LabQAR: A Manually Curated Dataset for Question Answering on Laboratory Test Reference Ranges and Interpretation

Presentation Time: 01:30 PM - 01:45 PM

Abstract Keywords: Clinical and Research Data Collection, Curation, Preservation, or Sharing, Informatics Research/Biomedical Informatics Research Methods, Knowledge Representation, Management, or Engineering, Machine Learning, Generative AI, and Predictive Modeling
Primary Track: Clinical Research Informatics
Programmatic Theme: Harnessing the Power of Large Language Models in Health Data Science

LabQAR is a manually curated dataset of 550 laboratory tests, including reference ranges, specimen types, and factors like age and gender. This dataset enables accurate predictions and classifications of lab results, supporting clinical applications such as diagnosis and personalized treatment, ultimately improving decision-making and patient outcomes.
Speaker(s):
Balu Bhasuran, Ph.D
Florida State University

Author(s):
Qiao Jin, M.D. - National Institutes of Health; Karim Hanna, MD - University of South Florida Health; Zhiyong Lu, PhD - National Library of Medicine, NIH; Zhe He, PhD, FAMIA - Florida State University; Xiaoyu Wang, MS - Florida State University;

LabGenie – A Patient-Facing AI-Powered Application for Enhancing Older Adults’ Comprehension of Laboratory Test Results

Presentation Time: 01:45 PM - 02:00 PM

Abstract Keywords: Advanced Data Visualization Tools and Techniques, Mobile Health, Wearable Devices and Patient-Generated Health Data, Learning Healthcare System
Primary Track: Data Science/Artificial Intelligence
Programmatic Theme: Digital Health Technologies for Patient Research

LabGenie is a web-based tool designed to help older adults interpret lab test results and engage in informed healthcare discussions. Built on accessibility, security, and personalized engagement, LabGenie uses de-identified EHR data and AI-generated questions to provide personalized insights. It features a user-friendly interface with secure data handling, visualizes lab trends, and offers tailored questions for consultations. Future studies will evaluate its usability, impact on comprehension, and effectiveness in enhancing patient-clinician communication.
Speaker(s):
Author(s):
Zhe He, PhD, FAMIA - Florida State University; Dhruv Kale, MS - Florida State University; Balu Bhasuran, Ph.D - Florida State University; Zhan Zhang, PhD; Mia Liza A. Lustria, PhD - Florida State University; Karim Hanna, MD - University of South Florida Health; Lisa Granville, MD - Florida State University; Xiao Luo, PhD - Oklahoma State University;

Enhancing Healthcare Data Integration: A Machine Learning Approach to Harmonizing Laboratory Labels

Presentation Time: 02:00 PM - 02:15 PM

Abstract Keywords: Data/System Integration, Standardization and Interoperability, Data Integration, Data Quality, Data Standards, Health Information and Biomedical Data Dissemination Strategies, Secondary Use of EHR Data, Natural Language Processing, Machine Learning, Generative AI, and Predictive Modeling
Primary Track: Data Science/Artificial Intelligence
Programmatic Theme: Harnessing the Power of Large Language Models in Health Data Science

Variations in laboratory test names across healthcare systems—caused by differing terminologies, abbreviations,
misspellings, and assay vendors—introduce major challenges to the integration and analysis of clinical data. These
discrepancies hinder interoperability and complicate efforts to extract meaningful insights for both clinical research
and patient care. In this study, we propose a machine learning-driven solution, enhanced by natural language
processing techniques, to standardize lab test names. By employing feature extraction methods that analyze both
string similarity and the distributional properties of test results, we improve the harmonization of test names,
resulting in a more robust dataset. Our model achieves a 99% accuracy rate in matching lab names, showcasing the
potential of AI-driven approaches in resolving long-standing standardization challenges. Importantly, this method
enhances the reliability and consistency of clinical data, which is crucial for ensuring accurate results in large-scale
clinical studies and improving the overall efficiency of informatics-based research and diagnostics.
Speaker(s):
Mehmet BAGCİ, Graduate Student
San Diego State University & University of California San Diego

Author(s):
Samantha Bagsic, Ph.D - Scripps Health; Anna Ritko, Scientiest - Dept. of Research Development, Scripps Health,; Brian Modena, MD - Modena Allergy + Asthma; Yusuf Ozturk, Professor - San Diego State University, ECE Dept.; Truong Nguyen, Professor - University of California San Diego, ECE Dept.; mehmet bagci, Ph.D - University of California San Diego, ECE Dept.;

AI-Empowered Autonomous Microscope Facilitates Real-Time Brain Cancer Diagnosis

Presentation Time: 02:15 PM - 02:30 PM

Abstract Keywords: Medical Imaging, Bioimaging Techniques and Applications, Clinical Decision Support for Translational/Data Science Interventions, Machine Learning, Generative AI, and Predictive Modeling
Primary Track: Data Science/Artificial Intelligence
Programmatic Theme: Health Data Science and Artificial Intelligence Innovation: From Single-Center to Multi-Site

