SHStelios Hadjidimitriou

Stelios Hadjidimitriou

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Stelios Hadjidimitriou has a total of 169 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Wearable-based sleep and activity tracking detects Parkinson's disease and associated pathological hallmarks

    by Apostolos Moustaklis, Ioannis Gerasimou, Charalampos Sotirakis, Paraskevi Chairta, Monica M. Kurtis, Natalia Del Campo, Tim Feige, Nils Schnalke, Maria-Luisa Almarcha Menargues, Kyriaki Michailidou, et al.

    Parkinson's disease (PD) is associated with sleep and motor function impairment, even in the early stages. Here, we investigated whether passive wearable-based sleep and activity tracking can detect PD and its associated pathological hallmarks, dopaminergic deficit and α-synuclein pathology. Sleep and activity metrics (SAM) were derived from free-living wrist accelerometry data of a U.S. cohort and used to train a machine learning model (SAM-PD) to distinguish persons with PD (n = 133) from healthy controls (n = 26). The model achieved an AUROC of 0.93 and an AUPRC of 0.98, demonstrated excellent test-retest reliability, and generalized without retraining to an independent European cohort (n = 40; 50% PwPD) monitored with a different device (AUROC, 0.95; AUPRC, 0.95). REM sleep duration, walking cadence, and low-level acceleration features were most predictive. Applied to a U.S. at-risk cohort, SAM-PD performed comparably to the resource-intensive Movement Disorder Society (MDS) prodromal PD risk score in detecting dopaminergic deficit (AUROC, 0.86 vs. 0.80; AUPRC, 0.45 vs. 0.46; n = 121; 12.4% prevalence) and α-synuclein pathology (AUROC, 0.76 for both; AUPRC 0.50 vs. 0.42; n = 114; 16.7% prevalence). A combination with the MDS score modestly improved detection. Findings support further evaluation of wearable monitoring for PD screening and biomarker enrichment in larger populations.

  • Open Access

    Is all knowledge, good knowledge?: Multinational Patient Survey on AI-Generated Prognostic Insights and Actionability for Parkinson’s Disease (Preprint)

    by Jamie Linnea Luckhaus, Anna Kharko, Therese Scott Duncan, Charlotte Blease, David M. Lyreskog, Margherita Fabbri, Monica Kurtis, Stelios Hadjidimitriou, John Zaras, Maria Hägglund, et al.

    BACKGROUND Predictive artificial intelligence (AI) is rapidly emerging in Parkinson's disease (PD) management, yet patient perspectives remain underexplored. Actionability is a key ethical consideration in prognostic disclosure. Understanding how people with PD (PwP) value and define actionable AI-generated prognostic information is essential for patient-centered design and implementation. OBJECTIVE This study aims to examine PwP’s preferences for AI-based prognostic information and explore how actionable insights are perceived and valued. METHODS In this convergent mixed-methods study, 149 PwP completed an online survey distributed through international patient networks between April and November 2025. Quantitative data were analysed using descriptive and inferential statistics, and free-text responses underwent qualitative content analysis. RESULTS Most expressed interest in knowing their motor prognosis (71%) and milestones (74%). Nearly half (47%) would only want feedback they can act upon. There was a significant association between education level and interest in knowing which factors may influence one’s prognosis (χ²(2)=6.69, p=.04), and between time since diagnosis and only wanting actionable insights (χ²(2)=7.38, p=.02). The meaning of actionability varied between individuals, including being described as the ability to influence disease progression, facilitating preparation, and as the credibility and personal applicability of insights. For some, actionability was necessary for wanting prognostic information, while for others information was valued regardless. CONCLUSIONS Respondents valued prognostic information beyond traditional intervention-focused definitions of actionability, though substantial individual variation remained. These findings support an individualized approach to communicating prognostic information, whereby healthcare professionals regularly assess and respond to patients' evolving preferences and informational needs throughout the disease course. CLINICALTRIAL N/A

  • Open Access

    Clinical AI is Not (Yet) Trustworthy-But It Could Be

    by Ali Saad, Sofia B Dias, Ghada Alhussein, David Lyreskog, Ioannis Gerasimou, Beatriz Alves, Μaarten de Vos, Ioannis Drivas, John Zaras, Andreas Stergioulas, et al.

