Biomedical Informatics Commons^™

Descriptor:Benchmarking Secure Neural Network Evaluation Methods For Protein Sequence Classification (Idash24), Arif Harmanci, Luyao Chen, Miran Kim, Xiaoqian Jiang

Faculty, Staff and Student Publications

To uniformly test and benchmark the secure evaluation of transformer-based models, we designed the iDASH24 homomorphic encryption track dataset. The dataset comprises a protein family classification model with a transformer architecture and an example dataset that is used to build and test the secure evaluation strategies. This dataset was used in the challenge period of iDASH24 Genomic Privacy Competition, where the teams designed secure evaluation of the classification model using a homomorphic encryption scheme. Combined with the benchmarking results and companion methods, iDASH24 dataset is a unique resource that can be used to benchmark secure evaluation of neural network models.

Transdisciplinary Perspectives For Health Systems Science, Jiang Bian, Nan Liu, Shauna M Overgaard, Rema Padman, Peter A D Steel, Victoria Tiase, Clifford Alan Whitcomb, Jiajie Zhang, Yiye Zhang

Faculty, Staff and Student Publications

Health systems science uses systems thinking as part of a transdisciplinary approach that transcends traditional disciplinary boundaries. It integrates and synthesizes knowledge from multiple disciplines to address real-world problems in healthcare with pragmatic solutions. This editorial defines health systems from the perspectives of systems thinking, science, and engineering, discusses their current challenges and opportunities, and envisions how health systems science as a field can advance the continuous learning and improvement of health systems.

Prioritizing Clinically Significant Lung Cancer Somatic Mutations For Targeted Therapy Through Efficient Ngs Data Filtering System, Jinlian Wang, Hui Li, Hongfang Liu

Faculty, Staff and Student Publications

In the realm of lung cancer treatment, where genetic heterogeneity presents formidable challenges, precision oncology demands an exacting approach to identify and hierarchically sort clinically significant somatic mutations. Current Next-Generation Sequencing (NGS) data filtering pipelines, while utilizing various external databases for mutation screening, often fall short in comprehensive integration and flexibility needed to keep pace with the evolving landscape of clinical data. Our study introduces a sophisticated NGS data filtering system, which not only aggregates but effectively synergizes diverse data sources, encompassing genetic variants, gene functions, clinical evidence, and an extensive body of literature. This system is distinguished by a …

Sequencing Conversational Turns In Peer Interactions: An Integrated Approach For Evidence-Based Conversational Agent For Just-In-Time Nicotine Cravings Intervention, Tavleen Singh, Michael Truong, Kirk Roberts, Sahiti Myneni

Faculty, Staff and Student Publications

BACKGROUND: Risky health behaviors place an enormous toll on public health systems. While relapse prevention support is integrated with most behavior modification programs, the results are suboptimal. Recent advances in artificial intelligence (AI) applications provide us with unique opportunities to develop just-in-time adaptive behavior change solutions.

METHODS: In this study, we present an innovative framework, grounded in behavioral theory, and enhanced with social media sequencing and communications scenario builder to architect a conversational agent (CA) specialized in the prevention of relapses in the context of tobacco cessation. We modeled peer interaction data (n = 1000) using the taxonomy of behavior …

Linking Cancer Clinical Trials To Their Result Publications, Evan Pan, Kirk Roberts

Faculty, Staff and Student Publications

The results of clinical trials are a valuable source of evidence for researchers, policy makers, and healthcare professionals. However, online trial registries do not always contain links to the publications that report on their results, instead requiring a time-consuming manual search. Here, we explored the application of pre-trained transformer-based language models to automatically identify result-reporting publications of cancer clinical trials by computing dense vectors and performing semantic search. Models were fine-tuned on text data from trial registry fields and article metadata using a contrastive learning approach. The best performing model was PubMedBERT, which achieved a mean average precision of 0.592 …

Community Scientist Program Provides Bi-Directional Communication And Co-Learning Between Researchers And Community Members, Jessica Alvarado, Larkin L Strong, Birnur Buzcu-Guven, Leonetta B Thompson, Erica Cantu, Chelsea C Carrier, Chiamaka D Chukwu, Cassandra L Harris, Luz K Melendez, Crystal L Roberson, Angela M Ross, Sophia C Russell, Pablo Sanchez, Amirali Tahanan, Blair C Zdenek, Belinda M Reininger, Lorna H Mcneill

