Periodic Reporting for period 2 - Digitising Hospitals For Better Patient Care (First modular, interconnected, data- and AI-driven software platform for the digitisation of patient care in the healthcare industry)
Période du rapport: 2024-07-01 au 2025-06-30
The healthcare industry is mired in technological backwardness where digital transformation is highly needed yet enormously underdeveloped. As per a recent McKinsey study, 83.1% of clinic executives rate the digitisation maturity and quality within their clinics as low to medium. Operational inefficiencies are rife, and the high administrative burden exacerbates time pressure during treatments (averaging 8 minutes per patient), increasing the risk of treatment errors up to 25% worldwide. Financial ramifications are also significant, with billing errors at a rate of 20%, amounting to avoidable costs of approximately €17 billion annually.
Avelios Medical aims to solve this problem by offering a trailblazing, holistic, modular plug & play clinic platform enabling comprehensive digitisation across all hospital departments. By interlinking intelligent input fields with medical libraries and databases, it achieves automatic data structuration to generate up to 2,000 data points per patient and treatment. This functions as a foundation for the self-learning AI-powered diagnostic support and workflow engine of Avelios Medical, facilitates actionable medical insights and research possibilities. Our software looked for saving up to 10 minutes on administrative tasks per patient treatment and to drastically improve the quality of care, ensuring that all data processing activities comply with GDPR, safeguarding patient privacy and maintaining high ethical standards. Key features include automatic generation of doctor’s letters, billing with diagnosis coding, and complete digitisation of patient data handling.
The main impacts of the project are:
• 20% less time spent on administrative tasks.
• Up to 80% fewer treatment errors.
• Hospitals save up to €650 per patient; (in global terms, the medical billing error reduction saves €2.8 billion annually).
• Enhanced treatment transparency and efficiency, longer patient interaction times (additional 10 minutes per treatment).
• First-time structuring and utilization of up to 2,000 data points per treatment for meaningful AI-driven medical research.
• Supports EU digital sovereignty and legislative mandates for digital health records.
The integration of social sciences and humanities into the project is fundamental. The interdisciplinary approach ensures that regulatory, ethical, and user-centric perspectives are considered, bridging the gap between technological innovation and humanistic care.
Avelios Medical aims to solve this problem by offering a trailblazing, holistic, modular plug & play clinic platform enabling comprehensive digitisation across all hospital departments. By interlinking intelligent input fields with medical libraries and databases, it achieves automatic data structuration to generate up to 2,000 data points per patient and treatment. This functions as a foundation for the self-learning AI-powered diagnostic support and workflow engine of Avelios Medical, facilitates actionable medical insights and research possibilities. Our software looked for saving up to 10 minutes on administrative tasks per patient treatment and to drastically improve the quality of care, ensuring that all data processing activities comply with GDPR, safeguarding patient privacy and maintaining high ethical standards. Key features include automatic generation of doctor’s letters, billing with diagnosis coding, and complete digitisation of patient data handling.
The main impacts of the project are:
• 20% less time spent on administrative tasks.
• Up to 80% fewer treatment errors.
• Hospitals save up to €650 per patient; (in global terms, the medical billing error reduction saves €2.8 billion annually).
• Enhanced treatment transparency and efficiency, longer patient interaction times (additional 10 minutes per treatment).
• First-time structuring and utilization of up to 2,000 data points per treatment for meaningful AI-driven medical research.
• Supports EU digital sovereignty and legislative mandates for digital health records.
The integration of social sciences and humanities into the project is fundamental. The interdisciplinary approach ensures that regulatory, ethical, and user-centric perspectives are considered, bridging the gap between technological innovation and humanistic care.
During the project, we have performed the following activities related to the product development:
• Workflow Engine Development and Integration
o Successfully integrated the workflow engine into the platform
o Developed AI-based workflow recommendation system for event-triggering via microservices, thus fully mapping and automating treatment processes
• Customisation of specific modules to individual clinic needs
• Workflow Interoperability (ensuring clinical and laboratory orders are interoperable with core treatment documentation)
• Interface Development
o Built standard interfaces for connecting existing HIS systems like SAP IS-H, Daedalus Orbis.
o Developed over 30 distinct interfaces
o Database construction
o Ontological Mapping and Workflow Development
o Platform rollout in different clinic’s departments and trials
o Software implemented across pilot hospital department
Our main achievements reached during the project include:
• A database of 18,263,800 data points was built based on 91,319 completed treatments (trials to validate data points conducted in at least three different clinics).
