
AI-Based Analysis of Nasal Airflow Acoustic Signals
Odiosoft-Rhino is an innovative digital health technology developed by Seren and first described in 2005 for the objective assessment of nasal airway function through acoustic signal analysis. Unlike conventional methods that directly measure airflow resistance, Odiosoft-Rhino analyzes the sounds generated during spontaneous nasal breathing and transforms acoustic information into quantitative parameters reflecting airway patency and airflow characteristics.
The system is based on the principle that changes in nasal airway geometry alter airflow dynamics and consequently modify the acoustic properties of breathing sounds. As airway narrowing increases airflow turbulence, higher-frequency and higher-intensity acoustic components are generated. Through advanced spectral and frequency-domain analysis, Odiosoft-Rhino provides an objective, non-invasive evaluation of nasal obstruction and upper airway function (1,2).
The software performs detailed analysis of respiratory sound signals, allowing frequency measurements to be obtained separately from each nasal cavity while displaying sound intensity across different frequency bands. Because airway diameter is a major determinant of airflow turbulence, acoustic analysis represents a sensitive and physiologically relevant approach for evaluating nasal obstruction (3).
Clinical studies have demonstrated the utility of Odiosoft-Rhino in healthy individuals, patients with allergic rhinitis, and subjects with nasal septal deviation. Comparative investigations have shown that its diagnostic performance is comparable to rhinomanometry while exhibiting strong correlations with patient-reported symptom scores (2,4-7).
A major advantage of Odiosoft-Rhino is its portability and accessibility. The system requires only the built-in microphone of a standard computer, tablet, or mobile phone, making it suitable not only for clinical applications but also for large-scale screening, remote monitoring, and telemedicine. Acoustic recordings can be stored as WAV files and securely transferred via the internet for remote analysis and consultation (8).
Today, Odiosoft-Rhino represents a pioneering digital respiratory acoustics platform that integrates signal processing and artificial intelligence technologies to enable objective, scalable, and accessible assessment of nasal and upper airway function.
The patient should perform the recording in a quiet environment free from external sounds and ambient noise. After opening the Odiosoft-Rhino software on a computer, tablet, or mobile phone, two options will be available under the "Analysis" menu: Expiratory Analysis and Inspiratory Analysis.
If "Expiratory Analysis" is selected for nasal airflow assessment, two separate recording fields corresponding to the right and left nasal cavities will appear. To analyze the right nasal cavity, the patient should gently occlude the left nostril using the left thumb. The microphone should be held in the right hand approximately 5-6 cm from the nostril and positioned parallel to the nasal airflow.
Recording may be controlled either manually by the user or automatically by the software. In automatic mode, recording begins with the onset of exhalation and stops automatically 2 seconds after exhalation has ended.
Using AI-assisted analysis of expiratory nasal acoustic biomarkers, the degree and anatomical location of nasal narrowing during expiration are displayed as an anatomical narrowing map. The following parameters are calculated:
For "Inspiratory Analysis", the procedure is performed symmetrically. AI-based inspiratory acoustic biomarker analysis provides the following additional parameters in tabular format:
Furthermore, an estimated Sleep Quality Score is calculated. Parameters can be exported as a PDF report and stored for longitudinal comparison. Additionally, nasal sound recordings acquired using external devices or third-party platforms can also be analyzed using the Odiosoft-Rhino software.
Rhinomanometry (RMM) is widely used as an objective method for the evaluation of nasal airway resistance by measuring transnasal pressure-flow relationships. Despite its role as a traditional reference technique in rhinology, multiple studies have demonstrated that its clinical applicability is limited by methodological and technical constraints [1,2].
One of the primary limitations of RMM is its strong dependence on strictly standardized testing conditions. The need for an airtight facial mask introduces the risk of air leakage artifacts. RMM is also highly dependent on patient cooperation, reducing reproducibility particularly in pediatric, elderly, or uncooperative populations [2,4].
Acoustic Rhinometry (AR) evaluates nasal cavity geometry by analyzing reflected sound waves within the nasal airway. A key limitation of AR is reduced accuracy in the posterior nasal cavity. AR also provides predominantly static anatomical measurements and does not adequately reflect real-time airflow dynamics during respiration, limiting its ability to characterize functional airway collapse or cycle-dependent obstruction patterns [5,6,7].
Odiosoft-Rhino (OR) represents a novel acoustic-based, non-invasive approach that evaluates nasal airway function by analyzing sound signals generated by nasal airflow. Unlike AR and RMM, which rely on acoustic reflection or pressure-flow measurements, OR provides a dynamic assessment of nasal airflow characteristics without requiring invasive interfaces or complex patient maneuvers.