Faculty, Staff and Student Publications

Community involvement in research is key to translating science into practice, and new approaches to engaging community members in research design and implementation are needed. The Community Scientist Program, established at the MD Anderson Cancer Center in Houston in 2018 and expanded to two other Texas institutions in 2021, provides researchers with rapid feedback from community members on study feasibility and design, cultural appropriateness, participant recruitment, and research implementation. This paper aims to describe the Community Scientist Program and assess Community Scientists' and researchers' satisfaction with the program. We present the analysis of the data collected from 116 Community Scientists …

Human Equilibrative Nucleoside Transporter 1: Novel Biomarker And Prognostic Indicator For Patients With Gemcitabine-Treated Pancreatic Cancer, Jianchun Xiao, Fangyu Zhao, Wenhao Luo, Gang Yang, Yicheng Wang, Jiangdong Qiu, Yueze Liu, Lei You, Lianfang Zheng, Taiping Zhang

Faculty, Staff and Student Publications

AIM: This article aimed to find appropriate pancreatic cancer (PC) patients to treat with Gemcitabine with better survival outcomes by detecting hENT1 levels.

METHODS: We collected surgical pathological tissues from PC patients who received radical surgery in our hospital from September 2004 to December 2014. A total of 375 PC tissues and paired adjacent nontumor tissues were employed for the construction of 4 tissue microarrays (TMAs). The quality of the 4 TMAs was examined by HE staining. We performed immunohistochemistry analysis to evaluate hENT1 expression in the TMAs. Moreover, we detected hENT1 expression level and proved the role of hENT1 …

Visualsphere: A Web-Based Interactive Visualization System For Clinical Research Data, Shiwei Lin, Shiqiang Tao, Wei-Chun Chou, Guo-Qiang Zhang, Xiaojin Li

Faculty, Staff and Student Publications

Clinical research data visualization is integral to making sense of biomedical research and healthcare data. The complexity and diversity of data, along with the need for solid programming skills, can hinder advances in clinical research data visualization. To overcome these challenges, we introduce VisualSphere, a web-based interactive visualization system that directly interfaces with clinical research data repositories, streamlining and simplifying the visualization workflow. VisualSphere is founded on three primary component modules: Connection, Configuration, and Visualization. An end-user can set up connections to the data repositories, create charts by selecting the desired tables and variables, and render visualization dashboards generated by …

A Real-World Disproportionality Analysis Of Everolimus: Data Mining Of The Public Version Of Fda Adverse Event Reporting System, Bin Zhao, Yumei Fu, Shichao Cui, Xiangning Chen, Shu Liu, Lan Luo

Faculty, Staff and Student Publications

Background: Everolimus is an inhibitor of the mammalian target of rapamycin and is used to treat various tumors. The presented study aimed to evaluate the Everolimus-associated adverse events (AEs) through data mining of the US Food and Drug Administration Adverse Event Reporting System (FAERS).

Methods: The AE records were selected by searching the FDA Adverse Event Reporting System database from the first quarter of 2009 to the first quarter of 2022. Potential adverse event signals were mined using the disproportionality analysis, including reporting odds ratio the proportional reporting ratio the Bayesian confidence propagation neural network and the empirical Bayes geometric …

Linking Artificial Sweetener Intake With Kidney Function: Insights From Nhanes 2003-2006 And Findings From Mendelian Randomization Research, Zhuoling Ran, Yuxuan Zheng, Lin Yu, Yuxian Zhang, Zhenjiang Zhang, Huijie Li, Xuhan Li, Jing Song, Li Zhang, Ran Zhang, Chang Lu, Yang Gong, Jian Gong

Faculty, Staff and Student Publications

BACKGROUND: The current investigation examines the association between artificial sweetener (AS) consumption and the likelihood of developing chronic kidney disease (CKD), along with its impact on kidney function.

