Periodic Reporting for period 3 - HosmartAI (Hospital Smart development based on AI)
Período documentado: 2023-04-01 hasta 2024-05-31
The main objective of HosmartAI is to promote an effective and efficient healthcare system transformation, by the use of AI technological developments and robotics. In order to achieve this transformation HosmartAI introduced an AI platform that allows for core facilities to be shared and linked composing smart services for healthcare professionals, patients, information system managers, and health organisation administrations.
HosmartAI created a common open Integration Platform with the necessary tools to facilitate and measure the benefits of integrating digital technologies (robotics and AI) in the healthcare system. A central hub will offer multi-faceted lasting functionalities (Marketplace, Co-creation space, Benchmarking) to healthcare stakeholders, combined with a collection of methods, tools and solutions to integrate and deploy AI-enabled solutions.
Eight Large-Scale Pilots will implement and evaluate improvements in medical diagnosis, surgical interventions, prevention and treatment of diseases, and support for rehabilitation and long-term care in several hospital and care settings. The project targets different medical aspects or manifestations such as Cancer (Pilot #1, #2 and #8); Gastrointestinal disorders (Pilot #1); Cardiovascular diseases (Pilot #1, #4, #5 and #7); Thoracic Disorders (Pilot #5); Neurological diseases (Pilot #3); Elderly Care and Neuropsychological Rehabilitation (Pilot #6); Fetal Growth Restriction and Prematurity (Pilot #1).
To ensure a user-centred approach, harmonization in the process (e.g. regarding ethical aspects, standardization, and robustness both from a technical and social and healthcare perspective), the living lab methodology was used. HosmartAI has identified appropriate KPIs that measure efficiency without undermining access or quality of care. Liaison and co-operation activities with relevant stakeholders and open calls will enable ecosystem building and industrial clustering.
HosmartAI brings together a consortium of leading organizations (3 large enterprises, 8 SMEs, 5 hospitals, 4 universities, 2 research centres and 2 associations) along with several more committed organizations. Two Open Calls were executed to further engage third parties.
HosmartAI created a common open Integration Platform with the necessary tools to facilitate and measure the benefits of integrating digital technologies (robotics and AI) in the healthcare system. A central hub will offer multi-faceted lasting functionalities (Marketplace, Co-creation space, Benchmarking) to healthcare stakeholders, combined with a collection of methods, tools and solutions to integrate and deploy AI-enabled solutions.
Eight Large-Scale Pilots will implement and evaluate improvements in medical diagnosis, surgical interventions, prevention and treatment of diseases, and support for rehabilitation and long-term care in several hospital and care settings. The project targets different medical aspects or manifestations such as Cancer (Pilot #1, #2 and #8); Gastrointestinal disorders (Pilot #1); Cardiovascular diseases (Pilot #1, #4, #5 and #7); Thoracic Disorders (Pilot #5); Neurological diseases (Pilot #3); Elderly Care and Neuropsychological Rehabilitation (Pilot #6); Fetal Growth Restriction and Prematurity (Pilot #1).
To ensure a user-centred approach, harmonization in the process (e.g. regarding ethical aspects, standardization, and robustness both from a technical and social and healthcare perspective), the living lab methodology was used. HosmartAI has identified appropriate KPIs that measure efficiency without undermining access or quality of care. Liaison and co-operation activities with relevant stakeholders and open calls will enable ecosystem building and industrial clustering.
HosmartAI brings together a consortium of leading organizations (3 large enterprises, 8 SMEs, 5 hospitals, 4 universities, 2 research centres and 2 associations) along with several more committed organizations. Two Open Calls were executed to further engage third parties.
After running for 40 months, HosmartAI has concluded its activities in May 2024.
Initially, relevant background knowledge has been analysed, user and technical requirements have been detailed, the conceptual architecture has been verified and KPIs have been defined. A living labs methodology has been applied and 4 sprints have been executed to co-create and refine the results.
Two versions of key pillars (on data models and semantic interoperability, AI tools, benchmarking framework, security/data protection privacy/traceability/governance, data sharing and blockchain) have been developed and used in the project platform and pilots.
Two versions of a set of AI-based tools to address specific healthcare challenges (on diagnosis revolution, logistic improvement, treatment improvement, surgical support, assistive care, personalised treatment) has been developed and used in the project pilots.
The reference architecture of the HosmartAI platform has been specified and three versions of the platform has been developed.
The 8 large scale HosmartAI pilots have been specified and their protocols have been approved. Then they were deployed, executed and evaluated. The pilots are focusing on applying AI and Robotics across their range of functions from screening and prevention (Pilot #6) and diagnosis, treatment and surgical support (Pilots #1, #3, #4, #5, #7 and #8) to organizational aspects and logistics in hospitals (Pilot #2).
Early in the project, plans for dissemination, communication and ecosystem building was prepared. Several activities followed and HosmartAI has built a large followers base. The first stakeholder’s event was held in 2022 and HosmartAI also had a large presence in Medica 2023. Standardization activities included participation in working groups and providing recommendations. Training and certification programs were prepared, and several professionals completed them. The datasets used in the various HosmartAI activities were identified and three versions of the data management plan were reported. Two Open Calls were executed (one for technology and one for pilots), providing 4 more technology results and 4 more pilots.
