Biomedical Informatics Commons^™

A Bayesian Deep Segmentation Framework For Glioblastoma Tumor Segmentation Using Follow-Up Mris, Tanjida Kabir, Kang-Lin Hsieh, Luis Nunez, Yu-Chun Hsu, Juan C Rodriguez Quintero, Octavio Arevalo, Kangyi Zhao, Jay-Jiguang Zhu, Roy F Riascos, Mahboubeh Madadi, Xiaoqian Jiang, Shayan Shams

Faculty, Staff and Student Publications

Background: Glioblastoma (GBM) is the most common malignant brain tumor with an abysmal prognosis. Since complete tumor cell removal is impossible due to the infiltrative nature of GBM, accurate measurement is paramount for GBM assessment. Preoperative magnetic resonance images (MRIs) are crucial for initial diagnosis and surgical planning, while follow-up MRIs are vital for evaluating treatment response. The structural changes in the brain caused by surgical and therapeutic measures create significant differences between preoperative and follow-up MRIs. In clinical research, advanced deep learning models trained on preoperative MRIs are often applied to assess follow-up scans, but their effectiveness in this …

Examining Educational And Career Transition Points Among A Diverse, Virtual Mentoring Network, Erika L Thompson, Toufeeq Ahmed Syed, Zainab Latif, Katie Stinson, Damaris Javier, Gabrielle Saleh, Jamboor K Vishwanatha

Faculty, Staff and Student Publications

Given the differences in trajectory for under-represented minorities in biomedical careers, we sought to explore how a virtual mentoring program, the National Research Mentoring Network (NRMN), and its platform (MyNRMN), may facilitate transitions in the science, technology, engineering, mathematics, and medicine (STEMM) pipeline. The purpose of this study was to describe how the size of an MyNRMN member’s mentoring network and level of engagement correlate with academic and career transitions. We examined MyNRMN platform user data from March 2020 to May 2021 (n = 2993). Logistic regression estimated the odds of a career or academic transition related to NRMN …

Key Considerations For Combination Therapy In Alzheimer’S Clinical Trials: Perspectives From An Expert Advisory Board Convened By The Alzheimer’S Drug Discovery Foundation, Jeffrey Cummings, Michael Gold, Mark Mintun, Michael Irizarry, Andrew Von Eschenbach, Suzanne Hendrix, Donald Berry, Cristina Sampaio, Kaycee Sink, Jaren Landen, Miia Kivipelto, Michael Grundman, Steven E Arnold, Allan Green, Katherine Partrick, Laura Nisenbaum, Aaron Burstein, Howard Fillit

Faculty, Staff and Student Publications

There is growing consensus in the Alzheimer's community that combination therapy will be needed to maximize therapeutic benefits through the course of the disease. However, combination therapy raises complex questions and decisions for study sponsors, from preclinical research through clinical trial design to regulatory, statistical, and operational considerations. In January 2024, the Alzheimer's Drug Discovery Foundation convened an expert advisory board to discuss the key considerations in each of these areas. Experts agreed on the need to prioritize a combination therapy approach that encompasses a wide range of targets associated with aging and the underlying biology of Alzheimer's disease. Progress …

A Serum Biomarker Panel And Miniarray Detection System For Tracking Disease Activity And Flare Risk In Lupus Nephritis, Chenling Tang, Gongjun Tan, Aygun Teymur, Jiechang Guo, Arturo Haces-Garcia, Weihang Zhu, Richard Williams, Jing Ning, Ramesh Saxena, Tianfu Wu

Faculty, Staff and Student Publications

Introduction: Lupus nephritis (LN) leads to end stage renal disease (ESRD), and early diagnosis and disease monitoring of LN could significantly reduce the risk. however, there is not such a system clinically. In this study we aim to develop a biomarker-panel based point-of-care system for LN.

Methods: Immunoassay screening combined with genomic expression databases and machine learning techniques was used to identify a biomarker panel of LN. A quantitative biomarker-panel mini-array (BPMA) system was developed and the sensitivity, specificity, reproducibility, and stability of the were examined. The performance of BPMA in disease monitoring was validated with machine models using a …

Effect Of Digital Health Coaching On Self-Efficacy And Patient-Reported Outcomes In Individuals With Acute Myeloid And Chronic Lymphocytic Leukemia: A Pilot Randomized Controlled Trial, Jennifer Marvin-Peek, Valerie Shelton, Kelly Brassil, Bryan Fellman, Austin Barr, Kelly Sharon Chien, Danielle Hammond, Mahesh Swaminathan, Nitin Jain, William Wierda, Alessandra Ferrajoli, Courtney Dinardo

Faculty, Staff and Student Publications

Introduction: Promotion of self-efficacy can enhance engagement with health care and treatment adherence in patients with cancer. We report the outcomes of a pilot trial of a digital health coach intervention in patients with leukemia with the aim of improving self-efficacy.