METHODS: We utilized data from the National Health and Nutrition Examination Survey from 2003-2006 to conduct covariance analysis and weighted adjusted logistic regression, aiming to assess the association between artificial sweetener intake and CKD risk, as well as kidney function indicators. Subsequently, we employed Mendelian randomization methods to validate the causal relationship between the intake of artificial sweeteners, CKD risk, and kidney function indicators. Instrumental variable analysis using inverse-variance weighting and Robust …

Modernizing And Harmonizing Regulatory Data Requirements For Genetically Modified Crops-Perspectives From A Workshop, Nicholas P Storer, Abigail R Simmons, Jordan Sottosanto, Jennifer A Anderson, Ming Hua Huang, Debbie Mahadeo, Carey A Mathesius, Mitscheli Sanches Da Rocha, Shuang Song, Ewa Urbanczyk-Wochniak

Faculty, Staff and Student Publications

Genetically modified (GM) crops that have been engineered to express transgenes have been in commercial use since 1995 and are annually grown on 200 million hectares globally. These crops have provided documented benefits to food security, rural economies, and the environment, with no substantiated case of food, feed, or environmental harm attributable to cultivation or consumption. Despite this extensive history of advantages and safety, the level of regulatory scrutiny has continually increased, placing undue burdens on regulators, developers, and society, while reinforcing consumer distrust of the technology. CropLife International held a workshop at the 16th International Society of Biosafety Research …

Vagus Nerve Stimulation For The Therapy Of Dravet Syndrome: A Systematic Review And Meta-Analysis, Shuang Chen, Man Li, Ming Huang

Faculty, Staff and Student Publications

OBJECTIVE: Dravet syndrome (DS) is a refractory developmental and epileptic encephalopathy characterized by seizures, developmental delay and cognitive impairment with a variety of comorbidities, including autism-like behavior, speech dysfunction, and ataxia. Vagus nerve stimulation (VNS) is one of the common therapies for DS. Here, we aim to perform a meta-analysis and systematic review of the efficacy of VNS in DS patients.

METHODS: We systematically searched four databases (PubMed, Embase, Cochrane and CNKI) to identify potentially eligible studies from their inception to January 2024. These studies provided the effective rate of VNS in treating patients with DS. The proportions of DS …

Dynamic Risk Score Modeling For Multiple Longitudinal Risk Factors And Survival, Cuihong Zhang, Jing Ning, Jianwen Cai, James E Squires, Steven H Belle, Ruosha Li

Faculty, Staff and Student Publications

Modeling disease risk and survival using longitudinal risk factor trajectories is of interest in various clinical scenarios. The capacity to build a prognostic model using the trajectories of multiple longitudinal risk factors, in the presence of potential dependent censoring, would enable more informed, personalized decision making. A dynamic risk score modeling framework is proposed for multiple longitudinal risk factors and survival in the presence of dependent censoring, where both events depend on participants' post-baseline clinical progression and form a competing risks structure. The model requires relatively few random effects regardless of the number of longitudinal risk factors and can therefore …

Pferm: A Fair Empirical Risk Minimization Approach With Prior Knowledge, Bojian Hou, Andrés Mondragón, Davoud Ataee Tarzanagh, Zhuoping Zhou, Andrew J Saykin, Jason H Moore, Marylyn D Ritchie, Qi Long, Li Shen

Faculty, Staff and Student Publications

Fairness is crucial in machine learning to prevent bias based on sensitive attributes in classifier predictions. However, the pursuit of strict fairness often sacrifices accuracy, particularly when significant prevalence disparities exist among groups, making classifiers less practical. For example, Alzheimer's disease (AD) is more prevalent in women than men, making equal treatment inequitable for females. Accounting for prevalence ratios among groups is essential for fair decision-making. In this paper, we introduce prior knowledge for fairness, which incorporates prevalence ratio information into the fairness constraint within the Empirical Risk Minimization (ERM) framework. We develop the Prior-knowledge-guided Fair ERM (PFERM) framework, aiming …

Feri: A Multitask-Based Fairness Achieving Algorithm With Applications To Fair Organ Transplantation, Can Li, Dejian Lai, Xiaoqian Jiang, Kai Zhang

