Department of Biomedical Engineering and Informatics Works
Permanent URI for this collection
Browse
Recent Submissions
Item Deep learning based analysis of sentiment dynamics in online cancer community forums: An experience(Sage, 2021) Balakrishnan, Athira; Idicula, Sumam Mary; Jones, Josette; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringOnline health communities (OHC) provide various opportunities for patients with chronic or life-threatening illnesses, especially for cancer patients and survivors. A better understanding of the sentiment dynamics of patients in OHCs can help in the precise formulation of the needs during their treatment. The current study investigated the sentiment dynamics in patients’ narratives in a Breast Cancer community group (Breastcancer.org) to identify the changes in emotions, thoughts, stress, and coping mechanisms while undergoing treatment options, particularly chemotherapy, radiation, and surgery. Sentiment dynamics of users’ posts was performed using a deep learning model. A sentiment change analysis was performed to measure change in the satisfaction level of the users. The deep learning model BiLSTM with sentiment embedding features provided a better F1-score of 91.9%. Sentiment dynamics can assess the difference in satisfaction level the users acquire by interacting with other users in the forum. A comparison of the proposed model with existing models revealed the effectiveness of this methodology.Item Editorial: Machine learning for peptide structure, function, and design(Frontiers Media, 2022-09-20) Ge, Ruiquan; Dong, Chuan; Wang, Juexin; Wei, Yanjie; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringItem Targeting the Cytoskeleton and Extracellular Matrix in Cardiovascular Disease Drug Discovery(Taylor & Francis, 2022) Khomtchouk, Bohdan B.; Lee, Yoon Seo; Khan, Maha L.; Sun, Patrick; Mero, Deniel; Davidson, Michael H.; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringIntroduction: Currently, cardiovascular disease (CVD) drug discovery has focused primarily on addressing the inflammation and immunopathology aspects inherent to various CVD phenotypes such as cardiac fibrosis and coronary artery disease. However, recent findings suggest new biological pathways for cytoskeletal and extracellular matrix (ECM) regulation across diverse CVDs, such as the roles of matricellular proteins (e.g. tenascin-C) in regulating the cellular microenvironment. The success of anti-inflammatory drugs like colchicine, which targets microtubule polymerization, further suggests that the cardiac cytoskeleton and ECM provide prospective therapeutic opportunities. Areas covered: Potential therapeutic targets include proteins such as gelsolin and calponin 2, which play pivotal roles in plaque development. This review focuses on the dynamic role that the cytoskeleton and ECM play in CVD pathophysiology, highlighting how novel target discovery in cytoskeletal and ECM-related genes may enable therapeutics development to alter the regulation of cellular architecture in plaque formation and rupture, cardiac contractility, and other molecular mechanisms. Expert opinion: Further research into the cardiac cytoskeleton and its associated ECM proteins is an area ripe for novel target discovery. Furthermore, the structural connection between the cytoskeleton and the ECM provides an opportunity to evaluate both entities as sources of potential therapeutic targets for CVDs.Item MedShift: Automated Identification of Shift Data for Medical Image Dataset Curation(IEEE, 2023) Guo, Xiaoyuan; Wawira Gichoya, Judy; Trivedi, Hari; Purkayastha, Saptarshi; Banerjee, Imon; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringAutomated curation of noisy external data in the medical domain has long been demanding as AI technologies should be validated on various sources with clean annotated data. To curate a high-quality dataset, identifying variance between the internal and external sources is a fundamental step as the data distributions from different sources can vary significantly and subsequently affect the performance of the AI models. Primary challenges for detecting data shifts are – (1) access to private data across healthcare institutions for manual detection, and (2) the lack of automated approaches to learn efficient shift-data representation without training samples. To overcome the problems, we propose an automated pipeline called MedShift to detect the top-level shift samples and evaluating the significance of shift data without sharing data between the internal and external organizations. MedShift employs unsupervised anomaly detectors to learn the internal distribution and identify samples showing significant shiftness for external datasets, and compared their performance. To quantify the effects of detected