Methods: Adult patients with newly diagnosed acute myeloid leukemia (AML) and chronic lymphocytic leukemia (CLL) were randomized 1:1 to a digital health coach intervention or standard of care. The primary outcome of self-efficacy was measured by the Cancer Behavior Inventory (CBI) score.

Results: A total of 147 patients (37 AML, 110 CLL) were enrolled from July 2020 to December …

Rethinking Parkinson's Disease Genetics In The Precision Medicine Era: Why Genomic Diversity Matters?, Camilla Teixeira Pinheiro Gusmão, Giselli Scaini, Everton Ferreira De Souza, Rafael Antônio Vicente Lacerda, Matheus De Almeida Costa, Raja Mehanna, João Quevedo, Howard Lopes Ribeiro Junior

Faculty, Staff and Student Publications

No abstract provided.

Topss: Tolerability Of Transcranial Direct Current Stimulation In Pediatric Stroke Survivors, Stuart Fraser, Anna Clearman, Melika Abrahams, Bernadette Gillick, Tia Lal, Sean Savitz, Nuray Yozbatiran

Faculty, Staff and Student Publications

Background: Transcranial direct current stimulation is a non-invasive neuromodulation technique with emerging therapeutic potential in neurodevelopmental conditions. While childhood-onset stroke survivors frequently experience long-term motor impairment, there are very few studies examining the safety and feasibility of transcranial direct current stimulation in this population.

Objective: To evaluate the safety, feasibility, and tolerability of bihemispheric transcranial direct current stimulation paired with occupational therapy in children and adolescents with chronic hemiparesis following childhood-onset arterial ischemic stroke or intracranial hemorrhage.

Methods: In this single-arm, open-label pilot study, five participants aged 6-19 years of age received five daily sessions of transcranial direct current stimulation …

Smoking Habit And Long-Term Colorectal Cancer Incidence By Exome-Wide Mutational And Neoantigen Loads: Evidence Based On The Prospective Cohort Incident-Tumour Biobank Method, Tsuyoshi Hamada, Tomotaka Ugai, Carino Gurjao, Satoko Ugai, Xuehong Zhang, Koichiro Haruki, Yasutoshi Takashima, Naohiko Akimoto, Mai Chan Lau, Kosuke Matsuda, Nobuhiro Nakazawa, Mayu Higashioka, Satoshi Miyahara, Keisuke Kosumi, Yohei Masugi, Li Liu, Yin Cao, Daniel Nevo, Molin Wang, Reiko Nishihara, Sachet A Shukla, Catherine J Wu, Levi A Garraway, Jeffrey A Meyerhardt, Edward L Giovannucci, Jonathan A Nowak, Charles S Fuchs, Andrew T Chan, Mingyang Song, Marios Giannakis, Shuji Ogino

Faculty, Staff and Student Publications

Objective: To test the hypothesis that the association of smoking with long-term colorectal cancer incidence may be stronger for tumours with higher mutational and neoantigen loads.

Methods and analysis: In the Nurses' Health Study (1980-2012) and the Health Professionals Follow-up Study (1986-2012), our novel prospective cohort incident-tumour biobank method (PCIBM) used 3053 incident colorectal carcinoma cases including 752 cases with whole-exome sequencing data. Using the multivariable duplication-method Cox regression model with the inverse probability weighting to adjust for the selection bias due to tissue availability, we assessed a differential association of cigarette smoking with colorectal carcinoma incidence by an exome-wide …

Reusable Generic Clinical Decision Support System Module For Immunization Recommendations In Resource-Constraint Settings, Samuil Orlioglu, Akash Shanmugan Boobalan, Kojo Abanyie, Richard D Boyce, Hua Min, Yang Gong, Dean F Sittig, Paul Biondich, Adam Wright, Christian Nøhr, Timothy Law, David Robinson, Arild Faxvaag, Nina Hubig, Ronald Gimbel, Lior Rennert, Xia Jing