21 exploitable results have been developed in HosmartAI. 10 key exploitable results were identified, and complete business plans and market analyses were developed for them.
HosmartAI has studied applicable laws and regulations as well as relevant ethical and social norms, and ensured its compliance with them, developing a compliance framework and executing an impact assessment. Required ethic reports have been prepared.
Initially, relevant background knowledge has been analysed, user and technical requirements have been detailed, the conceptual architecture has been verified and KPIs have been defined. A living labs methodology has been applied and 4 sprints have been executed to co-create and refine the results.
Two versions of key pillars (on data models and semantic interoperability, AI tools, benchmarking framework, security/data protection privacy/traceability/governance, data sharing and blockchain) have been developed and used in the project platform and pilots.
Two versions of a set of AI-based tools to address specific healthcare challenges (on diagnosis revolution, logistic improvement, treatment improvement, surgical support, assistive care, personalised treatment) has been developed and used in the project pilots.
The reference architecture of the HosmartAI platform has been specified and three versions of the platform has been developed.
The 8 large scale HosmartAI pilots have been specified and their protocols have been approved. Then they were deployed, executed and evaluated. The pilots are focusing on applying AI and Robotics across their range of functions from screening and prevention (Pilot #6) and diagnosis, treatment and surgical support (Pilots #1, #3, #4, #5, #7 and #8) to organizational aspects and logistics in hospitals (Pilot #2).
Early in the project, plans for dissemination, communication and ecosystem building was prepared. Several activities followed and HosmartAI has built a large followers base. The first stakeholder’s event was held in 2022 and HosmartAI also had a large presence in Medica 2023. Standardization activities included participation in working groups and providing recommendations. Training and certification programs were prepared, and several professionals completed them. The datasets used in the various HosmartAI activities were identified and three versions of the data management plan were reported. Two Open Calls were executed (one for technology and one for pilots), providing 4 more technology results and 4 more pilots.
21 exploitable results have been developed in HosmartAI. 10 key exploitable results were identified, and complete business plans and market analyses were developed for them.
HosmartAI has studied applicable laws and regulations as well as relevant ethical and social norms, and ensured its compliance with them, developing a compliance framework and executing an impact assessment. Required ethic reports have been prepared.
HosmartAI has achieved progress beyond the state of the art for the following exploitable results:
- Platform facilitating development of AI-powered solutions in healthcare
- AI-based pre-term birth prediction app with explainable ensemble eodels
- AI-based obstructive coronary artery disease prediction with explainable ensemble models
- AI-based decision support application that predicts adverse effects on cases of fetal growth restriction
- Video capsule endoscopy abnormality detector
- Device-independent AI-powered echocardiographic assessment of left ventricle function
- Automatic appointment scheduling for radiotherapy
- Home-automation system applied for smart rehabilitation of neurological patients with motor impairments in a hard-to-reach hospital
- New small dimension, flush mounted, Ultra-Wide Band Radar Device detects presence and people fall in indoor environment
- Smartphone camera-based body tracking for assessment of motor function
- REST API cloud interface based on the open standard KNX IoT 3rd party API
- Real-time analysis and visualization of electrophysiology mapping data during catheter ablation procedures using Machine Learning
- Socially assistive humanoid robot in grand round routine
- Socially assistive humanoid robot to support nurses during hospitalization
- Synthetic images generator for data augmentation
- Embedded blockchain on IoT
- Virtual assistant and coach for detection, prevention, therapeutic care and active ageing of older adults
- Smart cathlab assistant / Customer innovation lab
- Digital health research platform for cancer
- AI-based tumour image segmentation tool
- Tool for genetic analysis of tumour data
- Platform facilitating development of AI-powered solutions in healthcare
- AI-based pre-term birth prediction app with explainable ensemble eodels
- AI-based obstructive coronary artery disease prediction with explainable ensemble models
- AI-based decision support application that predicts adverse effects on cases of fetal growth restriction
- Video capsule endoscopy abnormality detector
- Device-independent AI-powered echocardiographic assessment of left ventricle function
- Automatic appointment scheduling for radiotherapy
- Home-automation system applied for smart rehabilitation of neurological patients with motor impairments in a hard-to-reach hospital
- New small dimension, flush mounted, Ultra-Wide Band Radar Device detects presence and people fall in indoor environment
- Smartphone camera-based body tracking for assessment of motor function
- REST API cloud interface based on the open standard KNX IoT 3rd party API
- Real-time analysis and visualization of electrophysiology mapping data during catheter ablation procedures using Machine Learning
- Socially assistive humanoid robot in grand round routine
- Socially assistive humanoid robot to support nurses during hospitalization
- Synthetic images generator for data augmentation
- Embedded blockchain on IoT
- Virtual assistant and coach for detection, prevention, therapeutic care and active ageing of older adults
- Smart cathlab assistant / Customer innovation lab
- Digital health research platform for cancer
- AI-based tumour image segmentation tool
- Tool for genetic analysis of tumour data