shift data, we train a multi-class classifier that learns internal domain knowledge and evaluating the classification performance for each class in external domains after dropping the shift data. We also propose a data quality metric to quantify the dissimilarity between the internal and external datasets. We verify the efficacy of MedShift with musculoskeletal radiographs (MURA) and chest X-rays datasets from more than one external source. Experiments show our proposed shift data detection pipeline can be beneficial for medical centers to curate high-quality datasets more efficiently. The code can be found at https://github.com/XiaoyuanGuo/MedShift. An interface introduction video to visualize our results is available at https://youtu.be/V3BF0P1sxQE.Item Altered Caveolin-1 Dynamics Result in Divergent Mineralization Responses in Bone and Vascular Calcification(Springer, 2023-08-19) Bakhshian Nik, Amirala; Kaiser, Katherine; Sun, Patrick; Khomtchouk, Bohdan B.; Hutcheson, Joshua D.; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringIntroduction: Though vascular smooth muscle cells adopt an osteogenic phenotype during pathological vascular calcification, clinical studies note an inverse correlation between bone mineral density and arterial mineral-also known as the calcification paradox. Both processes are mediated by extracellular vesicles (EVs) that sequester calcium and phosphate. Calcifying EV formation in the vasculature requires caveolin-1 (CAV1), a membrane scaffolding protein that resides in membrane invaginations (caveolae). Of note, caveolin-1-deficient mice, however, have increased bone mineral density. We hypothesized that caveolin-1 may play divergent roles in calcifying EV formation from vascular smooth muscle cells (VSMCs) and osteoblasts (HOBs). Methods: Primary human coronary artery VSMCs and osteoblasts were cultured for up to 28 days in an osteogenic media. CAV1 expression was knocked down using siRNA. Methyl β-cyclodextrin (MβCD) and a calpain inhibitor were used, respectively, to disrupt and stabilize the caveolar domains in VSMCs and HOBs. Results: CAV1 genetic variation demonstrates significant inverse relationships between bone-mineral density (BMD) and coronary artery calcification (CAC) across two independent epidemiological cohorts. Culture in osteogenic (OS) media increased calcification in HOBs and VSMCs. siRNA knockdown of CAV1 abrogated VSMC calcification with no effect on osteoblast mineralization. MβCD-mediated caveolae disruption led to a 3-fold increase of calcification in VSMCs treated with osteogenic media (p < 0.05) but hindered osteoblast mineralization (p < 0.01). Conversely, stabilizing caveolae by calpain inhibition prevented VSMC calcification (p < 0.05) without affecting osteoblast mineralization. There was no significant difference in CAV1 content between lipid domains from HOBs cultured in OS and control media. Conclusion: Our data indicate fundamental cellular-level differences in physiological and pathophysiological mineralization mediated by CAV1 dynamics. This is the first study to suggest that divergent mechanisms in calcifying EV formation may play a role in the calcification paradox. Supplementary information: The online version contains supplementary material available at 10.1007/s12195-023-00779-7.Item The bioinformatics toolbox for circRNA discovery and analysis(Oxford University Press, 2021) Chen, Liang; Wang, Changliang; Sun, Huiyan; Wang, Juexin; Lian, Yanchun; Wang, Yan; Wong, Garry; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringCircular RNAs (circRNAs) are a unique class of RNA molecule identified more than 40 years ago which are produced by a covalent linkage via back-splicing of linear RNA. Recent advances in sequencing technologies and bioinformatics tools have led directly to an ever-expanding field of types and biological functions of circRNAs. In parallel with technological developments, practical applications of circRNAs have arisen including their utilization as biomarkers of human disease. Currently, circRNA-associated bioinformatics tools can support projects including circRNA annotation, circRNA identification and network analysis of competing endogenous RNA (ceRNA). In this review, we collected about 100 circRNA-associated bioinformatics tools and summarized their current attributes and capabilities. We also performed network analysis and text mining on circRNA tool publications in order to reveal trends in their ongoing development.Item CETP and SGLT2 inhibitor combination therapy increases glycemic control: a 2x2 factorial Mendelian randomization analysis(Frontiers Media, 2024-06-19) Khomtchouk, Bohdan B.; Sun, Patrick; Maggio, Zane A.; Ditmarsch, Marc; Kastelein, John J. P.; Davidson, Michael H.; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringIntroduction: