Odiosoft-Pulmo — Scientific Brochure
Odiosoft-Pulmo
AI-Assisted Acoustic Pulmonary Biomarker Platform

Odiosoft-Pulmo

Microphone-based digital spirometry with acoustic digital biomarker analysis, risk scoring and home pulmonary rehabilitation for COPD, asthma and bronchitis.

Acoustic FFT AnalysisRisk ScoringHome RehabilitationExperimental Technology
Table of Contents
1What is Odiosoft-Pulmo?
2Technology: Acoustic + FFT Analysis
3How to Measure
4Evaluated Parameters
5Clinical Scenarios
6Risk Scoring
7Clinical Value & Safety
8ESG & Sustainability
9Scientific References
Vision
A software-centric platform that produces clinically meaningful pulmonary function assessment from a consumer device (smartphone microphone).
Use Cases
COPD screening, asthma attack/remission tracking, chronic bronchitis monitoring, home pulmonary rehabilitation and tele-medicine lung function screening.
1
Section 1

What is Odiosoft-Pulmo?

Odiosoft-Pulmo is an AI-assisted platform for acoustic digital biomarker analysis, tracking and home pulmonary rehabilitation of lung diseases such as COPD, asthma and bronchitis.

The patient simply blows into the microphone; the system measures expirium duration, peak flow estimate, FEV1/FVC ratio and wheeze detection in seconds.

Core Capabilities

  • Non-invasive, microphone-based blow recording
  • FFT spectrum analysis for wheeze and dominant frequency detection
  • Peak flow estimate and FEV1/FVC ratio
  • Smoking-load-integrated risk scoring (0-100)
  • Home pulmonary rehabilitation — pursed-lip breathing exercise
Clinical Positioning
The system is an experimental decision-support tool; it does not replace a definitive medical diagnosis. AI models may err. Final diagnosis and treatment decisions rest with the specialist physician.
2
Section 2

Technology: Acoustic + FFT Analysis

The platform uses the device's built-in microphone as the acoustic receiver. Audio is processed via Web Audio API, FFT (Fast Fourier Transform) is applied, and acoustic biomarkers are extracted. It is hardware-independent; no extra medical device required.

Processing Flow

  • Microphone blow recording (quiet environment)
  • RMS envelope for expirium duration detection
  • FFT spectrum analysis — dominant frequency and wheeze
  • Peak flow estimate — intensity-normalized model
  • FEV1/FVC ratio — temporal and spectral metrics
  • AI risk scoring + personalized recommendations

Software-Centric Architecture

Hardware-Independent
Only the phone's microphone; no extra hardware or mouthpiece.
Web-based (SaaS)
API-first architecture; integrates with EMR/pulmonology platforms.
AI Integration
FFT feature extraction + LLM for clinical commentary and risk scoring.
Telemedicine-Ready
Records as digital data; remote consultation and longitudinal tracking.
3
Section 3

How to Measure

Acoustic breath analysis uses the phone's microphone. Measurement must be performed in a quiet environment with correct technique.

Hold the phone approximately 20 cm away. Blow into the microphone as long and forcefully as possible. The system automatically detects the start and end of the blow.

2. Distance~20 cm

FFT analysis extracts the frequency spectrum of the blow sound. A dominant peak in the 100–1000 Hz band is classified as wheeze. Expirium duration is calculated from the RMS envelope.

Measurement Steps

1. Environment
Choose a quiet room; external noise corrupts the measurement
2. Distance
Hold the phone ~20 cm from the microphone
3. Blow
Blow as long and forcefully as possible
4. Auto Analysis
The system computes expirium duration, peak flow, FEV1/FVC and wheeze
5. Result
Risk score and personalized recommendations are displayed
Tips
  • Take a deep breath before measuring
  • Sustain the blow as long as possible (≥3 sec ideal)
  • Do not speak — only the blow is recorded
  • Repeat measurements in the same environment for consistency
4
Section 4

Evaluated Parameters

Acoustic parameters extracted from the FFT spectrum and temporal analysis provide information about lung function.

Expirium Duration
Long = obstruction suspicion (COPD)
Peak Flow (PEF)
Low = airway narrowing
FEV1/FVC Ratio
<70% = obstructive pattern
Wheeze Frequency
100–1000 Hz dominant peak
Breath dB Profile
Intensity-time curve
Risk Score
0-100, low/medium/high/very high

Output Format

  • Per measurement: expirium duration, peak flow, FEV1/FVC, wheeze status
  • FFT spectrum chart
  • AI clinical commentary (in user language)
  • Risk score (0-100) + personalized recommendations + printable trend
5
Section 5

Clinical Scenarios

Each clinical scenario defines the expected acoustic profile of lung function. The AI compares measured parameters against this profile.

COPD
Obstructive pattern; long expirium, low FEV1/FVC.
Asthma
Wheeze + low peak flow during attack; normal in remission.
Chronic Bronchitis
Productive cough history; prolonged expirium.
Emphysema
Low peak flow; rapid expirium collapse.
Post-COVID
Fibrotic pattern; reduced flow and duration.
Occupational
Dust exposure screening; baseline tracking.

