Recently, researchers and students at the VinUni-Illinois Smart Health Center (VISHC) have published 05 new studies in prestigious AI and Medicine Journals & Conferences, which include Applied Physics Letters (Q1, IF 4.0), IEEE Journal of Biomedical and Health Informatics (Q1, IF 7.7), Trends in Analytical Chemistry (Q1, IF 13.53), Nature Scientific Reports (Q1, IF 4.6), and IEEE Computer Vision and Pattern Recognition Conference (CVPR) Workshop. Our research spans a diverse array of AI and Medicine topics, encompassing both foundational studies and various facets of intelligent healthcare applications. Below we are enthusiastic about sharing our latest publications with the community. A full list of publications can be found here.
01. Viet Anh Nguyen, Viet Hoang Le; Eirini Sarelli, Loïc Malgrey, Dang-Khue Luu, Ha Linh Chu, Truong Tuan Vu, Cong Quang Tong, Dinh Lam Vu, Christian Seassal, Quynh Le-Van, Hai Son Nguyen. “Direct observation of exceptional points in photonic crystal by cross-polarization imaging in momentum space.” Applied Physics Letters 123.19 (2023).
Abstract: This study explores exceptional points (EPs) in photonic crystals (PhCs) and introduces an experimental technique for their single-shot observation. Exceptional points are spectral singularities found in non-Hermitian systems, such as leaky PhC slabs. However, directly observing EPs in PhC systems using regular reflectivity spectroscopy is a considerable challenge due to interference between guided resonances and background signals. In this work, we present a simple, nondestructive technique that employs crossed polarizations to directly observe EPs in momentum-resolved resonant scattering. This approach effectively suppresses the background signal, enabling exclusive probing of the guided resonances where EPs manifest. Our results demonstrate the formation of EPs in both energy-momentum mapping and isofrequency imaging. All experimental findings align seamlessly with numerical simulations and analytical models. Our approach holds great potential as a robust tool for studying non-Hermitian physics in the PhC platform.
Full paper: https://pubs.aip.org/aip/apl/article-abstract/123/19/191105/2920108/
02. Nguyen, Son Hai, Van-Nhat Nguyen, and Mai Thi Tran. “Dual-channel fluorescent sensors based on chitosan-coated Mn-doped ZnS micromaterials to detect ampicillin.” Scientific Reports 14.1 (2024): 10066.
Abstract: The global threat of antibiotic resistance has increased the importance of the detection of antibiotics. Conventional methods to detect antibiotics are time-consuming and require expensive specialized equipment. Here, we present a simple and rapid biosensor for detecting ampicillin, a commonly used antibiotic. Our method is based on the fluorescent properties of chitosan-coated Mn-doped ZnS micromaterials combined with the β-lactamase enzyme. The biosensors exhibited the highest sensitivity in a linear working range of 13.1–72.2 pM with a limit of detection of 8.24 pM in deionized water. In addition, due to the biological specificity of β-lactamase, the proposed sensors have demonstrated high selectivity over penicillin, tetracycline, and glucose through the enhancing and quenching effects at wavelengths of 510 nm and 614 nm, respectively. These proposed sensors also showed promising results when tested in various matrices, including tap water, bottled water, and milk. Our work reports for the first time the cost-effective (Mn:ZnS)Chitosan micromaterial was used for ampicillin detection. The results will facilitate the monitoring of antibiotics in clinical and environmental contexts.