    The growing emphasis on trustworthy artificial intelligence (AI) in health care reflects a shift away from models optimized for predictive performance toward governable and auditable systems that can be adopted and sustained in clinical practice. Nonetheless, many clinical AI applications continue to privilege technical performance while underaddressing ethical, regulatory, and societal considerations, leading to concerns around robustness, transparency, and clinical adoption. To address this, governance frameworks such as the Assessment List for Trustworthy Artificial Intelligence (ALTAI) have been proposed to operationalize trust-related requirements across the AI lifecycle. However, evidence on the practical use of these frameworks remains limited. In this Viewpoint, we describe the application of ALTAI as a procedural governance framework within the Horizon Europe AI-PROGNOSIS project, which aims to support Parkinson disease diagnosis and care through predictive models and digital biomarkers derived from everyday devices. The seven ALTAI requirements (ie, human agency and oversight; technical robustness and safety; privacy and data governance; transparency; diversity, nondiscrimination, and fairness; societal and environmental well-being; and accountability) were mapped to key stages of the AI lifecycle within the project, including design and specification, data preparation, model development and validation, user interface and user experience deployment, external prospective validation, and overarching management and workflow. To examine how these requirements were perceived in practice, we conducted a structured internal survey among AI developers and data scientists involved in the AI-PROGNOSIS project (n=10). Participants rated the relevance of the 17 ALTAI subdomains using a three-point prioritization scale. Technical accuracy, data governance, and privacy were consistently rated as highly relevant, whereas societal impact received the lowest prioritization. This pattern reflects a documented tension in AI development, where technical teams tend to deprioritize broader societal concerns under delivery and performance constraints. Nonetheless, this work should be interpreted as a context-specific case study rather than a validation of ALTAI. The small sample size and project-specific setting limit generalizability, and these findings should not be considered as representative of broader clinical AI development. Overall, by making prioritization gaps explicit and embedding multidisciplinary review across lifecycle checkpoints, this case study illustrates how structured governance frameworks can surface implementation tensions and support accountable AI development. While these approaches do not resolve all of the aforementioned challenges, they provide practical guidance for integrating trust-related considerations into clinical AI projects.

  • Open Access

    Drivers of Patients’ Intention To Use Digital Serious Games in Psoriatic Arthritis Care

    by Sofia Balula Dias, Bárbara Ramalho, Jolanda Luime, Ijia Tchetverikov, Laura C. Coates, Ana M. Rodrigues, Theodoros Dimitroulas, Ioannis T. Rekanos, Stelios Hadjidimitriou, Vasileios Charisis, et al.

    Digital serious games (SGs) are emerging tools for chronic disease self‑management, yet their adoption among people with psoriatic arthritis (PsA) remains unknown. This lack of evidence on what drives SG uptake limits the design of interventions patients will actually use. Here we examined whether interaction technology in SGs shapes PsA patients’ enjoyment and intention to use SGs by mediating the effects of core serious and game design elements. Data from a cross‑sectional web‑based survey across four European countries (n=381 PsA patients) were analysed with a hybrid structural equation modelling framework to test the hypothesised pathways. Bootstrapped path coefficients (0.133–0.864; all p<0.001) and adjusted R² (up to 0.863), supported all associations. Interaction technology emerged as the central mediator, linking serious and game elements to enjoyment and, most strongly, to intention to use SGs. Reliability, validity, effect sizes, predictive relevance, and subgroup invariance were also confirmed. Interaction technology drives SG engagement, showing that therapeutic content must be paired with engaging interactive features for personalised PsA self‑management.

  • Open Access

    Dynamic Difficulty Adjustment in Psoriatic Arthritis Serious Games Using Learning-Driven Policies

    by Samuel Gomes, Barbara Ramalho, Joao Caldeira, Filipa Magalhaes, Marta Vicente, Sandra Gama, Maria Mytilinaiou, Theodoros Dimitroulas, Vasileios Charisis, Stelios Hadjidimitriou, et al.