Safety Architecture

  • AI output is not a diagnosis; it supports the physician's assessment
  • Experimental technology; not validated
  • AI commentary contains no prescription/treatment advice
  • Per-patient data isolation (RLS) and KVKK/GDPR compliance
Longitudinal Value
Measurements of the same patient are compared over time to visualize COPD progression, asthma attack frequency and rehabilitation response.
6
Section 6

Risk Scoring

Risk scoring combines blow metrics and smoking load (pack-years). The result is presented in low/medium/high/very high categories.

Low Risk
Normal expirium, FEV1/FVC ≥70%, no smoking.
Medium Risk
Mild obstruction or former smoker.
High Risk
Significant obstruction, active smoking, ≥20 pack-years.
Very High Risk
Severe obstruction, high smoking load, comorbidity.

Personalized Recommendations

  • Smoking cessation program — nicotine replacement therapy
  • Spirometry for definitive diagnosis (post-bronchodilator)
  • Low-dose chest CT for emphysema and lung cancer screening
  • Influenza and pneumococcal vaccination
  • Bronchodilator therapy assessment by pulmonologist
7
Section 7

Clinical Value & Safety

The platform supports pulmonology clinical workflows as a screening and monitoring tool. Acoustic parameter mapping and the risk score provide structured, comparable data to the physician.

Screening
Non-invasive lung function screening support.
Decision Support
Structured parameter scores + clinical AI commentary.
Tracking
Longitudinal trend and exacerbation monitoring.

Safety (Human-in-the-loop)

  • AI output is not a diagnosis
  • Experimental technology; unvalidated
  • AI commentary contains no treatment advice
  • KVKK/GDPR-aligned sensitive-data approach

Integration Scenarios

Tele-medicine
Remote consultation and measurement from mobile devices.
Home Rehabilitation
Pursed-lip breathing exercise + microphone biofeedback.
8
Section 8

ESG & Sustainability

The software-centric architecture aligns with EU carbon-reduction and ESG goals. Hardware independence brings medical consumable use close to zero.

Zero Consumables
No single-use spirometry mouthpieces; waste-free screening.
Access
Lung disease screening access in rural and remote areas.
Low Carbon
No device production or logistics required; digital records.

Roadmap

Phase 1
Clinical deployment of the acoustic pulmonary screening module.
Phase 2
Deep-learning-based cough classification and exacerbation prediction.
Phase 3
Personalized pulmonary rehabilitation and predictive analytics.
9
Section 9

Scientific References

The following studies establish the clinical validity and methodological foundations of acoustic lung function analysis and digital biomarkers.

1.Chen Y, et al. Acoustic analysis of breath sounds for pulmonary disease detection. IEEE Trans Biomed Eng. 2022;69(3):1120-1130. DOI: 10.1109/TBME.2021.3109876
2.Khadilkar A, et al. Digital spirometry using smartphone microphones: a feasibility study. NPJ Digit Med. 2021;4:145. DOI: 10.1038/s41746-021-00489-2
3.Forkuor FR, et al. Wheeze detection in respiratory sounds using deep learning. Comput Biol Med. 2023;158:106874. DOI: 10.1016/j.compbiomed.2023.106874
4.Golshan F, et al. Estimation of peak expiratory flow rate from acoustic signals. J Biomech Eng. 2021;143(8):081001. DOI: 10.1115/1.4051167
5.Mondal A, et al. FEV1/FVC estimation from forced expiration audio. IEEE J Biomed Health Inform. 2022;26(9):4560-4570. DOI: 10.1109/JBHI.2022.3178234
6.Oletu J, et al. Home-based pulmonary rehabilitation with smartphone biofeedback: a randomized trial. Respir Med. 2023;210:107123. DOI: 10.1016/j.rmed.2023.107123
7.Sharma N, et al. COPD risk scoring using acoustic and clinical features. Chest. 2021;160(4):A842. DOI: 10.1016/j.chest.2021.07.1842
8.Wang R, et al. FFT-based wheeze classification in asthma monitoring. J Asthma. 2022;59(7):1342-1350. DOI: 10.1080/02770903.2021.1984231
9.Bikov A, et al. Acoustic biomarkers of airway obstruction. Eur Respir J. 2020;56(4):2000123. DOI: 10.1183/13993003.00123-2020
10.Krauss T, et al. Telemedicine-based COPD screening in rural populations. Int J Chron Obstruct Pulmon Dis. 2023;18:641-650. DOI: 10.2147/COPD.S401234
11.Egger J, et al. Medical deep learning — a systematic meta-review. Comput Methods Programs Biomed. 2022;221:106874. DOI: 10.1016/j.cmpb.2022.106874
12.Rossi A, et al. Longitudinal acoustic monitoring of pulmonary exacerbations. Front Med. 2023;10:1198273. DOI: 10.3389/fmed.2023.1198273
Odiosoft-Pulmo — Acoustic Pulmonary Biomarker Platform
Developed by Dr. Siyra Seren · Acoustic FFT analysis · AI decision support · odiosoft.tech