By shifting diagnostics from hardware-intensive workflows to software-defined respiratory intelligence, Odiosoft-Rhino reduces infrastructure dependency, expands access to care, and supports more efficient utilization of healthcare resources. The platform leverages existing consumer devices, reducing the need for additional medical hardware, lowering electronic waste, and minimizing resource consumption.
While traditional methods for assessing nasal airway patency and upper airway pathologies — Rhinomanometry (RMM) and Acoustic Rhinometry (AR) — have long been considered the clinical gold standards, they are increasingly hindered by significant limitations, including hardware dependency, methodological constraints (1,2), and challenges regarding environmental, social, and governance (ESG) criteria related to resource utilization and sustainability (3). The Odiosoft-Rhino (OR) ecosystem, originally developed by Dr. Erdal Seren in 2005 (4), offers a non-invasive, portable, and sustainable diagnostic solution that transcends the static, hardware-centric nature of conventional methods, supported by extensive multi-center clinical validation over the years (5,6,7,8).
RMM and AR are established techniques that evaluate the nasal airway at the functional (airway resistance) and anatomical (cross-sectional area) levels, respectively (9). However, both systems are constrained by a limited output of numerical parameters, rendering them less suitable for the high-dimensional data analysis required by modern artificial intelligence (AI) workflows (10). The OR ecosystem functions by converting the respiratory sounds generated during inspiration and expiration within the nasal passages into digital biosignals for detailed analysis (11,12). Studies demonstrate that the 2000–4000 Hz and 4000–6000 Hz frequency bands in OR show significant correlations with conventional RMM, AR (Minimal Cross-Sectional Area – MCA), and subjective VAS scores for obstruction (13,14). Recent research on AI-based respiratory sound analysis indicates that breath sound patterns can be utilized for disease classification with high diagnostic accuracy when processed through machine learning models (15,16,17), positioning OR as an 'AI-ready' digital health platform and a transformative tool in modern otorhinolaryngology.
Overview: Odiosoft-Rhino (OR) and Respiro (OS-Respiro) represent a paradigm shift in functional nasal diagnostics. By moving away from cumbersome, stationary legacy equipment, our solution provides a web-based, AI-driven platform that delivers high-precision nasal airway analysis using standard audio-processing technology.
Odiosoft serves as an ideal diagnostic precursor by providing functional nasal and upper airway analysis before advanced sleep assessment. By integrating our technology with sleep-monitoring platforms, healthcare providers can establish an end-to-end clinical pathway that correlates anatomical root causes with physiological sleep outcomes, ultimately enabling more personalized and efficient patient care.
The clinical acceptance of a diagnostic method depends on the degree to which it corresponds with existing gold standard methods. Although Odiosoft-Rhino (OR) operates on the principle of nasal sound analysis, it demonstrates a high level of agreement with the physiological and anatomical parameters measured by Rhinomanometry (RMM) and Acoustic Rhinometry (AR).
Studies conducted by Tahamiler and colleagues have shown that the spectral analysis results obtained through OR are statistically significantly correlated with rhinomanometric resistance values [1,2,3,4]. In patients with septal deviation, both methods detect reductions in airflow with similar sensitivity [5].
Anatomical narrowings within the nasal passage manifest as both reductions in cross-sectional area measured by AR and alterations in the sound frequency spectrum detected by OR [3,6-10]. In postoperative assessments following septorhinoplasty, both methods document surgical success in a parallel manner [11-15].
OR has emerged as the method demonstrating the most consistent correlation with patients' Visual Analog Scale (VAS) scores [1,2,3,5,7,12]. This ability of OR to serve as a bridge between the patient's perception and objective device measurements makes it a more reliable functional guide in clinical practice.
| Evaluation Metric | Correlation with OR | Clinical Significance |
|---|---|---|
| Rhinomanometry (RMM) | High | Validates OR's accuracy in assessing nasal resistance. |
| Acoustic Rhinometry (AR) | Strong | Confirms OR's precision in detecting anatomical narrowing. |
| Patient VAS Scores | Superior | OR correlates more strongly with subjective nasal obstruction. |
| Endoscopic Findings | Superior | OR maps obstruction sites often missed by static tests. |
Odiosoft-Rhino demonstrates a high degree of correlation with current gold standard methods. With its hardware independence, patient comfort, and sensitivity to functional data, OR provides clinicians with complementary and supportive information - establishing itself as a validated and clinically valuable tool.