  • EduGenZ: A Gamified Educational Framework for GenZ Using Affective and Immersive Technology

    by Sofia J. Hadjileontiadou, Sofia B. Dias, Anastasios Mikropoulos, Ana Balula, Anna Klothou, Charalambos Sakonidis, Ioannis Dimitrios Kapsidis, Ilias Kitsas, Vasileios Charisis, Stelios Hadjidimitriou, et al.

  • Open Access

    Stakeholder Perspectives on Trustworthy AI for Parkinson Disease Management Using a Cocreation Approach: Qualitative Exploratory Study

    by Beatriz Alves, Ghada Alhussein, Sara Riggare, Therese Scott Duncan, Ali Saad, David M Lyreskog, Christos Chatzichristos, Ioannis Gerasimou, Stelios Hadjidimitriou, Leontios J Hadjileontiadis, et al.

    Background: Parkinson disease (PD) is the fastest-growing neurodegenerative disorder in the world, with prevalence expected to exceed 12 million by 2040, which poses significant health care and societal challenges. Artificial intelligence (AI) systems and wearable sensors hold potential for PD diagnosis, personalized symptom monitoring, and progression prediction. Nonetheless, ethical AI adoption requires several core principles, including user trust, transparency, fairness, and human oversight. Objective: This study aims to explore and synthesize the perspectives of diverse stakeholders, such as individuals living with PD, health care professionals, AI experts, and bioethicists. The aim was to guide the development of AI-driven digital health solutions, emphasizing transparency, data security, fairness, and bias mitigation while ensuring robust human oversight. These efforts are part of the broader Artificial Intelligence-Based Parkinson’s Disease Risk Assessment and Prognosis (AI-PROGNOSIS) European project, dedicated to advancing ethical and effective AI applications in PD diagnosis and management. Methods: An exploratory qualitative approach, based on 2 datasets constructed from cocreation workshops, engaged key stakeholders with diverse expertise to gather insights, ensuring a broad range of perspectives and enriching the thematic analysis. A total of 24 participants participated in the cocreation workshops, including 11 (46%) people with PD, 6 (25%) health care professionals, 3 (13%) AI technical experts, 1 (4%) bioethics expert, and 3 (13%) facilitators. Using a semistructured guide, key aspects of the discussion centered on trust, fairness, explainability, autonomy, and the psychological impact of AI in PD care. Results: Thematic analysis of the cocreation workshop transcripts identified 5 key main themes, each explored through various corresponding subthemes. AI trust and security (theme 1) was highlighted, focusing on data safety and the accuracy and reliability of the AI systems. AI transparency and education (theme 2) emphasized the need for educational initiatives and the importance of transparency and explainability of AI technologies. AI bias (theme 3) was identified as a critical theme, addressing issues of bias and fairness and ensuring equitable access to AI-driven health care solutions. Human oversight (theme 4) stressed the significance of AI-human collaboration and the essential role of human review in AI processes. Finally, AI’s psychological impact (theme 5) examined the emotional impact of AI on patients and how AI is perceived in the context of PD care. Conclusions: Our findings underline the importance of implementing robust security measures, developing transparent and explainable AI models, reinforcing bias mitigation and reduction strategies and equitable access to treatment, integrating human oversight, and considering the psychological impact of AI-assisted health care. These insights provide actionable guidance for developing trustworthy and effective AI-driven digital PD diagnosis and management solutions.


  • Open Access

    Stakeholder Perspectives on Trustworthy AI for Parkinson Disease Management Using a Cocreation Approach: Qualitative Exploratory Study (Preprint)

    by Beatriz Alves, Ghada Alhussein, Sara Riggare, Therese Scott Duncan, Ali Saad, David M Lyreskog, Christos Chatzichristos, Ioannis Gerasimou, Stelios Hadjidimitriou, Leontios J Hadjileontiadis, et al.

