Validation of Remote Photoplethysmography for Non-Invasive Estimation of Blood Glucose and HbA1c

Официальное название

Validation of Remote Photoplethysmography for Non-Invasive Estimation of Blood Glucose and HbA1c in a Community-Based Population in Jakarta

Обзор

The goal of this observational study is to evaluate whether a non-invasive facial scan technology using remote photoplethysmography (rPPG) can accurately estimate blood glucose and HbA1c levels in adults living in the community in Jakarta. The study focuses on adults aged 18 years and older, including individuals with or without diabetes. The main questions it aims to answer are: 1. Can rPPG-based facial scan estimates of blood glucose and HbA1c match results from standard laboratory blood tests? 2. How well can rPPG identify individuals with high blood sugar or diabetes risk based on established clinical cut-off values? Researchers will compare results from the rPPG facial scan with standard laboratory measurements of fasting blood glucose and HbA1c to determine how accurate and reliable the technology is for screening purposes. Participants will: 1. Provide basic information such as age, sex, and medical history 2. Undergo a non-invasive facial scan using a smartphone-based system 3. Have a blood sample taken to measure fasting blood glucose and HbA1c 4. Complete all assessments during a single study visit This study aims to determine whether rPPG can serve as a simple, non-invasive, and accessible tool for early detection and monitoring of diabetes in community settings.

Подробное описание

Introduction Type 2 diabetes mellitus (T2DM) represents a major global health burden characterized by chronic hyperglycemia and associated complications. Standard monitoring methods, such as fasting blood glucose and glycated hemoglobin (HbA1c), rely on invasive blood sampling and access to laboratory facilities, which may reduce patient adherence and limit early detection. Remote photoplethysmography (rPPG), a non-contact optical technique using facial video analysis, has emerged as a promising alternative for estimating physiological and metabolic parameters. However, evidence regarding its validity in assessing glycemic markers remains limited .

Objective This study aims to evaluate the validity and diagnostic performance of rPPG-based facial scan technology in estimating blood glucose and HbA1c levels compared with standard laboratory measurements.

Methods This study employs an analytical observational design with a cross-sectional diagnostic validation approach conducted in Kelurahan Semanan, Jakarta. A total of 150-300 adult participants will be recruited using a community-based sampling method. Each participant will undergo venous blood sampling for laboratory measurement of fasting blood glucose and HbA1c, alongside a non-contact rPPG facial scan using a smartphone-based system. Agreement between methods will be assessed using Bland-Altman analysis, while correlation analysis (Pearson/Spearman) will evaluate the strength of association. Diagnostic performance, including sensitivity and specificity, will be calculated using clinical cut-offs (≥126 mg/dL for glucose and ≥6.5% for HbA1c).

Expected Results It is expected that rPPG-derived estimates will demonstrate moderate to good correlation with laboratory measurements, with acceptable agreement for screening purposes. The technology is anticipated to show reasonable diagnostic performance in identifying individuals with high glycemic risk. These findings may support the feasibility of rPPG as a non-invasive, accessible screening tool for diabetes monitoring in community settings.

Первичные конечные точки

• Agreement Between rPPG-Derived and Laboratory Blood Glucose [Срок оценки: Single assessment at baseline (during study visit)]
• Agreement Between rPPG-Derived and Laboratory HbA1c [Срок оценки: Single assessment at baseline (during study visit)]
• Correlation and Validation of rPPG Estimates with Laboratory Blood Glucose and HbA1c [Срок оценки: Single assessment at baseline (during study visit)]
• Diagnostic Performance of rPPG for Detecting Hyperglycemia and Diabetes Risk [Срок оценки: Single assessment at baseline (during study visit)]

Критерии участия

Критерии включения

• Adults aged ≥18 years
• Willing to participate and provide informed consent
• Able to undergo facial scan and blood examination
• Stable clinical condition

Критерии исключения

• Facial conditions interfering with rPPG signal (e.g., wounds, deformities)
• Use of facial coverings obstructing camera detection
• Inability to remain still during facial scan
• Incomplete data or withdrawal from study

Критерии приведены из реестра в оригинале (на английском). Окончательную оценку соответствия проводит исследовательский центр.

Здоровые добровольцы: Да

Дизайн исследования

Модель наблюдения

Экологическое

Центры проведения

Индонезия · 1 центр

• Kelurahan Semanan — Jakarta

Публикации

• Zeynali M, Alipour K, Tarvirdizadeh B, Ghamari M. Non-invasive blood glucose monitoring using PPG signals with various deep learning models and implementation using TinyML. Sci Rep. 2025 Jan 2;15(1):581. doi: 10.1038/s41598-024-84265-8. PMID 39753714
• Zanelli S, Ammi M, Hallab M, El Yacoubi MA. Diabetes Detection and Management through Photoplethysmographic and Electrocardiographic Signals Analysis: A Systematic Review. Sensors (Basel). 2022 Jun 29;22(13):4890. doi: 10.3390/s22134890. PMID 35808386
• Shi B, Dhaliwal SS, Soo M, Chan C, Wong J, Lam NWC, Zhou E, Paitimusa V, Loke KY, Chin J, Chua MT, Liaw KCS, Lim AWH, Insyirah FF, Yen SC, Tay A, Ang SB. Assessing Elevated Blood Glucose Levels Through Blood Glucose Evaluation and Monitoring Using Machine Learning and Wearable Photoplethysmography Sensors: Algorithm Development and Validation. JMIR AI. 2023 Oct 27;2:e48340. doi: 10.2196/48340. PMID 38875549
• Santillan A, Travez Proano EI, Jaramillo Encalada IN, Abril Lopez PA, Tricallotis J, Acosta-Espana JD. Structured telemonitoring reduces HbA1c and emergency visits in insulin-treated type 2 diabetes: a controlled cohort study in Ecuador's public hospital. Front Clin Diabetes Healthc. 2026 Feb 9;7:1734589. doi: 10.3389/fcdhc.2026.1734589. eCollection 2026. PMID 41737571
• Qawqzeh YK, Bajahzar AS, Jemmali M, Otoom MM, Thaljaoui A. Classification of Diabetes Using Photoplethysmogram (PPG) Waveform Analysis: Logistic Regression Modeling. Biomed Res Int. 2020 Aug 11;2020:3764653. doi: 10.1155/2020/3764653. eCollection 2020. PMID 32851065
• Kwon TH, Kim KD. Machine-Learning-Based Noninvasive In Vivo Estimation of HbA1c Using Photoplethysmography Signals. Sensors (Basel). 2022 Apr 12;22(8):2963. doi: 10.3390/s22082963. PMID 35458947
• Farenden E, Kelly J, Russell A, Menon A. Remote Monitoring for Type 2 Diabetes: What Do Patients, Healthcare Professionals, and Executives Think? Stud Health Technol Inform. 2024 Jan 25;310:1526-1527. doi: 10.3233/SHTI231276. PMID 38269728
• Chu J, Yang WT, Lu WR, Chang YT, Hsieh TH, Yang FL. 90% Accuracy for Photoplethysmography-Based Non-Invasive Blood Glucose Prediction by Deep Learning with Cohort Arrangement and Quarterly Measured HbA1c. Sensors (Basel). 2021 Nov 24;21(23):7815. doi: 10.3390/s21237815. PMID 34883817

Идентификаторы

NCT: NCT07502690 · 20260324

Первоисточники (государственные реестры)

• ClinicalTrials.gov (NIH, США)

Открыть это исследование на ClinicalTrials.gov ↗