Sinonasal outcomes on nasal acoustics and resonance (SONAR)

The nasal and paranasal cavities constitute integral components of the vocal tract resonator system, yet their precise contribution to voice quality remains a subject of debate. While early anatomical and acoustic models (1-3) suggested that sinus coupling through open ostia may introduce anti-resonances and spectral alterations, subsequent investigations employing cadaveric dissections, physical simulations, and three-dimensional replicas (4, 5) have confirmed notable effects on frequency response and formant balance. Clinical studies, including those examining patients with chronic rhinosinusitis undergoing functional endoscopic sinus surgery (FESS), have shown that surgical management can improve both sinonasal and vocal quality of life without adverse effects on voice characteristics (6,7). However, conventional acoustic measures such as jitter and shimmer frequently fail to detect such changes (8). In contrast, Mel-frequency cepstral coefficients (MFCCs), which transform the speech spectrum into a perceptually weighted domain, have shown superior sensitivity to resonance-related shifts and hypernasality (9). Based on this evidence, we investigated whether MFCCs and spectral flatness can more accurately identify post-operative alterations in resonance that are not captured by traditional acoustic parameters.