Faculty, Staff and Student Publications

Clinical decision support systems (CDSS) are routinely employed in clinical settings to improve quality of care, ensure patient safety, and deliver consistent medical care. However, rule-based CDSS, currently available, do not feature reusable rules. In this study, we present CDSS with reusable rules. Our solution includes a common CDSS module, electronic medical record (EMR) specific adapters, CDSS rules written in the clinical quality language (CQL) (derived from CDC immunization recommendations), and patient records in fast healthcare interoperability resources (FHIR) format. The proposed CDSS is entirely browser-based and reachable within the user's EMR interface at the client-side. This helps to avoid …

Exploring The Inequitable Impact Of Data Missingness On Fairness In Machine Learning, Sitao Min, Hafiz Asif, Jaideep Vaidya

Faculty, Staff and Student Publications

Today, data-driven models and artificial intelligence / machine learning underlie decision making in almost all aspects of society. However, significant concerns have been raised over the fairness of such models. While various aspects of algorithmic fairness have been studied, the effect of missing data on fairness remains understudied. This is a significant problem since data in real-world settings is almost never complete, and may often suffer from systemic missingness. This article systematically evaluates how missing data, particularly when correlated with protected classes and outcome variables, affects the fairness of classifiers. Utilizing a comprehensive framework covering various missing data patterns, rates, …

Disposition Outcomes Following Prehospital Use Of Naloxone In A Large Metropolitan City In The United States, James R Langabeer, Christine Bakos-Block, A Sarah Cohen, Ishmam Alam, Bhanumathi Gopal, Marylou Cardenas-Turanzas, Arlo F Weltge, David Persse, Tiffany Champagne-Langabeer

Faculty, Staff and Student Publications

Objectives: During a drug overdose, research suggests individuals may not call 9-1-1 out of fear of criminal justice concerns. Of those that call, research is inconclusive about the disposition of the emergency transport. We evaluated transport outcomes for adults with opioid-related overdose in the Emergency Medical Services (EMS) of a large metropolitan city in the United States.

Methods: We reviewed the EMS incident report database from the patient care record system for years 2018 to 2022. We queried all records, searching for relevant terms, and two reviewers cross-checked the database to identify cases that did not result in death at …

A Scoping Review Of Patient Safety Checklists In Pediatrics, Kawtar Zouaidi, Tate W Miner, Muhammad F Walji, Kristin N Ray, Elsbeth Kalenderian, Donald B Rindal, Katie J Suda

Faculty, Staff and Student Publications

Objective: To examine the utilization and effectiveness of safety checklists in pediatric clinical care.

Methods: A comprehensive literature search was conducted using Medline to identify studies related to the development and/or implementation of patient safety checklists in pediatrics. All study designs were included for citations published through September 2023.

Results: Following abstract and full-text screening, 74 studies remained for data extraction and analysis. Pediatric surgery emerged as the main setting for checklists use (n = 35), followed by Intensive Care Units (n = 21), and Emergency Departments (n = 9). Of the 74 reviewed papers, 37 (50%) designed and developed …

Defining Spine Cancer Pain Syndromes: A Systematic Review And Proposed Terminology, Markian Pahuta, Ilya Laufer, Sheng-Fu Larry Lo, Stefano Boriani, Charles Fisher, Nicolas Dea, Michael H Weber, Dean Chou, Arjun Sahgal, Laurence Rhines, Jeremy Reynolds, Aron Lazary, Alessandro Gasbarrinni, Jorrit-Jan Verlaan, Ziya Gokaslan, Chetan Bettegowda, Mohamed Sarraj, Ori Barzilai, Ao Spine Knowledge Forum Tumor

Faculty, Staff and Student Publications

STUDY DESIGN: Systematic Review.

OBJECTIVES: Formalized terminology for pain experienced by spine cancer patients is lacking. The common descriptors of spine cancer pain as mechanical or non-mechanical is not exhaustive. Misdiagnosed spinal pain may lead to ineffective treatment recommendations for cancer patients.

METHODS: We conducted a systematic review of pain terminology that may be relevant to spinal oncology patients. We provide a comprehensive and unbiased summary of the existing evidence, not limited to the spine surgery literature, and subsequently consolidate these data into a practical, clinically relevant nomenclature for spine oncologists.