Faculty, Staff and Student Publications

Liver transplantation often faces fairness challenges across subgroups defined by sensitive attributes such as age group, gender, and race/ethnicity. Machine learning models for outcome prediction can introduce additional biases. Therefore, we introduce Fairness through the Equitable Rate of Improvement in Multitask Learning (FERI) algorithm for fair predictions of graft failure risk in liver transplant patients. FERI constrains subgroup loss by balancing learning rates and preventing subgroup dominance in the training process. Our results show that FERI maintained high predictive accuracy with AUROC and AUPRC comparable to baseline models. More importantly, FERI demonstrated an ability to improve fairness without sacrificing accuracy. …

Outcomes For Patients With Head And Neck Sarcoma Treated Curatively With Radiation Therapy And Surgery, Alison K Yoder, Ahsan Farooqi, Devarati Mitra, J Andrew Livingston, Dejka M Araujo, Erich M Sturgis, Ryan Goepfert, Andrew J Bishop, B Ashleigh Guadagnolo

Faculty, Staff and Student Publications

PURPOSE: Soft tissue sarcomas (STSs) of the head and neck (H&N) are rare malignancies that are challenging to manage. We sought to describe the outcomes of patients treated with curative intent using combined surgery and radiation therapy (RT) for H&N STS.

METHODS AND MATERIALS: We performed a single-institution retrospective review of patients with nonmetastatic STS of the H&N who were treated from 1968 to 2020. The Kaplan-Meier method was used to estimate disease-specific survival (DSS) and local control (LC). Multivariable analyses (MVAs) were conducted using Cox proportional hazards model.

RESULTS: One hundred ninety-two patients had a median follow-up of 82 …

Engravings, Secrets, And Interpretability Of Neural Networks, Nathaniel Hobbs, Periklis A Papakonstantinou, Jaideep Vaidya

Faculty, Staff and Student Publications

This work proposes a definition and examines the problem of undetectably engraving special input/output information into a Neural Network (NN). Investigation of this problem is significant given the ubiquity of neural networks and society's reliance on their proper training and use. We systematically study this question and provide (1) definitions of security for secret engravings, (2) machine learning methods for the construction of an engraved network, (3) a threat model that is instantiated with state-of-the-art interpretability methods to devise distinguishers/attackers. In this work, there are two kinds of algorithms. First, the constructions of engravings through machine learning training methods. Second, …

Igamt: Privacy-Preserving Electronic Health Record Synthesization With Heterogeneity And Irregularity, Wenjie Wang, Pengfei Tang, Jian Lou, Yuanming Shao, Lance Waller, Yi-An Ko, Li Xiong

Faculty, Staff and Student Publications

Utilizing electronic health records (EHR) for machine learning-driven clinical research has great potential to enhance outcome predictions and treatment personalization. Nonetheless, due to privacy and security concerns, the secondary use of EHR data is regulated, constraining researchers’ access to EHR data. Generating synthetic EHR data with deep learning methods is a viable and promising approach to mitigate privacy concerns, offering not only a supplementary resource for downstream applications but also sidestepping the privacy risks associated with real patient data. While prior efforts have concentrated on EHR data synthesis, significant challenges persist: addressing the heterogeneity of features including temporal and non-temporal …

Federated Node Classification Over Distributed Ego-Networks With Secure Contrastive Embedding Sharing, Han Xie, Li Xiong, Carl Yang

Faculty, Staff and Student Publications

Federated learning on graphs (a.k.a., federated graph learning- FGL) has recently received increasing attention due to its capacity to enable collaborative learning over distributed graph datasets without compromising local clients' data privacy. In previous works, clients of FGL typically represent institutes or organizations that possess sets of entire graphs (e.g., molecule graphs in biochemical research) or parts of a larger graph (e.g., sub-user networks of e-commerce platforms). However, another natural paradigm exists where clients act as remote devices retaining the graph structures of local neighborhoods centered around the device owners (i.e., ego-networks), which can be modeled for specific graph applications …

Bridging The Gap: Rademacher Complexity In Robust And Standard Generalization, Jiancong Xiao, Ruoyu Sun, Qi Long, Weijie J Su

