Revolutionizing respiratory care with AI-driven compliance and data insights

Coach Name

EU Organization

Sosteco

Members

  • Roberto Medina Bujalance
  • Sandra Lohnes
  • Victor González Toledo ( advisor )

US Organization

Sonoma State University

Members

  • Dr. Farid Farhamand

Project Overview

SmartBreathAI is a digital health project designed to modernize respiratory rehabilitation by transforming incentive spirometry into an intelligent, connected therapy tool. The platform provides real-time feedback to patients, monitors adherence, and supports clinicians with data-driven insights to improve recovery and reduce postoperative complications.

Developed entirely with open-source technologies, SmartBreathAI integrates a patient- and clinician-facing interface, a secure backend, and an anonymized data analytics pipeline, all compliant with HIPAA and GDPR standards. The project also produced an open-access GitHub repository, a comprehensive compliance roadmap, and a scientific paper summarizing the methodology and outcomes.

With registered trademarks in both the EU and the USA, SmartBreathAI establishes a strong foundation for future clinical validation and the potential creation of a dedicated startup focused on AI-powered respiratory care.

Methods and approaches

Human-Centered Design Framework

Co-created the patient and clinician interfaces through iterative feedback with healthcare professionals in surgery, pulmonology, and internal medicine departments.

AI and Machine Learning Pipeline

Developed an explainable AI model using anonymized spirometry data for adherence classification and recovery prediction, following reproducible ML workflows.

Secure Data Infrastructure

Implemented a local private server with AES-256 encryption, JWT authentication, and HIPAA/GDPR-compliant data management for privacy and integrity.

Open-Source Development

Built the platform entirely with open-source technologies (React, Node.js, TypeScript, SQLite) and released the repository publicly on GitHub to ensure transparency and reproducibility.

Compliance and Ethical AI Design

Integrated privacy, fairness, and transparency principles into every development stage through a dedicated compliance roadmap and responsible data governance.

Participatory Validation and Knowledge Transfer

Validated the prototype through structured pilot feedback, internal workshops, and documentation in the Knowledge Transfer Handbook to ensure sustainability and reproducibility.

Key Achievements

Developed a functional digital spirometry prototype

Integrated real-time patient feedback, clinician dashboards, and AI-based adherence tracking into a unified secure platform.

Ensured full data protection compliance

Implemented HIPAA and GDPR standards with encrypted data storage, secure authentication, and role-based access control.

Created an open-source repository

Published the SmartBreathAI code, anonymized datasets, and documentation on GitHub to support reproducibility and transparency.

Produced a scientific paper and compliance report

Submitted the research paper “Design, Implementation, and Evaluation of a Digital Incentive Spirometry Application for Respiratory Therapy” and finalized a complete HIPAA/GDPR compliance roadmap.

Strengthened international collaboration

Maintained active cooperation between EU, US, and Canadian partners through mentoring sessions, data exchange, and continuous feedback.

Established the foundation for startup creation

 Registered trademarks in the EU and USA, defining a roadmap toward clinical validation, commercialization, and future company formation.

Impact & Results

Scientific Impact

Advanced the field of digital respiratory monitoring by demonstrating how AI and real-time feedback can improve patient adherence and clinical decision-making.

Industrial Impact

Provided a scalable model for healthcare technology integration using open-source tools, enabling hospitals and startups to adopt cost-effective and compliant digital health systems.

Economic Impact

Showed potential for reducing hospital readmissions and clinician workload, creating measurable efficiency gains and laying the foundation for future startup-driven job creation.

Social Impact

Empowered patients through accessible, user-friendly design and continuous motivation, promoting better self-care and improved recovery outcomes.

Policy and Compliance Impact

Aligned with HIPAA and GDPR frameworks, setting a benchmark for ethical, privacy-preserving AI applications in healthcare and reinforcing trust in digital medical systems.

Publications and Open-Source Contributions

Project web : https://smartbreath.ai/

Research article: “Design, Implementation, and Evaluation of a Digital Incentive Spirometry Application for Respiratory Therapy” (to be published)

GitHub Repository: https://github.com/SmartBreathAI/SmartBreathAI

Workshop materials and stakeholder input summaries (referenced in report)

Future directions

Clinical Validation

Expand pilot testing with partner hospitals to validate SmartBreathAI’s effectiveness in real-world postoperative and rehabilitation settings.

Regulatory Pathway

Initiate steps toward medical device certification and formal clinical trial design in alignment with EU MDR and FDA frameworks.

AI Model Expansion

Refine and retrain the AI algorithms using larger, more diverse datasets and publish enhanced models on Hugging Face for community validation.

Startup Development

Establish the SmartBreathAI startup to support deployment, maintenance, and commercialization, building the foundation for sustainable growth.

EHR Integration

Implement interoperability with hospital Electronic Health Record systems using FHIR standards for seamless clinical data exchange.

International Collaboration

Continue the strong EU–US and EU–Canada partnerships to co-develop new research projects focused on AI ethics, data governance, and health innovation.

Subscribe to our newsletter

Horizon Europe – Grant Agreement number 101092887

Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Union’s Horizon Europe research and innovation programme. Neither the European Union nor the granting authority can be held responsible for them.