RESULTS: Our literature search identified 3515 unique citations. Through …

The Influence Of The Microbiome On Radiotherapy And Dna Damage Responses, Aadil Sheikh, Michael A Curran

Faculty, Staff and Student Publications

Colorectal cancer (CRC) is one of the most prevalent cancers in terms of diagnosis and mortality. Radiotherapy (RT) remains a mainstay of CRC therapy. As RT relies on DNA damage to promote tumor cell death, the activity of cellular DNA damage repair pathways can modulate cancer sensitivity to therapy. The gut microbiome has been shown to influence intestinal health and is independently associated with CRC development, treatment responses and outcomes. The microbiome can also modulate responses to CRC RT through various mechanisms such as community structure, toxins and metabolites. In this review we explore the use of RT in the …

Patient-Reported Outcomes During Pelvic Radiation Therapy: A Secondary Analysis On Sexual Function From Nrg-Rtog 1203, Kelsey L Corrigan, Rebecca Paulus, Ann H Klopp, Lari B Wenzel, Anamaria R Yeung, J Spencer Thompson, Desiree E Doncals, Vijayananda Kundapur, Nancy H Wiggers, Dasarahally S Mohan, Sharad A Ghamande, Shannon N Westin, Kara L Schnarr, Michael L Haas, David K Gaffney, Steven E Waggoner, Pamela J Vanderwall, Noha T Jastaniyah, Stephanie L Pugh, Lisa A Kachnic

Faculty, Staff and Student Publications

Purpose: NRG-RTOG 1203 reported that intensity-modulated radiation therapy (IMRT) reduced patient-reported GI toxicities in patients with cervical/endometrial cancer receiving postoperative RT, compared with 3-dimensional conformal radiation therapy (3DRT). We conducted a secondary analysis of patient-reported sexual function (PR-SF) among treatment groups to identify factors associated with sexual dysfunction.

Methods and materials: Patients on NRG-RTOG 1203 were randomly assigned to 3DRT versus IMRT and completed Patient-Reported Outcomes (PRO)-Common Terminology Criteria for Adverse Events (CTCAE) and FACT-Cx surveys at baseline, week 5 of RT, and at 4-6 weeks, 1 year, and 3 years after RT. Patient responses to FACT-Cx sexual function questions …

Reproducibility And Repeatability Of 18f-(2s, 4r)-4-Fluoroglutamine Pet Imaging In Preclinical Oncology Models, Gregory D Ayers, Allison S Cohen, Seong-Woo Bae, Xiaoxia Wen, Alyssa Pollard, Shilpa Sharma, Trey Claus, Adria Payne, Ling Geng, Ping Zhao, Mohammed Noor Tantawy, Seth T Gammon, H Charles Manning

Faculty, Staff and Student Publications

Introduction: Measurement of repeatability and reproducibility (R&R) is necessary to realize the full potential of positron emission tomography (PET). Several studies have evaluated the reproducibility of PET using 18F-FDG, the most common PET tracer used in oncology, but similar studies using other PET tracers are scarce. Even fewer assess agreement and R&R with statistical methods designed explicitly for the task. 18F-(2S, 4R)-4-fluoro-glutamine (18F-Gln) is a PET tracer designed for imaging glutamine uptake and metabolism. This study illustrates high reproducibility and repeatability with 18F-Gln for in vivo research.

Methods: Twenty mice bearing colorectal cancer cell line xenografts were injected with ~9 …

Patterns Of Loco-Regional Progression And Patient Outcomes After Definitive-Dose Radiation Therapy For Anaplastic Thyroid Cancer, Julianna K Bronk, Alexander Augustyn, Abdallah S R Mohamed, C David Fuller, Adam S Garden, Amy C Moreno, Anna Lee, William H Morrison, Jack Phan, Jay P Reddy, David I Rosenthal, Michael T Spiotto, Steven J Frank, Ramona Dadu, Naifa Busaidy, Mark Zafereo, Jennifer R Wang, Anastasios Maniakas, Renata Ferrarotto, Priyanka C Iyer, Maria E Cabanillas, G Brandon Gunn

Faculty, Staff and Student Publications

Background: The aim of this study is to characterize the patterns of loco-regional progression (LRP) and outcomes after definitive-dose intensity modulated radiation therapy (IMRT) for anaplastic thyroid cancer (ATC) with macroscopic neck disease at the time of IMRT.