Faculty, Staff and Student Publications

Training Deep Neural Networks (DNNs) with adversarial examples often results in poor generalization to test-time adversarial data. This paper investigates this issue, known as adversarially robust generalization, through the lens of Rademacher complexity. Building upon the studies by Khim and Loh (2018); Yin et al. (2019), numerous works have been dedicated to this problem, yet achieving a satisfactory bound remains an elusive goal. Existing works on DNNs either apply to a surrogate loss instead of the robust loss or yield bounds that are notably looser compared to their standard counterparts. In the latter case, the bounds have a …

Privacy-Preserving Fingerprinting Against Collusion And Correlation Threats In Genomic Data, Tianxi Ji, Erman Ayday, Emre Yilmaz, Pan Li

Faculty, Staff and Student Publications

Sharing genomic databases is critical to the collaborative research in computational biology. A shared database is more informative than specific genome-wide association studies (GWAS) statistics as it enables “do-it-yourself” calculations. Genomic databases involve intellectual efforts from the curator and sensitive information of participants, thus in the course of data sharing, the curator (database owner) should be able to prevent unauthorized redistributions and protect individuals’ genomic data privacy. As it becomes increasingly common for a single database be shared with multiple recipients, the shared genomic database should also be robust against collusion attack, where multiple malicious recipients combine their individual copies …

Fairness-Aware Estimation Of Graphical Models, Zhuoping Zhou, Davoud Ataee Tarzanagh, Bojian Hou, Qi Long, Li Shen

Faculty, Staff and Student Publications

This paper examines the issue of fairness in the estimation of graphical models (GMs), particularly Gaussian, Covariance, and Ising models. These models play a vital role in understanding complex relationships in high-dimensional data. However, standard GMs can result in biased outcomes, especially when the underlying data involves sensitive characteristics or protected groups. To address this, we introduce a comprehensive framework designed to reduce bias in the estimation of GMs related to protected attributes. Our approach involves the integration of the pairwise graph disparity error and a tailored loss function into a nonsmooth multi-objective optimization problem, striving to achieve fairness across …

Facilitating Clinical Information Extraction With Synthetic Data And Ontology Using Large Language Models, Yan Hu, Huan He, Qingyu Chen, Xiaoqian Jiang, Kirk Roberts, Hua Xu

Faculty, Staff and Student Publications

The rapid growth of unstructured clinical text in electronic health records necessitates robust information extraction systems, yet their development is hindered by the scarcity of high-quality annotated data. This study explores the potential of large language models to generate synthetic data for clinical named entity recognition and examines its impact on model performance. We propose a novel framework that integrates self-verified synthetic data generation with domain-specific semantic mapping using SNOMED-CT. By leveraging GPT-4o-mini for synthetic data creation and refining its quality through iterative verification and anomaly detection, we systematically evaluate the influence of synthetic data quality and quantity on fine-tuning …

Narrative Feature Or Structured Feature? A Study Of Large Language Models To Identify Cancer Patients At Risk Of Heart Failure, Ziyi Chen, Mengyuan Zhang, Mustafa Mohammed Ahmed, Yi Guo, Thomas J George, Jiang Bian, Yonghui Wu

Faculty, Staff and Student Publications

Cancer treatments are known to introduce cardiotoxicity, negatively impacting outcomes and survivorship. Identifying cancer patients at risk of heart failure (HF) is critical to improving cancer treatment outcomes and safety. This study examined machine learning (ML) models to identify cancer patients at risk of HF using electronic health records (EHRs), including traditional ML, Time-Aware long short-term memory (T-LSTM), and large language models (LLMs) using novel narrative features derived from the structured medical codes. We identified a cancer cohort of 12,806 patients from the University of Florida Health, diagnosed with lung, breast, and colorectal cancers, among which 1,602 individuals developed HF …

Descriptor: Synthetic Genomic Dataset With Diverse Ancestry (Syngen6), Xinyue Wang, Sitao Min, Jaideep Vaidya

Faculty, Staff and Student Publications

Advancements in genomic analysis techniques and data-driven research are driving precision medicine. However, in many cases, these advances are not equitable and do not help all subpopulations, since many existing genomic datasets lack diversity, limiting their applicability for studying populations beyond those of European ancestry. Thus, to advance genomic analysis and to allow for a fair benchmarking of novel proposed approaches, there is a significant demand for balanced and representative datasets. To address this issue, we developed,

The Baseline Hemoglobin Level Is A Positive Biomarker For Immunotherapy Response And Can Improve The Predictability Of Tumor Mutation Burden For Immunotherapy Response In Cancer, Yin He, Tong Ren, Chengfei Ji, Li Zhao, Xiaosheng Wang

Faculty, Staff and Student Publications

Purpose: Because only a subset of cancer patients can benefit from immunotherapy, identifying predictive biomarkers of ICI therapy response is of utmost importance.