We developed an AI-powered autonomous microscope that enables real-time brain cancer evaluation during surgery. The system digitizes pathology slides and analyzes tissue samples using deep-learning models, achieving high accuracy in differentiating glioma subtypes. This low-cost solution provides crucial diagnostic support to neurosurgeons and pathologists, democratizing access to expert-level pathology evaluations, especially in resource-limited settings.
Speaker(s):
Kian Weihrauch, Undergraduate Engineering Student
Harvard Medical School

Author(s):

AI Mapping of In-House Codes to LOINC Codes Using Laboratory Test Results Excluding Test Names: Toward International Sharing of Medical Data

Presentation Time: 02:30 PM - 02:45 PM

Abstract Keywords: Data/System Integration, Standardization and Interoperability, Machine Learning, Generative AI, and Predictive Modeling, Secondary Use of EHR Data
Primary Track: Data Science/Artificial Intelligence
Programmatic Theme: Health Data Science and Artificial Intelligence Innovation: From Single-Center to Multi-Site

There is an increasing demand for automatic mapping to standardized codes such as LOINC codes to create integrated medical databases across multiple facilities. However, natural language processing (NLP) in Japanese presents greater challenges compared with English with a limited Japanese corpus for medical terms, such as test names. To address this limitation, we developed a machine learning-based method that maps in-house codes to LOINC codes by leveraging test result values, without relying on test names that would require NLP. Using this approach, we achieved high mapping accuracy (70% or higher) for 80.4% of the analytes targeted in this study. The proposed method facilitates easier mapping to standardized codes in languages where NLP is challenging, ensuring accurate mapping to LOINC codes regardless of the source data language.
Speaker(s):
Noriyuki SHIDO, Master of Engineering
Mitsubishi Electric Software Corporation

Author(s):
Noriyuki Shido, M.Eng. - Mitsubishi Electric Software Corporation; Yuma Iwahashi, M.Eng. - Mitsubishi Electric Software Corporation; Hidenari Ohsawa, B.Eng. - Mitsubishi Electric Software Corporation; Katsushige Furuya, M.A. - Mitsubishi Electric Software Corporation; Yasumichi Sakai, B.Eng. - Mitsubishi Electric Software Corporation; Masamichi Ishii, Ph.D. in Global Health - National Center for Global Health and Medicine; Hiroyuki Hoshimoto, M.Sc. - National Center for Global Health and Medicine, Japan; Nobukazu Namiki, B.Eng - Mitsubishi Electric Software Corporation; Kengo Miyo, Ph.D. - National Center for Global Health and Medicine;

A Pathology Foundation Model for Cancer Diagnosis and Prognosis Prediction

Presentation Time: 02:45 PM - 03:00 PM

Abstract Keywords: Machine Learning, Generative AI, and Predictive Modeling, Medical Imaging, Clinical Decision Support for Translational/Data Science Interventions
Primary Track: Data Science/Artificial Intelligence
Programmatic Theme: Health Data Science and Artificial Intelligence Innovation: From Single-Center to Multi-Site

Pathology image evaluation is indispensable for cancer diagnoses, but standard artificial intelligence (AI) methods for analyzing these images have limited generalizability to samples from different populations. To address this challenge, we established the Clinical Histopathology Imaging Evaluation Foundation (CHIEF) model, a versatile weakly supervised pathology AI framework. CHIEF outperformed state-of-the-art methods by up to 36.1% in diagnostic and prognostic prediction tasks across 24 patient cohorts, providing a generalizable foundation for cancer pathology evaluation.
Speaker(s):
Kun-Hsing Yu, MD, PhD
Harvard Medical School

Author(s):
Xiyue Wang, PhD - Harvard Medical School; Junhan Zhao, PhD - Harvard Medical School; Eliana Marostica, MD - Harvard Medical School; Wei Yuan, BS - Sichuan University; Jietian Jin, MD - Sun Yat-sen University Cancer Center; Jiayu Zhang, PhD - Sichuan University; Ruijiang Li, PhD - Stanford University; Hongping Tang, MD - Shenzhen Maternity & Child Healthcare Hospital; Kanran Wang, PhD - Chongqing University Cancer Hospital; Yu Li, PhD - Chongqing University Cancer Hospital; Fang Wang, MD - Yuhuangding Hospital; Yulong Peng, MD - The First Affiliated Hospital of Jinan University; Junyou Zhu, MD - The First Affiliated Hospital, Sun Yat-sen University; Jing Zhang, PhD - Sichuan University; Christopher Jackson, MD, MS - Harvard Medical School; Jun Zhang, PhD - Tencent AI Lab; Deborah Dillon, MD - Brigham and Women’s Hospital;