Methods: Disease/treatment characteristics and outcomes for patients with unresected or incompletely resected ATC who received IMRT (≥45 Gy) were retrospectively reviewed. For those with LRP after IMRT, progressive/recurrent gross tumor volumes (rGTV) were contoured on diagnostic CTs and co-registered with initial planning CTs using deformable image registration. rGTVs were classified based on established spatial/dosimetric criteria.

Results: Forty patients treated between 2010-2020 formed …

Selective Inhibition Of Canonical Stat3 Signaling Suppresses K-Ras Mutant Lung Tumorigenesis And Reinvigorates Anti-Tumor Immunity, Michael J Clowers, Zahraa Rahal, Sung-Nam Cho, Avantika Krishna, Bo Yuan, Leticia G Hamana Zorrilla, T Kris Eckols, Moses M Kasembeli, Samuel Liu, Stephen Peng, Marco Ramos-Castaneda, Annamarie L Thompson, Carlos Ignacio Rodriguez Reyna, Katherine E Larsen, Maria T Grimaldo, Shanshan Deng, Nastaran Karimi, Cody Chou, Walter V Velasco, Melody Zarghooni, Sayan Alekseev, Luisa M Solis Soto, Edwin J Ostrin, Humam Kadara, Suhendan Ekmekcioglu, David J Tweardy, Seyed Javad Moghaddam

Faculty, Staff and Student Publications

Introduction: K-ras mutant lung adenocarcinoma (KM-LUAD) is a difficult-to-treat cancer subtype in which chronic inflammation pervades the tumor immune microenvironment (TIME). Pro-inflammatory pathways dampen the response to treatments, including immune checkpoint inhibitors, necessitating therapies that target this inflammatory signaling network in the TIME. One of the lynchpins of chronic inflammation in KM-LUAD is signal transducer and activator of transcription 3 (STAT3).

Methods: Here, we tested the anti-tumor and early immunotherapeutic efficacy of TTI-101, a selective small-molecule inhibitor of canonical STAT3 signaling, in a K-rasG12D mutant lung cancer mouse model (CC-LR).

Results: Treatment of CC-LR mice with TTI-101 resulted in reduced …

Dedifferentiated Liposarcomas Treated With Immune Checkpoint Blockade: The Md Anderson Experience, Madeline B Torres, Cheuk Hong Leung, Marianne Zoghbi, Rossana Lazcano, Davis Ingram, Khalida Wani, Emily Z Keung, M Alejandra Zarzour, Christopher P Scally, Kelly K Hunt, Anthony Conley, Andrew J Bishop, B Ashleigh Guadagnolo, Ahsan Farooqi, Devarati Mitra, Alison K Yoder, Michael S Nakazawa, Dejka Araujo, Andrew Livingston, Ravin Ratan, Shreyaskumar Patel, Vinod Ravi, Alexander J Lazar, Christina L Roland, Neeta Somaiah, Elise F Nassif Haddad

Faculty, Staff and Student Publications

Background: Dedifferentiated liposarcoma (DDLPS) is one of the most common types of soft tissue sarcoma (STS) characterized by liposarcomatous differentiation and a predilection for the retroperitoneum. Despite the growing number of histology-specific immune checkpoint blockade (ICB) trials in STS, it is still difficult to identify the radiographic objective response rate (ORR) for DDLPS in the real world setting. This study aimed to evaluate the ORR and survival of patients with DDLPS treated with ICB at a single center.

Methods: We conducted a retrospective study of 31 patients with pathologically confirmed DDLPS treated with ICB at MD Anderson Cancer Center between …

A Plain Language Summary Of The Final Analysis Of The Griffin Study Of Daratumumab Plus Lenalidomide, Bortezomib, And Dexamethasone For People With Newly Diagnosed Multiple Myeloma, Peter M Voorhees, Douglas W Sborov, Jacob Laubach, Jonathan L Kaufman, Brandi Reeves, Cesar Rodriguez, Rebecca Silbermann, Luciano J Costa, Larry D Anderson, Nitya Nathwani, Nina Shah, Naresh Bumma, Yvonne A Efebera, Sarah A Holstein, Caitlin Costello, Andrzej Jakubowiak, Tanya M Wildes, Robert Z Orlowski, Kenneth H Shain, Andrew J Cowan, Shira Dinner, Katharine S Gries, Huiling Pei, Annelore Cortoos, Sharmila Patel, Thomas S Lin, Saad Z Usmani, Paul G Richardson

Faculty, Staff and Student Publications

No abstract provided.