Methods: We analyzed the association between hemoglobin (HGB) levels and clinical outcomes in 1,479 ICIs-treated patients across 16 cancer types. We explored the dose-dependent associations between HGB levels and survival and immunotherapy response using the spline-based cox regression analysis. Furthermore, we investigated the associations across subgroups of patients with different clinicopathological characteristics, treatment programs and cancer types using the bootstrap resampling method.

Results: HGB levels correlated positively with clinical outcomes in cancer patients receiving immunotherapy but not …

Federated Multiple Imputation For Variables That Are Missing Not At Random In Distributed Electronic Health Records, Yi Lian, Xiaoqian Jiang, Qi Long

Faculty, Staff and Student Publications

Large electronic health records (EHR) have been widely implemented and are available for research activities. The magnitude of such databases often requires storage and computing infrastructure that are distributed at different sites. Restrictions on data-sharing due to privacy concerns have been another driving force behind the development of a large class of distributed and/or federated machine learning methods. While missing data problem is also present in distributed EHRs, albeit potentially more complex, distributed multiple imputation (MI) methods have not received as much attention. An important advantage of distributed MI, as well as distributed analysis, is that it allows researchers to …

Large Language Models Struggle In Token-Level Clinical Named Entity Recognition, Qiuhao Lu, Rui Li, Andrew Wen, Jinlian Wang, Liwei Wang, Hongfang Liu

Faculty, Staff and Student Publications

Large Language Models (LLMs) have revolutionized various sectors, including healthcare where they are employed in diverse applications. Their utility is particularly significant in the context of rare diseases, where data scarcity, complexity, and specificity pose considerable challenges. In the clinical domain, Named Entity Recognition (NER) stands out as an essential task and it plays a crucial role in extracting relevant information from clinical texts. Despite the promise of LLMs, current research mostly concentrates on document-level NER, identifying entities in a more general context across entire documents, without extracting their precise location. Additionally, efforts have been directed towards adapting ChatGPTfor token-level …

Barriers And Facilitators Of Digital Health Use For Self-Management Of Hypertensive Disorders By Black Pregnant Women, Morgan A Foreman, Angela Ross, Angela P H Burgess, Sahiti Myneni, Amy Franklin

Faculty, Staff and Student Publications

Although digital health tools are increasingly common for managing health conditions, these applications are often developed without consideration of differences across user populations. A reproducible framework is needed to support tailoring applications to include cultural considerations, potentially leading to better adoption and more effective use. As a first step, this study captures a snapshot of Black women's barriers and facilitators in using digital health products for self-management of hypertensive disorders of pregnancy (HDP). One-on-one semi-structured interviews were conducted with 17 Black pregnant women with HDP. We established a unique model for cultural tailoring with these experiences using Black feminist theory …

Enhancing Patient Medication Safety At Home: A Patient-Facing Technology Architecture Integrating Redcap, Visualization Dashboards, And An Ai- Driven Chatbot, Yue Yu, Yuheng Shi, Eric Yang, Katie Gahn, Heidi Mason, Yun Jiang, Yang Gong

Faculty, Staff and Student Publications

This work demonstrates a novel architecture for a patient-facing technology (PFT) that supports patients with cancer in self-managing medication concerns and symptoms after care transitions back home. Patient-generated data are collected and stored using the framework of Research Electronic Data Capture (REDCap), which serves as a key component of the architecture. Individual patient and administrator dashboards are integrated with REDCap to visualize medication or symptom-related data and generate customized report summaries. Additionally, a Q&A chatbot, using a retrieval augmented generation (RAG) framework, is intended to integrate into the architecture to further enhance the interactivity of the PFT. The proposed architecture …