Computational & Digital Health

#sleep apnea screening #glucose prediction #glycemic dynamics #PMS #FemTech #digital phenotyping #digital biomarkers #sound therapy #stress #depression #anxiety #N-of-1 #fNIRS

Core Question: How can AI transform physiological and behavioral data into clinically meaningful health insights?

🏆Funded by JSPS KAKENHI, JSPS Postdoctoral Fellowships for Research in Japan (Summer Program)

🤝International Collaborations with Delaware University (USA)

🏅Young Investigator Award @ Asian Society of Sleep Medicine (ASSM)

Wearable PPG and Questionnaire Based Sleep Apnea Screening/Prediction

Obstructive sleep apnea (OSA) is a common yet underdiagnosed sleep disorder that can lead to serious cardiovascular, metabolic, and cognitive health complications if left untreated. This project investigates accessible, non-invasive approaches for early OSA screening using data collected from consumer wearable devices and self-reported questionnaires. We develop machine learning models that analyze photoplethysmography (PPG) signals and demographic or clinical information to estimate an individual's risk of sleep apnea, reducing the need for costly and inconvenient overnight sleep studies. Our goal is to enable scalable, at-home screening tools that facilitate early detection, timely clinical referral, and more accessible sleep healthcare for the general population.

Selected Publications

[Q1] Hoang NH, Liang Z. (2025) Detection and Severity Classification of Sleep Apnea Using Continuous Wearable SpO2 Signals: A Multi-Scale Feature Approach. Sensors 25(6), 1698. DOI: 10.3390/s25061698. [PubMed/SCI/Scopus]【First author is PhD student】
[Q1] Hoang NH, Liang Z. (2025) AI-driven sleep apnea screening with overnight blood oxygen saturation: current practices and future directions. Front. Digit. Health - Health Informatics. [PubMed/SCI/Scopus]【First author is PhD student】
Hoang NH, Liang Z. (2026) Decision Threshold Optimization in Machine Learning Models for Questionnaire-Based OSA Screening. In Proceedings of the 19th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2026), Marbella, Spain. [Scopus] 【First author is PhD student】(Full paper acceptance rate = 25%)
Hoang N, Liang Z. (2025) Reconsidering AHI as an Indicator of Sleep Apnea Severity: Insights from Mining Large, Longitudinal Sleep Datasets. In Proceedings of the 18th International Conference on Health Informatics (HEALTHINF 2025), Porto, Portugal. 【First author is PhD student】 [Scopus]
[Q1] Liang Z. (2024) Developing Probabilistic Ensemble Machine Learning Models for Home-Based Sleep Apnea Screening using Overnight SpO2 Data at Varying Data Granularity. Sleep and Breathing. Doi: 10.1007/s11325-024-03141-x. [SCI/Scopus/PubMed]
[Q1] Liang Z. (2023) Novel method combining multiscale attention entropy of overnight blood oxygen level and machine learning for easy sleep apnea screening. Digital Health 9: 1-19. [PubMed/SCI/Scopus]

🏆Funded by KUAS Advanced Research Grant

🤝 International Collaborations with the Ohio State University (USA)

🏅Editor's Choice Article Award 2025

🏅Best Presentation Award @ WSSE 2025

🏅Oral Presentation Award @ GCCE 2025

🏅Finalist Team @ IEEE CIS Student Grand Competition 2025

🏅Best Presentation Award @ BDCI 2024

Non-Invasive Glucose Monitoring and Prediction Using Wearable AI

Continuous glucose monitoring has transformed diabetes care, but current systems often rely on invasive sensors that can be costly and uncomfortable for long-term use. This research explores how consumer wearable devices and artificial intelligence can enable non-invasive glucose monitoring and prediction in free-living environments. By combining multimodal physiological signals—including heart rate, activity, sleep, and circadian information—with advanced machine learning and deep learning models, we develop personalized approaches for estimating and forecasting glucose levels. Our work also investigates feature selection, physiological complexity, interoperable AI frameworks, and model generalizability across healthy, prediabetic, and diabetic populations. The long-term goal is to create accessible, scalable, and personalized digital health solutions that support proactive metabolic health management without the need for invasive sensing.