    BACKGROUND Parkinson disease (PD) is the fastest-growing neurodegenerative disorder in the world, with prevalence expected to exceed 12 million by 2040, which poses significant health care and societal challenges. Artificial intelligence (AI) systems and wearable sensors hold potential for PD diagnosis, personalized symptom monitoring, and progression prediction. Nonetheless, ethical AI adoption requires several core principles, including user trust, transparency, fairness, and human oversight. OBJECTIVE This study aims to explore and synthesize the perspectives of diverse stakeholders, such as individuals living with PD, health care professionals, AI experts, and bioethicists. The aim was to guide the development of AI-driven digital health solutions, emphasizing transparency, data security, fairness, and bias mitigation while ensuring robust human oversight. These efforts are part of the broader Artificial Intelligence-Based Parkinson’s Disease Risk Assessment and Prognosis (AI-PROGNOSIS) European project, dedicated to advancing ethical and effective AI applications in PD diagnosis and management. METHODS An exploratory qualitative approach, based on 2 datasets constructed from cocreation workshops, engaged key stakeholders with diverse expertise to gather insights, ensuring a broad range of perspectives and enriching the thematic analysis. A total of 24 participants participated in the cocreation workshops, including 11 (46%) people with PD, 6 (25%) health care professionals, 3 (13%) AI technical experts, 1 (4%) bioethics expert, and 3 (13%) facilitators. Using a semistructured guide, key aspects of the discussion centered on trust, fairness, explainability, autonomy, and the psychological impact of AI in PD care. RESULTS Thematic analysis of the cocreation workshop transcripts identified 5 key main themes, each explored through various corresponding subthemes. AI trust and security (theme 1) was highlighted, focusing on data safety and the accuracy and reliability of the AI systems. AI transparency and education (theme 2) emphasized the need for educational initiatives and the importance of transparency and explainability of AI technologies. AI bias (theme 3) was identified as a critical theme, addressing issues of bias and fairness and ensuring equitable access to AI-driven health care solutions. Human oversight (theme 4) stressed the significance of AI-human collaboration and the essential role of human review in AI processes. Finally, AI’s psychological impact (theme 5) examined the emotional impact of AI on patients and how AI is perceived in the context of PD care. CONCLUSIONS Our findings underline the importance of implementing robust security measures, developing transparent and explainable AI models, reinforcing bias mitigation and reduction strategies and equitable access to treatment, integrating human oversight, and considering the psychological impact of AI-assisted health care. These insights provide actionable guidance for developing trustworthy and effective AI-driven digital PD diagnosis and management solutions. CLINICALTRIAL

  • Open Access

    An agile co-creation approach for designing a comprehensive digital motor assessment test for Parkinson's Disease patients

    by Beatriz Alves, Ghada Alhussein, Filomena Carnide, Nikolaos Grammalidis, Kosmas Dimitropoulos, Dhaval Trivedi, Mónica Kurtis Urra, Stelios Hadjidimitriou, Leontios Hadjileontiadis, Sofia B. Dias


  • Novel Digital Biomarkers for Fine Motor Skills Assessment in Psoriatic Arthritis: The DaktylAct Touch-Based Serious Game Approach

    by Eleni Vasileiou, Sofia B. Dias, Stelios Hadjidimitriou, Vasileios Charisis, Nikolaos Karagkiozidis, Stavros Malakoudis, Patty de Groot, Stelios Andreadis, Vassilis Tsekouras, Georgios Apostolidis, et al.

    Psoriatic Arthritis (PsA) is a chronic, inflammatory disease affecting joints, substantially impacting patients' quality of life, with European guidelines for managing PsA emphasizing the importance of assessing hand function. Here, we present a set of novel digital biomarkers (dBMs) derived from a touchscreen-based serious game approach, DaktylAct, intended as a proxy, gamified, objective assessment of hand impairment, with emphasis on fine motor skills, caused by PsA. This is achieved by its design, where the user controls a cannon to aim at and hit targets using two finger pinch-in/out and wrist rotation gestures. In-game metrics (targets hit and score) and statistical features (mean, standard deviation) of gameplay actions (duration of gestures, applied pressure, and wrist rotation angle) produced during gameplay serve as informative dBMs. DaktylAct was tested on a cohort comprising 16 clinically verified PsA patients and nine healthy controls (HC). Correlation analysis demonstrated a positive correlation between average pinch-in duration and disease activity (DA) and a negative correlation between standard deviation of applied pressure during wrist rotation and joint inflammation. Logistic regression models achieved 83% and 91% classification performance discriminating HC from PsA patients with low DA (LDA) and PsA patients with and without joint inflammation, respectively. Results presented here are promising and create a proof-of-concept, paving the way for further validation in larger cohorts.