• Over 250 data value sets were developed that are mapped to SNOMED CT, LOINC and HPO as much as possible.
• AI-based workflow recommendation systems were developed using soft decision trees.
• A total of 6,268 cases were examined and a 100% success rate of making automatic treatment sequence suggestion was achieved. The rate of accepting automatic treatment sequence was 79%.
• Clinical staff feedback was incorporated to gathered robustness requirements.
• A support a training method for federated machine learning was developed as part of our AI-based diagnostic support ensuring compliance with GDPR and with the highest standards regarding ethic requirements.
• Workflow Engine Development and Integration
o Successfully integrated the workflow engine into the platform
o Developed AI-based workflow recommendation system for event-triggering via microservices, thus fully mapping and automating treatment processes
• Customisation of specific modules to individual clinic needs
• Workflow Interoperability (ensuring clinical and laboratory orders are interoperable with core treatment documentation)
• Interface Development
o Built standard interfaces for connecting existing HIS systems like SAP IS-H, Daedalus Orbis.
o Developed over 30 distinct interfaces
o Database construction
o Ontological Mapping and Workflow Development
o Platform rollout in different clinic’s departments and trials
o Software implemented across pilot hospital department
Our main achievements reached during the project include:
• A database of 18,263,800 data points was built based on 91,319 completed treatments (trials to validate data points conducted in at least three different clinics).
• Over 250 data value sets were developed that are mapped to SNOMED CT, LOINC and HPO as much as possible.
• AI-based workflow recommendation systems were developed using soft decision trees.
• A total of 6,268 cases were examined and a 100% success rate of making automatic treatment sequence suggestion was achieved. The rate of accepting automatic treatment sequence was 79%.
• Clinical staff feedback was incorporated to gathered robustness requirements.
• A support a training method for federated machine learning was developed as part of our AI-based diagnostic support ensuring compliance with GDPR and with the highest standards regarding ethic requirements.
By providing a structured, interoperable system with clear access to historical patient data and standard operating procedures, we directly contribute to reduce treatment errors - such as incorrect medication administration. This has an important social impact as it directly addresses public health, something essential for the economic development and the social wellbeing. According to WHO, 4 in 10 patients were harmed in primary and outpatient health care and up to 80% of this harm was preventable. Most errors were related to diagnosis, prescription, and the use of medicine, being directly addressed by our solution due to automated and AI-backed decision support.
Therefore, the project enhanced treatment transparency and efficiency, increasing the patient interaction time (additional 10 minutes per treatment). In sum, the Avelios platform enables up to 20% reduction on the time spent on administrative tasks and up to 80% fewer treatment errors, which in terms of economic impact, allows hospitals to save up to €650 per patient. By reducing administrative burdens up to 20% and enhancing patient care, the platform not only improves operational efficiency but also contributes to better patient outcomes. Healthcare providers have reported increased staff satisfaction and reduced burnout, as the platform frees up valuable time for patient interaction and care.
Additionally, our solution supports EU digital sovereignty and legislative mandates for digital health records addressing the EU strategy “A Europe Fit for the Digital Age”. Across Europe, only 19% of advanced medical tasks were automated and the use of supporting AI was only 5% because structured data was required for the development of meaningful AI, yet none of the current solutions were able to generate such.
Therefore, the project enhanced treatment transparency and efficiency, increasing the patient interaction time (additional 10 minutes per treatment). In sum, the Avelios platform enables up to 20% reduction on the time spent on administrative tasks and up to 80% fewer treatment errors, which in terms of economic impact, allows hospitals to save up to €650 per patient. By reducing administrative burdens up to 20% and enhancing patient care, the platform not only improves operational efficiency but also contributes to better patient outcomes. Healthcare providers have reported increased staff satisfaction and reduced burnout, as the platform frees up valuable time for patient interaction and care.
Additionally, our solution supports EU digital sovereignty and legislative mandates for digital health records addressing the EU strategy “A Europe Fit for the Digital Age”. Across Europe, only 19% of advanced medical tasks were automated and the use of supporting AI was only 5% because structured data was required for the development of meaningful AI, yet none of the current solutions were able to generate such.