Cholesteryl ester transfer protein (CETP) inhibitors, initially developed for treating hyperlipidemia, have shown promise in reducing the risk of new-onset diabetes during clinical trials. This positions CETP inhibitors as potential candidates for repurposing in metabolic disease treatment. Given their oral administration, they could complement existing oral medications like sodium-glucose cotransporter 2 (SGLT2) inhibitors, potentially delaying the need for injectable therapies such as insulin. Methods: We conducted a 2x2 factorial Mendelian Randomization analysis involving 233,765 participants from the UK Biobank. This study aimed to evaluate whether simultaneous genetic inhibition of CETP and SGLT2 enhances glycemic control compared to inhibiting each separately. Results: Our findings indicate that dual genetic inhibition of CETP and SGLT2 significantly reduces glycated hemoglobin levels compared to controls and single-agent inhibition. Additionally, the combined inhibition is linked to a lower incidence of diabetes compared to both the control group and SGLT2 inhibition alone. Discussion: These results suggest that combining CETP and SGLT2 inhibitor therapies may offer superior glycemic control over SGLT2 inhibitors alone. Future clinical trials should investigate the potential of repurposing CETP inhibitors for metabolic disease treatment, providing an oral therapeutic option that could benefit high-risk patients before they require injectable therapies like insulin or glucagon-like peptide-1 (GLP-1) receptor agonists.Item CrossMP: Enabling Cross-Modality Translation between Single-Cell RNA-Seq and Single-Cell ATAC-Seq through Web-Based Portal(MDPI, 2024-07-05) Lyu, Zhen; Dahal, Sabin; Zeng, Shuai; Wang, Juexin; Xu, Dong; Joshi, Trupti; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringIn recent years, there has been a growing interest in profiling multiomic modalities within individual cells simultaneously. One such example is integrating combined single-cell RNA sequencing (scRNA-seq) data and single-cell transposase-accessible chromatin sequencing (scATAC-seq) data. Integrated analysis of diverse modalities has helped researchers make more accurate predictions and gain a more comprehensive understanding than with single-modality analysis. However, generating such multimodal data is technically challenging and expensive, leading to limited availability of single-cell co-assay data. Here, we propose a model for cross-modal prediction between the transcriptome and chromatin profiles in single cells. Our model is based on a deep neural network architecture that learns the latent representations from the source modality and then predicts the target modality. It demonstrates reliable performance in accurately translating between these modalities across multiple paired human scATAC-seq and scRNA-seq datasets. Additionally, we developed CrossMP, a web-based portal allowing researchers to upload their single-cell modality data through an interactive web interface and predict the other type of modality data, using high-performance computing resources plugged at the backend.Item Cardioinformatics Advancements in Healthcare and Biotechnology(American Heart Association, 2023) Khomtchouk, Bohdan B.; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringItem IEEE Access Special Section Editorial: Smart Health Sensing and Computational Intelligence: From Big Data to Big Impacts(IEEE, 2021) Zhang, Qingxue; Piuri, Vincenzo; Clancy, Edward A.; Zhou, Dian; Penzel, Thomas; Hu, Wenchuang Walter; Zheng, Hui; Biomedical Engineering and Informatics, Luddy School of Informatics, Computing, and EngineeringSmart health big data is paving a promising way for ubiquitous health management, leveraging exciting advances in biomedical engineering technologies, such as convenient bio-sensing, health monitoring, in-home monitoring, biomedical signal processing, data mining, health trend tracking, and evidence-based medical decision support. To build and utilize the smart health big data, advanced data sensing and data mining technologies are closely coupled key enabling factors. In smart health big data innovations, challenges arise in how to informatively and robustly build the big data with advanced sensing technologies, and how to automatically and effectively decode patterns from the big data with intelligent computational methods. More specifically, advanced sensing techniques should be able to capture more modalities that can reflect rich physiological and behavioral states of humans, and enhance the signal robustness in daily wearable applications. In addition, intelligent computational techniques are required to unveil patterns deeply hidden in the data and nonlinearly convert the patterns to high-level medical insights.