Selected Publications

[Q1] Karunarathna TS, Liang Z. (2025) Development of Non-invasive Continuous Glucose Prediction Models Using Multi-Modal Wearable Sensors in Free-living Conditions. Sensors. [PubMed/SCI/Scopus]【First author is undergraduate student】
Karunarathna T, Carvalho CF, Liang Z. (2025) Short-Term Glucose Prediction in Diabetes: Comparing Deep Learning Training Paradigms. In the Proceedings of the 2025 IEEE 14th Global Conference on Consumer Electronics (GCCE 2025), Osaka. Japan.【First author is master student】
Carvalho CF, Karunarathna T, Liang Z. (2025) Multiscale Entropy Analysis of Continuous Glucose Monitoring Data: A Comparative Study of Diabetic and Healthy Populations. In Proceedings of the 18th International Conference on Health Informatics (HEALTHINF 2025), Porto, Portugal.【First author is master student】 [Scopus]
Karunarathna TS, Liang Z. (2025) Bottom-Up Feature Selection for Continuous Glucose Prediction Using Wearable Data. In the Proceedings of the 2025 IEEE 14th Global Conference on Consumer Electronics (GCCE 2025), Osaka. Japan.【First author is undergraduate student】
Liang Z, Karunarathna TS, Baig A, Beten A, Lococo L. AIM-W: an Input-Adaptive Framework for Interoperable Physiological Modeling on Wearables. In Proceedings of the 2025 IEEE 11th World Forum on Internet of Things (WF-IoT 2025), Chengdu, China. [Scopus]
Karunarathna T, Carvalho CF, Liang Z. (2025) Predicting Glucose Levels: The Role of Physiological, Circadian, Behavioral and Demographic Features. In Proceedings of the 9th International Conference on Medical and Health Informatics (ICMHI 2025), Kyoto, Japan.【First author is undergraduate student】(reviewed)
Karunarathna T, Liang Z. (2024) Developing Consumer Wearable Compatible Computational Models for Non-Invasive Glucose Estimation in Diabetes Patients. In Proceedings of 2024 IEEE Consumer Life Tech Conference (ICLT 2024), Sydney, Australia.【First author is undergraduate student】[Scopus]
Carvalho CF, Karunarathna TS, Liang Z. (2024) LSTM-Based Glucose Prediction for Prediabetes. In Proceedings of the 13th Global Conference on Consumer Electronics (GCCE 2024), Kitakyushu, Japan.【First author is master student】[SCI/Scopus]
Carvalho C, Liang Z. (2024) Glucose Prediction with Long Short-Term Memory (LSTM) Models on Three Distinct Populations. In Proceedings of The 11th International Electronic Conference on Sensors and Applications (ECSA-11), online.【First author is master student】[Scopus]

🏆Funded by Exploration France by Embassy of France in Japan

🤝 International Collaborations with A*STAR (Singapore), Université Paris Cité (France)

Foundation Models for Digital Phenotyping and Clinical AI

Understanding complex health conditions often requires integrating diverse clinical information rather than relying on a single data source. This project explores the use of large language models (LLMs) and foundation models to advance digital phenotyping—the automated identification and characterization of health conditions from multimodal health data. We develop AI frameworks that combine structured clinical data, synthesized medical knowledge, and reasoning capabilities to perform robust disease phenotyping, including the identification of multiple coexisting conditions in zero-shot settings. By leveraging foundation models for clinical reasoning, this research aims to create more generalizable, interpretable, and data-efficient AI systems that support precision medicine and next-generation healthcare analytics.

Selected Publications

[Q1, top 5%] Priyadarshana YHPP, Zhou N, Krishnaswamy P, Jiang R, Liang Z. (2026) MediPhen: Prompt-Based LLM Reasoning with Synthesized Multimodal Clinical Knowledge for Zero-Shot Multi-morbidity Phenotyping. IEEE Journal of Biomedical and Health Informatics. [SCI/Scopus]【First author was PhD student】 (Acceptance rate = 14.4%)

🏅Honorable Mention Award @ IEEE CIS Student Grand Competition 2024

🏅Best Paper Award @ WEBIST 2023

AI for Digital Mental Health and Computational Psychiatry

Mental health assessment increasingly relies on digital interactions, creating opportunities for AI to support early detection and personalized care. This research develops advanced natural language processing and large language model (LLM) techniques to identify depressive symptoms and other mental health indicators from conversational and textual data. Our work spans graph neural networks for multi-party dialogue analysis, retrieval-augmented LLMs, prompt engineering, explainable AI, and knowledge-enhanced reasoning to improve the accuracy, robustness, and transferability of digital mental health models. By combining computational linguistics with clinical insights, we aim to develop trustworthy AI systems that enable scalable mental health screening, support clinical decision-making, and advance the future of computational psychiatry.

