healthcare has significantly in-creased, emphasizing the importance of early disease detection through lung sound analysis. Lung sounds serve as a critical biomarker in the diagnosis of pulmonary diseases, and numerous deep learning-based approaches have been actively explored for this purpose. Existing lung sound classification models have demonstrated high accuracy, benefiting from recent advances in artificial intelligence (AI) technologies. However, these models often rely on transmitting data to computationally intensive servers for processing, introducing potential security risks due to the transfer of sensitive medical information over networks. To mitigate these concerns, on-device AI has garnered growing attention as a promising solution for protecting healthcare data. On-device AI enables local data processing and inference directly on the device, thereby enhancing data security compared to server-based schemes. Despite these advantages, on-device AI is inherently limited by computational constraints, while conventional models typically require substantial processing power to maintain high performance. In this study, we propose a lightweight lung sound classification model designed specifically for on-device environments. The proposed scheme extracts audio features using Mel spectrograms, chromagrams, and Mel-Frequency Cepstral Coefficients (MFCC), which are converted into image representations and stacked to form the model input. The lightweight model performs convolution operations tailored to both temporal and frequency–domain characteristics of lung sounds. Comparative experimental results demonstrate that the proposed model achieves superior inference performance while maintaining a significantly smaller model size than conventional classification schemes, making it well-suited for deployment on resource-constrained devices.
31 May 2023 in Clinical and Experimental Otorhinolaryngology
Objectives. To compare the efficacy and safety of high-intensity focused ultrasound (HIFU) therapy with coblation for the treatment of inferior turbinate hypertrophy (ITH).Methods. In this randomized controlled clinical trial, 20 patients underwent inferior turbinate surgery, which consisted of either HIFU or coblation therapy. Efficacy, safety, and tolerability were evaluated by subjective symptom scores, acoustic rhinometry, and nasal endoscopy.Results. The modified nasal obstruction symptom evaluation (NOSE) score and nasal obstruction visual analog scale (NO-VAS) significantly decreased in both groups 12 weeks postoperatively. The between-group differences in the evaluation scores were not statistically significant. On nasal endoscopy, the HIFU patients showed improvements in mucosal swelling sooner than the patients undergoing coblation therapy. Nasal crusting significantly increased in the patients undergoing coblation compared to the patients undergoing HIFU therapy until postoperative week 4. Mucosal preservation was superior in the HIFU patients. Although HIFU was less painful than coblation therapy during the procedure, the difference was not significant (4.9 vs. 6.3, P=0.143). The difference in global satisfaction between the two groups was not statistically significant, although satisfaction was slightly higher among the HIFU patients than among the coblation patients (4.6 vs. 4.1, P=0.393).Conclusion. HIFU provided results similar to those of coblation therapy for patients with nasal obstruction due to ITH, but HIFU therapy caused less discomfort during the procedure. HIFU therapy appears to be a good noninvasive alternative to the current surgical modalities for ITH.