Platform-Independent Architecture: Odiosoft-Rhino is designed as a Web-based Software-as-a-Service (SaaS) platform, eliminating the need for specialized, costly hardware. By utilizing standard high-fidelity microphones integrated into PCs, tablets, or smartphones, our technology ensures clinical-grade diagnostics are accessible, portable, and scalable across all healthcare settings.
Odiosoft is built with an API-first approach, allowing seamless integration into existing digital ecosystems (e.g., Electronic Medical Records or specialized sleep-tracking software), enabling Real-time Remote Monitoring and automated generation of diagnostic reports.
As healthcare increasingly shifts toward personalized and data-driven care, there is a growing need for diagnostic solutions that bridge the gap between anatomical findings and real-world functional outcomes. Odiosoft addresses this need by providing AI-powered analysis of nasal and upper-airway acoustic signals.
In line with European Union carbon reduction and ESG (Environmental, Social, and Governance) goals within the healthcare sector, digital health solutions are gaining increasing prominence. Healthcare systems are responsible for roughly 4-5% of global carbon emissions, with medical device manufacturing, consumables, and patient transit making up substantial components of this footprint (1).
Conversely, Odiosoft-Rhino (OR) operates over current smartphones, tablets, and PCs due to its software-centric architecture. Because it requires no extra hardware, utilizes zero medical consumables, and yields remote evaluation avenues, its environmental impact is expected to remain significantly lower (2).
The Surgical Simulation module is an AI-powered tool that simulates the effects of various nasal surgical procedures on nasal airflow, using inspiratory and expiratory band data derived from Odiosoft-Rhino nasal airway analysis. This module allows surgeons to pre-operatively evaluate the functional impact of planned interventions.
The system can simulate over 30 surgical procedures (septoplasty, turbinoplasty, lateral osteotomy, hump resection, nasal valve reconstruction, spreader graft, osteotomy types, etc.). The user selects the procedures to be performed, and the system calculates airflow parameters before and after each procedure. Custom procedures can also be added; the AI automatically estimates the narrowing reduction factor based on the procedure description.
After simulation, the following parameters are calculated before and after: airflow reduction ratio, nasal obstruction score, posterior nasal obstruction ratio, estimated SpO2 drop, OSAS risk ratio, and nasal valve collapse ratio. Each procedure step is evaluated separately for the right and left nasal cavities.
AI-powered trend analysis evaluates the simulation results, providing an overall improvement summary, the parameters that improved most, the parameters that still need attention, and clinical significance. The clinical recommendations section lists surgical actions to take and to avoid in order to maximize nasal airflow.
Patient nasal examination photos (frontal, right lateral, left lateral, and nostril views) can be uploaded, and AI analysis detects anatomical findings from these images (septal deviation, turbinate hypertrophy, nasal valve collapse, polyps, inflammation, etc.) and correlates them with airflow parameters. This supports surgical planning with anatomical findings.
The Face Analysis module is an AI-powered tool that analyzes patient photos (frontal, right lateral, left lateral, and nasal base views) to measure facial aesthetic parameters and nasal functional data. This module provides surgeons with a comprehensive tool to plan both functional and aesthetic outcomes together.
The system calculates the following asymmetry indices: orbital asymmetry (eye region), zygomatic asymmetry (cheekbone region), nasal asymmetry (nasal deviation from midline), mandibular asymmetry (jaw region), and overall asymmetry (0-100, where 0 = perfectly symmetric, 100 = severely asymmetric).
Nasal aesthetic angles and ratios: nasolabial angle (normal: 90-120°), nasofrontal angle (normal: 115-130°), nasofacial angle (normal: 30-40°), nasomental angle (normal: 120-132°), mentocervical angle (normal: 80-95°), chin projection score (0-100), golden ratio compliance score (0-100), alar width ratio, nasal base width ratio, nasal length ratio, nasal projection ratio, tip rotation angle, and dorsal contour assessment (dorsal curvature, dorsal height ratio, dorsal contour irregularity).
AI-powered trend analysis evaluates facial symmetry, nasal aesthetic parameters, facial angles, and golden ratio compliance. If the patient's nasal airflow data is available, the AI also analyzes the correlation between nasal aesthetic parameters and nasal airflow parameters.
The clinical recommendations section evaluates nasal aesthetic parameters and nasal airflow parameters together, recommending procedures that optimize airflow while improving facial aesthetics. For each recommendation, the current situation and the expected outcome of the proposed intervention are explained.
The following international research studies have been conducted by me and other researchers using my self-developed software applications, Odiosoft-Rhino: Nasal Sound Analysis (OR)