Grape-Pi: Graph-Based Neural Networks For Enhanced Protein Identification In Proteomics Pipelines, Chunhui Gu, Seyyed Mahmood Ghasemi, Yining Cai, Johannes F Fahrmann, James P Long, Hiroyuki Katayama, Chong Wu, Jody Vykoukal, Jennifer B Dennison, Samir Hanash, Kim-Anh Do, Ehsan Irajizad

Faculty, Staff and Student Publications

Motivation: Protein identification via mass spectrometry (MS) is the primary method for untargeted protein detection. However, the identification process is challenging due to data complexity and the need to control false discovery rates (FDR) of protein identification. To address these challenges, we developed a graph neural network (GNN)-based model, Graph Neural Network using Protein-Protein Interaction for Enhancing Protein Identification (Grape-Pi), which is applicable to all proteomics pipelines. This model leverages protein-protein interaction (PPI) data and employs two types of message-passing layers to integrate evidence from both the target protein and its interactors, thereby improving identification accuracy.

Results: Grape-Pi achieved significant …

A Critical Assessment Of Artificial Intelligence In Magnetic Resonance Imaging Of Cancer, Chengyue Wu, Meryem Abbad Andaloussi, David A Hormuth, Ernesto A B F Lima, Guillermo Lorenzo, Casey E Stowers, Sriram Ravula, Brett Levac, Alexandros G Dimakis, Jonathan I Tamir, Kristy K Brock, Caroline Chung, Thomas E Yankeelov

Faculty, Staff and Student Publications

Given the enormous output and pace of development of artificial intelligence (AI) methods in medical imaging, it can be challenging to identify the true success stories to determine the state-of-the-art of the field. This report seeks to provide the magnetic resonance imaging (MRI) community with an initial guide into the major areas in which the methods of AI are contributing to MRI in oncology. After a general introduction to artificial intelligence, we proceed to discuss the successes and current limitations of AI in MRI when used for image acquisition, reconstruction, registration, and segmentation, as well as its utility for assisting …

Inhibition Of Nitric Oxide Synthase Transforms Carotid Occlusion-Mediated Benign Oligemia Into De Novo Large Cerebral Infarction, Ha Kim, Jinyong Chung, Jeong Wook Kang, Dawid Schellingerhout, Soo Ji Lee, Hee Jeong Jang, Inyeong Park, Taesu Kim, Dong-Seok Gwak, Ji Sung Lee, Sung-Ha Hong, Kang-Hoon Je, Hee-Joon Bae, Joohon Sung, Eng H Lo, James Faber, Cenk Ayata, Dong-Eog Kim

Faculty, Staff and Student Publications

No abstract provided.

Upregulation Of Delta Opioid Receptor By Meningeal Interleukin-10 Prevents Relapsing Pain, Kufreobong E Inyang, Jaewon Sim, Kimberly B Clark, Matan Geron, Karli Monahan, Christine Evans, Patrick O'Connell, Sophie Laumet, Bo Peng, Jiacheng Ma, Cobi J Heijnen, Robert Dantzer, Grégory Scherrer, Annemieke Kavelaars, Matthew Bernard, Yasser A Aldhamen, Joseph K Folger, Alexis Bavencoffe, Geoffroy Laumet

Faculty, Staff and Student Publications

Chronic pain often includes periods of transient amelioration and even remission that alternate with severe relapsing pain. While most research on chronic pain has focused on pain development and maintenance, there is a critical unmet need to better understand the mechanisms that underlie pain remission and relapse. We found that interleukin (IL)-10, a pain resolving cytokine, is produced by resident macrophages in the spinal meninges during remission from pain and signaled to IL-10 receptor-expressing sensory neurons. Using unbiased RNA-sequencing, we identified that IL-10 upregulated expression and antinociceptive activity of δ-opioid receptor (δOR) in the dorsal root ganglion. Genetic or pharmacological …

Privacy-Preserving Collaborative Population Stratification With Dynamic Algorithm And Hyperparameter Selection, Maryam Ghasemian, Lynette Hammond Gerido, Erman Ayday

