AI-PROGNOSIS participated in the IEEE HealthCom 2025 Conference, held 21–23 October 2025 and hosted by Khalifa University in Abu Dhabi, United Arab Emirates. The event brought together experts in digital health, artificial intelligence, and biomedical technologies to explore innovative solutions for non-communicable diseases.

The AI-PROGNOSIS teams from the Signal Processing & Biomedical Technology Unit (AUTH) and from the Centre for Research and Technology Hellas contributed through scientific presentations, an award-winning paper, a panel discussion on digital health collaboration, and the co-organisation of a thematic workshop.

Capturing effects of medication change in Parkinson's with wearable data

Apostolos Moustaklis from the Signal Processing & Biomedical Technology Unit, Aristotle University of Thessaloniki (AUTH) presented the paper “On capturing effects of medication change in Parkinson’s disease with wrist accelerometry-based digital biomarkers”.

By analysing data from the Verily Study Watch cohort of the Parkinson’s Progression Markers Initiative (PPMI), the study showed that diurnal acceleration metrics linked to slowness of movement, captured using a single wrist wearable, can indicate the effects of increased levodopa equivalent daily dose in people with Parkinson’s disease.

Assessing rest tremor in daily life using wearable data (Best workshop paper award)

The paper “A Deep Learning Approach for Parkinsonian Tremor Assessment Using Wearables” by Thomas-Theocharis Chatzis and the team at the Centre for Research & Technology Hellas (CERTH) received the Best Workshop Paper Award.

The work analysed data from the Verily Study Watch (PPMI) and the Michael J. Fox Levodopa Response Study, proposing a deep learning–based framework for detecting and quantifying tremor in Parkinson’s disease using accelerometry data from a single wrist wearable.

Panel Discussion

Project Coordinator Leontios Hadjileontiadis, Scientific and Technical Manager Stelios Hadjidimitriou, and Vasileios Charisis from AUTH, together with Kosmas Dimitropoulos from CERTH and Prof. Mohdamed Seghier from Khalifa University, took part in the panel “Digital Bridges for Better Health: AI Innovation from Europe to the Middle East and North Africa (MENA)”.

The discussion explored:

• Integration challenges of AI tools with existing healthcare IT systems

• Collaboration between stakeholders to ensure clinical relevance and usability

• The importance of patient engagement and digital literacy

• Socio-cultural factors influencing adoption of AI in different regions

• Regulatory challenges and opportunities for harmonisation between Europe and MENA

During the discussion, the AI-PROGNOSIS project and the AI-enabled digital health tools it develops for assisting Parkinson's disease screening and care were highlighted as a case study of AI solutions with the potential of improving health outcomes of persons with non-communicable diseases.

Workshop Chairing

Stelios Hadjidimitriou (AUTH) co-chaired the workshop “AI-enabled Digital Health Tools for Non-Communicable Diseases: From Concepts to Impact”, organised jointly with the iPROLEPSIS Horizon Europe project.

The session included research contributions on Parkinson’s disease, psoriatic arthritis, multiple sclerosis, and post-stroke rehabilitation.

Two of the presented papers originated from AI-PROGNOSIS research:

• Chatzis et al., A Deep Learning Approach for Parkinsonian Tremor Assessment Using Wearables

• Moustaklis et al., On capturing effects of medication change in Parkinson’s disease with wrist accelerometry-based digital biomarkers

The AI-PROGNOSIS and iPROLEPSIS teams from AUTH participated jointly at IEEE HealthCom 2025, showcasing the projects’ shared efforts in developing AI-enabled solutions to improve the understanding, detection, and management of chronic diseases.