Selected Publications

[Q1, top 10%] Priyadarshana YHPP, Liang Z, Piumarta I. (2026) DoRA: Dual-encoder demonstration Retriever Architecture to Transfer Large Language Models for Depressive Symptom Detection. ACM Transactions on Computing for Healthcare 7(1), Article No.10, p.1-21. [SCI/Scopus]【First author is PhD student】
[Q1] Priyadarshana YHPP, Liang Z, Piumarta I. (2025) GEM: Graph Attention Encoder for Multi-task Depression Severity Detection in Multi-party Conversations. Social Network Analysis and Mining. [SCI/Scopus]【First author is PhD student】
[Q1] Priyadarshana YHPP, Liang Z, Piumarta I. (2025) ExDoRA: Enhancing transferability of large language models for depression detection using free-text explanations. Frontiers in Artificial Intelligence. [PubMed/SCI/Scopus]【First author is PhD student】
[Q2] Priyadarshana YHPP, Liang Z, Piumarta I. (2025) Who Says What (WSW) 2.0 : Utterance Semantic Modelling for Speaker Identification in Text-Based Multi-Party Conversations. SN Computer Science. [Scopus]【First author is PhD student】
[Q1] Priyadarshana YHPP, Senanayake A, Liang Z, Piumarta I. (2024) Prompt Engineering for Digital Mental Health: A Short Review. Frontiers in Digital Health - Digital Mental Health 6:1410947. Doi: 10.3389/fdgth.2024.1410947. [SCI/Scopus/PubMed]【First author is PhD student】

🏅Honorable Mention for Best Poster Award @ HEALTHINF 2025

AI for Anxiety Detection from Digital Behavior

Digital platforms provide rich longitudinal records of how people express thoughts, emotions, and experiences, creating new opportunities for early mental health assessment. This research develops artificial intelligence and natural language processing techniques to detect anxiety from social media and other text-based digital interactions. Our work investigates emotion dynamics, temporal language patterns, contextualized language embeddings, and longitudinal behavioral modeling to identify subtle changes associated with anxiety over time. By moving beyond single-post analysis to model how emotional states evolve, this research aims to enable more accurate, timely, and explainable digital mental health screening while supporting early intervention and personalized mental healthcare.

Selected Publications

Ashala Lakmini Senanayake, Zilu Liang. CIERTe: Continuous Inter-Event Relative Temporal Anchoring for Anxiety Detection from Social Media Posts. In Proceedings of the 17th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM BCB), Rende, Italy. [Scopus] 【First author is master student】(Regular paper acceptance rate = 20%)
Senanayake A, Liang Z. (2025) Anxiety Detection in Reddit Posts through Emotion Dynamics Analysis. In Proceedings of the 18th International Conference on Health Informatics (HEALTHINF 2025), Porto, Portugal.【First author is master student】[Scopus]
Senanayake A, Liang Z. (2025) CE-LAH: Contextualized Embeddings for Longitudinal Analysis of Human Anxiety. In the Proceedings of the 2025 IEEE 14th Global Conference on Consumer Electronics (GCCE 2025), Osaka. Japan. 【First author is master student】

🤝Collaborations with Institute of Science Tokyo (東京工業大学)

🏅Oral Presentation Award @ GCCE 2025

Human Voice as Stress Therapy: Multimodal Stress Sensing and Digital Stress Management

Stress is a major contributor to both physical and mental health problems, yet objective assessment outside laboratory settings remains challenging. This research investigates how multimodal physiological sensing, behavioral signals, and digital interventions can be combined to better understand, measure, and reduce human stress. Our work spans the development of novel stress induction protocols, multimodal analyses of speech and paralanguage, wearable-based estimation of physiological stress biomarkers such as cortisol, and systematic evaluations of sound-based interventions for stress reduction. By integrating wearable sensing, psychophysiology, artificial intelligence, and behavioral science, we aim to develop objective, personalized, and scalable technologies for continuous stress monitoring and effective stress management in everyday life.