Faculty, Staff and Student Publications

We present a privacy-preserving selection layer for collaborative population stratification under 𝜖-local differential privacy (LDP). Rather than fixing a single pipeline (e.g., PCA+K-Means with preset 𝐾), our framework lets parties choose among three DP pipelines: PCA→Noise, Noise→PCA, and Noise-Only, according to their resources, and has an honest-but-curious server aggregate only DP shares to automatically select the clustering algorithm (K-Means, GMM, or Hierarchical) and 𝐾 that maximize internal metrics (Silhouette, Calinski–Harabasz, Davies–Bouldin). Because selection operates on DP data, it adds no further privacy loss. On openSNP (942 samples, 28,396 SNPs), the PCA-augmented pipelines yield higher utility …

On The Empirical Power Of Goodness-Of-Fit Tests In Watermark Detection, Weiqing He, Xiang Li, Tianqi Shang, Li Shen, Weijie Su, Qi Long

Faculty, Staff and Student Publications

Large language models (LLMs) raise concerns about content authenticity and integrity because they can generate human-like text at scale. Text watermarks, which embed detectable statistical signals into generated text, offer a provable way to verify content origin. Many detection methods rely on pivotal statistics that are i.i.d. under human-written text, making goodness-of-fit (GoF) tests a natural tool for watermark detection. However, GoF tests remain largely underexplored in this setting. In this paper, we systematically evaluate eight GoF tests across three popular watermarking schemes, using three open-source LLMs, two datasets, various generation temperatures, and multiple post-editing methods. We find that general …

Privacy-Preserving Verification Of Ml Preprocessing Via Model Behavior Indicators, Wenbiao Li, Anisa Halimi, Jaideep Vaidya, Xiaoqian Jiang, Erman Ayday

Faculty, Staff and Student Publications

We present a privacy-preserving framework to verify whether a declared data preprocessing pipeline was correctly applied before training a machine learning model on sensitive data. The verifier has only black-box query access to the model and combines three behavior indicators: shift in prediction accuracy, Kullback-Leibler (KL) divergence between output distributions, and explanation vectors from Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP). The method requires neither the original training records nor ground-truth labels. It supports two tasks: (i) a binary decision on correctness and (ii) a multi-class diagnosis identifying which step is missing. Experiments on three tabular datasets …

Safeguarding Privacy In Genome Research: A Comprehensive Framework For Authors, Maryam Ghasemian, Lynette Hammond Gerido, Erman Ayday

Faculty, Staff and Student Publications

As genomic research continues to advance, sharing of genomic data and research outcomes has become increasingly important for fostering collaboration and accelerating scientific discovery. However, such data sharing must be balanced with the need to protect the privacy of individuals whose genetic information is being utilized. This paper presents a bidirectional framework for evaluating privacy risks associated with data shared (both in terms of summary statistics and research datasets) in genomic research papers, particularly focusing on re-identification risks such as membership inference attacks (MIA). The framework consists of a structured workflow that begins with a questionnaire designed to capture researchers' …

Not Fully Synthetic: Llm-Based Hybrid Approaches Towards Privacy-Preserving Clinical Note Sharing, Atiquer Rahman Sarkar, Yao-Shun Chuang, Xiaoqian Jiang, Noman Mohammed

Faculty, Staff and Student Publications

The publication and sharing of clinical notes are crucial for healthcare research and innovation. However, privacy regulations such as HIPAA and GDPR pose significant challenges. While de-identification techniques aim to remove protected health information, they often fall short of achieving complete privacy protection. Similarly, the current state of synthetic clinical note generation can lack nuance and content coverage. To address these limitations, we propose an approach that combines de-identification, filtration, and synthetic clinical note generation. Variations of this approach currently retain 36%-61% of the original note's content and fill the remaining gaps using an LLM, ensuring high information coverage. We …

Enhancing Cross-Domain Generalizability In Social Determinants Of Health Extraction With Prompt-Tuning Large Language Models, Cheng Peng, Zehao Yu, Kaleb E Smith, Wei-Hsuan Lo-Ciganic, Jiang Bian, Yonghui Wu

Faculty, Staff and Student Publications

The progress in natural language processing (NLP) using large language models (LLMs) has greatly improved patient information extraction from clinical narratives. However, most methods based on the fine-tuning strategy have limited transfer learning ability for cross-domain applications. This study proposed a novel approach that employs a soft prompt-based learning architecture, which introduces trainable prompts to guide LLMs toward desired outputs. We examined two types of LLM architectures, including encoder-only GatorTron and decoder-only GatorTronGPT, and evaluated their performance for the extraction of social determinants of health (SDoH) using a cross-institution dataset from the 2022 n2c2 challenge and a cross-disease dataset from …