Full paper: https://www.nature.com/articles/s41598-024-59772-3
03. Linh Thi Phuong Le, Anh Hoang Quan Nguyen, Le Minh Tu Phan, Hien Thi Thanh Ngo, Xing Wang, Brian Cunningham, Enrique Valera, Rashid Bashir, Andrew W. Taylor-Robinson, Cuong Danh Do. “Current smartphone-assisted point-of-care cancer detection: Towards supporting personalized cancer monitoring” Trends in Analytical Chemistry174, 117681, 2024. 19 pages. doi: 10.1016/j.trac.2024.117681 (Q1 journal; IF 13.53)
Abstract: With the rising incidence of cancer-related mortality, new enabling technologies are necessary to offer comprehensive molecular profiles of patients in order to assist clinicians to establish an early presumptive diagnosis. Biosensors that are technically comparable to conventional laboratory diagnostics have been developed. However, because the manufacture and operation processes of these biosensors were not well adjusted for end-users as patients, the approach is not optimized in these newly built platforms. Hence, smartphone-assisted biosensors have been developed for point-of-care utilization. They are faster, simpler, and more affordable than standard examinations and first-generation biosensors, however, they raise numerous concerns regarding their applicability for early cancer detection. Therefore, this review focuses primarily on cutting edge developments in smartphone-assisted biosensing platforms that are most relevant to early cancer diagnosis, including optical and electrochemical biosensors, and cancer imaging. What is needed to bring this important technology to realization as early cancer diagnosis tool is discussed.
Full paper: https://www.sciencedirect.com/science/article/abs/pii/S0165993624001638
04. Hong Nguyen, Hoang Nguyen, Melinda Chang, Hieu Pham, Shrikanth Narayanan, Michael Pazzani – “ConPro: Learning Severity Representation for Medical Images using Contrastive Learning and Preference Optimization” – IEEE Computer Vision and Pattern Recognition Conference (CVPR) Workshop on Domain adaptation, Explainability, Fairness in AI for Medical Image Analysis.
Abstract: Understanding the severity of conditions shown in images in medical diagnosis is crucial, serving as a key guide for clinical assessment, treatment, as well as evaluating longitudinal progression. This paper proposes Con- PrO: a novel representation learning method for severity assessment in medical images using Contrastive learningintegrated Preference Optimization. Different from conventional contrastive learning methods that maximize the distance between classes, ConPrO injects into the latent vector the distance preference knowledge between various severity classes and the normal class. We systematically examine the key components of our framework to illuminate how contrastive prediction tasks acquire valuable representations. We show that our representation learning framework offers valuable severity ordering in the feature space while outperforming previous state-of-the-art methods on classification tasks. We achieve a 6% and 20% relative improvement compared to a supervised and a self-supervised baseline, respectively. In addition, we derived discussions on severity indicators and related applications of preference comparison in the medical domain.
Full paper (preprint): https://arxiv.org/pdf/2404.18831
05. Hieu X Nguyen, Duong V Nguyen, Hieu H Pham, Cuong D Do – “MPCNN: A Novel Matrix Profile Approach for CNN-based Sleep Apnea Classification” – IEEE Journal of Biomedical and Health Informatics (J-BHI).
Abstract: Sleep apnea (SA) is a significant respiratory condition that poses a major global health challenge. Previous studies have investigated several machine and deep learning models for electrocardiogram (ECG)-based SA diagnoses. Despite these advancements, conventional feature extractions derived from ECG signals, such as R-peaks and RR intervals, may fail to capture crucial information encompassed within the complete PQRST segments. In this study, we propose an innovative approach to address this diagnostic gap by delving deeper into the comprehensive segments of the ECG signal. The proposed methodology draws inspiration from Matrix Profile algorithms, which generate an Euclidean distance profile from fixed-length signal subsequences. From this, we derived the Min Distance Profile (MinDP), Max Distance Profile (MaxDP), and Mean Distance Profile (MeanDP) based on the minimum, maximum, and mean of the profile distances, respectively. To validate the effectiveness of our approach, we use the modified LeNet-5 architecture as the primary CNN model, along with two existing lightweight models, BAFNet and SE-MSCNN, for ECG classification tasks. Our extensive experimental results on the PhysioNet Apnea-ECG dataset revealed that with the new feature extraction method, we achieved a per-segment accuracy up to 92.11% and a per-recording accuracy of 100%. Moreover, it yielded the highest correlation compared to state-of-the-art methods, with a correlation coefficient of 0.989. By introducing a new feature extraction method based on distance relationships, we enhanced the performance of certain lightweight models, showing potential for home sleep apnea test (HSAT) and SA detection in IoT devices.
Full paper (preprint): https://arxiv.org/pdf/2311.15041
Congratulations to all authors and co-authors!