Selected Publications

[Q1, top 10%] Saskovets M, Saponkova I, Liang Z. (2025) Effects of Sound Interventions on the Mental Stress Response in Adults: A Scoping Review. JMIR Mental Health. 22/11/2024:69120. DOI: 10.2196/preprints.69120. [SCI/Scopus/PubMed]【First author is PhD student】
[Q2] Saskovets M, Mykhailo Lohachov, Liang Z. (2025) Paralanguage as a Tool for Shaping Stress Response in Listeners: Multimodal Physiological Sensing Study. Comprehensive Psychoneuroendocrinology. [PubMed/SCI/Scopus]【First author is PhD student】
[Q2] Marina Saskovets, Mykhailo Lohachov, Zilu Liang. (2025) Validation of a New Stress Induction Protocol using Speech Improvisation. Brain Sciences. [PubMed/SCI/Scopus]【First author is PhD student】
Saskovets M, Lohachov M, Liang Z. (2025) Cortisol Estimation Using Wearable Wristbands: Comparing Multilevel Modelling and Machine Learning. In the Proceedings of the 2025 IEEE 14th Global Conference on Consumer Electronics (GCCE 2025), Osaka. Japan.【First author is PhD student】
Saskovets M, Liang Z, Piumarta I, Saponkova I. (2024) Effects of Sound Interventions on the Mental Stress Response in Adults: Protocol for A Scoping Review. JMIR Research Protocols 13:e54030. Doi:10.2196/54030. [SCI/Scopus/PubMed]【First author is PhD student】

🏆Funded by JSPS KAKENHI, KUAS Internal Funding for Research Acitivty Start-up

Neural Informatics: Wearable Neuroimaging for Understanding Human Physiology and Psychology

The human brain provides valuable insights into both physiological processes and psychological states, yet monitoring brain activity in everyday environments remains challenging. This research explores the use of wearable functional near-infrared spectroscopy (fNIRS) to investigate the neural mechanisms underlying human physiology, cognition, emotion, and behavior. By combining portable neuroimaging with computational analysis and machine learning, we study how changes in cerebral hemodynamics reflect internal states such as stress, cognitive workload, and sleep-related brain activity. Our long-term goal is to establish wearable neuroimaging as a practical tool for advancing neuroscience, human-computer interaction, and next-generation digital health technologies.

Selected Publications

[Q2] Liang Z. (2021) What does sleeping brain tell about stress? A pilot fNIRS study into stress-related cortical hemodynamic features during sleep. Frontiers in Computer Science (Section: Mobile and Ubiquitous Computing) 3:774949. Doi: 10.3389/fcomp.2021.774949. [SCI /Scopus]

🤝 International Collaborations with Université de Toulouse (France)

Sleep Health Informatics

blablabla

Selected Publications

[Q1, top 10%] Nhung, H. H., Liang Z. (2023) Knowledge discovery in ubiquitous and personal sleep-tracking: a scoping review. JMIR mHealth and uHealth 11:e42750 [PubMed/SCI/Scopus]【First author is master student】
[Q2] Liang Z. (2022). Context-aware sleep health recommender systems (CASHRS): a narrative review. Electronics 2022, 11(20), 3384. Doi: 10.3390/electronics1120338. [SCI/Scopus]
Liang Z., Nhung H. H., Bertrand L., Cleyet-Marrel N. (2022) Context-aware sleep analysis with intraday steps and heart rate time series data from consumer activity trackers. In Proceedings of the 15th International Conference on Health Informatics (HEALTHINF 2022), Cyberspace. (Full paper acceptance rate = 23%) [SCI/Scopus]
Liang Z. (2022) Correlation analysis of nested consumer health data: a new look at an old problem. In Proceedings of the IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech 2022), Osaka, Japan. [SCI/Scopus]

🤝 International Collaborations with the Ohio State University (USA)

🏅Best Presentation Award @ ICMHI 2026

Continuous Glucose Monitoring: Computational Modeling and Visual Analytics

Continuous glucose monitoring (CGM) offers unprecedented insights into how diet, sleep, physical activity, and individual physiology influence metabolic health. This research develops computational methods and visual analytics tools to better understand glucose dynamics in everyday life. Our work includes within-person modeling of postprandial glucose responses, interactive dashboards for exploring CGM data, visual analytics for dietary and glucose patterns, and data mining techniques to uncover actionable relationships between lifestyle behaviors and metabolic outcomes. By combining wearable sensing, artificial intelligence, and human-centered visualization, we aim to transform complex metabolic data into personalized insights that support healthier lifestyles and precision nutrition.

