Diagnostic Model Based on Efect Size Across Phases of Respiratory Sounds Solunum Seslerinin Fazlari Boyunca Etki Büyüklü?güne Dayali Tani Modeli
34th Signal Processing and Communications Applications Conference, SIU 2026, İstanbul, Türkiye, 7 - 10 Temmuz 2026, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/siu71813.2026.11636412
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Anahtar Kelimeler: asthma, Cohens'd, diagnostic classification, effect size, Pulmonary sounds
- Yıldız Teknik Üniversitesi Adresli: Evet
Özet
The diagnostic accuracy of stethoscopes and pulmonary function tests used to detect respiratory diseases may be inadequate. This poses a significant limitation to early and accurate diagnosis. This study proposes an approach based on the effect size between the sub-phases of the respiratory cycle. As part of the study, 14-channel respiratory sounds from 135 participants (75 healthy individuals and 60 asthma patients) were divided into sub-phases of the respiratory cycle. Cohen's d effect size was then calculated between sub-phases based on their power spectral density estimates and used as a feature. The features were used as input for five different classifiers. The results demonstrated that the feature composed of effect sizes calculated between consecutive sub-phases achieved successful performance with an F1 score of 0.878, accuracy of 0.889, precision of 0.900, and specificity of 0.880.