Selected Publications

[Q1] Liang Z. Within-Person Modeling of Postprandial Glucose Using Multimodal Wearable Data. Frontiers in Digital Health. [SCI/Scopus]
Liang Z, Lococo M, Karunarathna TS. (2026) Visual Analytics of Sugar Intake and Postprandial Interstitial Glucose Dynamics Using 3D Surface Plots. In Proceedings of The The 10th International Conference on Medical and Health Informatics (ICMHI 2026), Kyoto, Japan. [Scopus]
Karunarathna TS, Hoang NH, Liang Z. (2026) GlucoPatrol: A Multi-Stakeholder Dashboard for Contextual Exploration of Continuous Glucose Monitoring Data. In Proceedings of the 19th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2026), Marbella, Spain. [Scopus]【First author is master student】
Carvalho CF, Karunarathna T, Liang Z. (2025) Exploring Gender Differences in Interstitial Glucose Dynamics through Multiscale Entropy Analysis. In Proceedings of the 9th International Conference on Medical and Health Informatics (ICMHI 2025), Kyoto, Japan.【First author is master student】(reviewed)
Nhung H. H., Liang Z. (2023) Contrast set mining for actionable insights into associations between sleep and glucose in a normoglycemic population. In Proceedings of the 16th International Conference on Health Informatics (HEALTHINF 2023), Lisbon, Portugal.【First author is master student】[SCI/Scopus]

Biochemical Health Informatics: Advanced Analytics for Personalized Nutrition

Advances in consumer health technologies have made it possible for individuals to monitor biochemical markers such as glucose, vitamins, minerals, and other nutritional indicators outside clinical settings. This research investigates how biochemical self-tracking can support self-knowledge and healthier decision-making through consumer-friendly sensing technologies, visual analytics, and human-centered design. Our work explores continuous glucose monitoring, at-home micronutrient testing, dietary assessment, and personal health dashboards to understand how people interpret biomarker data, identify meaningful behavioral patterns, and translate insights into sustainable lifestyle changes. By combining personal informatics, wearable technologies, and artificial intelligence, we aim to transform biochemical measurements into actionable knowledge for personalized health management.

Selected Publications

Liang Z, Hoang NH, Karunarathna TS. (2026) Biochemical Self-Tracking as Personal Informatics: Sensemaking and Actionability of At-Home Micronutrient Tests. In Proceedings of the 19th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2026), Marbella, Spain. [Scopus]
Liang Z. (2025) Assessing Dietary Protein Intake: A Comparative Study of Two Consumer Mail-In Nutrition Test Kits. In Proceedings of the 18th International Conference on Health Informatics (HEALTHINF 2025), Porto, Portugal. [Scopus]
Liang Z, Hoang NH, Karunarathna T. (2025) A Pilot Study on Sensemaking of Consumer Micronutrient Tests: Sufficiencies, Deficiencies, and Barriers to Improvement. In Proceedings of the Envisioning the Future of Interactive Health Workshop in conjunction with CHI 2025, Yokohama, Japan. (reviewed)
Liang Z (2024) Exploring the Impact of Wearable Continuous Glucose Monitoring on Glucose Regulation and Eating Behavior in Healthy Individuals: A Pilot Study. In Proceedings of The 2nd Behavior Transformation by IoT International Workshop (BTIW 2024), Tokyo, Japan. [SCI/Scopus]
Bertrand L, Cleyet-Marrel N, Liang Z. (2021) The role of continuous glucose monitoring (CGM) in automatic detection of eating activities. In Proceedings of the IEEE 3rd Global Conference on Life Sciences and Technologies (LifeTech 2021), Nara, Japan. 【First authors are visiting master students】[SCI /Scopus]

Page updated

Google Sites

Report abuse