Presentation Type: LIEAF

Toward Trustworthy Medical AI: A Framework to Test the Responsiveness of Clinical AI Models

Click to View Presentation

2025 Annual Symposium On Demand

Presentation Time: 03:30 PM - 03:45 PM

Abstract Keywords: Evaluation, Machine Learning, Clinical Decision Support
Primary Track: Foundations
Programmatic Theme: Academic Informatics / LIEAF

An intelligent system should alert physicians when a patient's condition deteriorates (e.g., low oxygen levels). However, the lack of a comprehensive machine learning model testing framework limits the ability of researchers and clinicians to evaluate them before deployment. In this work, we develop systematic approaches for generating new test cases beyond the original dataset to assess the responsiveness of machine learning models to critical health conditions that may occur in clinical settings.

Speaker:
Humayera Islam, PhD
The University of Chicago

Authors:
Tanmoy Sarkar Pias, PhD Candidate (MSc completed) - Virginia Tech; Sharmin Afrose, PhD - Oak Ridge National Laboratory; Moon Das Tuli, MBBS - Greenlife Medical College & Hospital; Ipsita Hamid Trisha, MPH, MBBS - College of Medicine, University of Arizona; Xinwei Deng, PhD - Department of Statistics, Virginia Tech; Charles B. Nemeroff, PhD - Dell Medical School, University of Texas at Austin; Danfeng (Daphne) Yao, PhD - Virginia Tech;

---

VaxKG: Integrating The Vaccine Ontology And VIOLIN For Advanced Vaccine Queries And LLM-Powered Chat Systems

Click to View Presentation

2025 Annual Symposium On Demand

Presentation Time: 03:45 PM - 04:00 PM

Abstract Keywords: Knowledge Representation and Information Modeling, Controlled Terminologies, Ontologies, and Vocabularies, Large Language Models (LLMs), Data transformation/ETL, Data Sharing, Data Standards, Artificial Intelligence, Bioinformatics
Primary Track: Foundations
Programmatic Theme: Academic Informatics / LIEAF

Vaccine research faces challenges in integrating diverse biomedical datasets. While the Vaccine Investigation and Online Information Network (VIOLIN) provides comprehensive vaccine data, implemented in traditional relational models limit complex analysis. Similarly, the Vaccine Ontology (VO) offers standardized semantic frameworks but lacks comprehensive empirical data. This study addresses these limitations by developing a vaccine knowledge graph (VaxKG) that integrates VIOLIN's dataset with VO's standardized terminology. Using Neo4j, we transformed 12 core VIOLIN tables into a graph structure enriched with VO concepts. The resulting knowledge graph comprises 28,123 VIOLIN data nodes and 101,282 VO resource nodes, connected by 412,865 relationships. Our comparative analysis of Brucella and Influenza vaccines demonstrates VaxKG's ability to enable complex semantic queries and reveal insights unavailable from either resource alone. We further demonstrate VaxKG's utility through VaxChat, a large language model application that leverages the VaxKG as Retrieval-Augmented Generation (RAG) for intuitive vaccine information access.

Speaker:
Feng-Yu Yeh, Masters Degree
University of Michigan Medical School - He Lab

Authors:
Feng-Yu Yeh, Masters Degree - University of Michigan Medical School - He Lab; Matthew Asato, Computer Engineering - He Group; Jie Zheng, PhD - University of Michigan; Yongqun He, PhD - University of Michigan;

---

Enhancing Breast Cancer Recurrence Prediction Across Treatment Scenarios with Weighted Cox Mixtures

Click to View Presentation

2025 Annual Symposium On Demand

Presentation Time: 04:00 PM - 04:15 PM

Abstract Keywords: Cancer Prevention, Machine Learning, Clinical Decision Support
Primary Track: Policy
Programmatic Theme: Academic Informatics / LIEAF

Breast cancer treatment involves surgery, radiation, chemotherapy, and endocrine therapy, with recurrence risk depending on treatment execution. We propose a weighted Cox mixtures model that integrates treatment plans and clinical data to estimate recurrence risk. Data from Mayo Clinic (US) and the National Institute of Oncology (Morocco) inform the model. We enhance expectation maximization within the Cox mixtures model using three weighting strategies: Inverse Probability of Treatment Weighting, Adaptive Weights with focal loss, and Prioritizing Subgroups. In the Mayo Clinic cohort, Adaptive Weights improve predictive accuracy (C-index: 0.67–0.88), outperforming the standard Cox model. In the Moroccan cohort, Adaptive Weights also enhance C-index values (0.60–0.71), though with larger confidence intervals. Our findings demonstrate that weighting strategies refine recurrence risk prediction, particularly in imbalanced cohorts. Expanding datasets, especially in underrepresented populations, is crucial for improving model reliability and clinical applicability.

Speaker:
Hasna EL HAJI, PhD
International University of Rabat

Authors:
Hasna EL HAJI, PhD - International University of Rabat; Amara Tariq, Ph.D. - Mayo Clinic Arizona; Amine Souadka, MD - Surgical Oncology Department, National Institute of Oncology, Rabat, Morocco; Nada Sbihi, PhD - International University of Rabat; Felipe Batalini, MD - Mayo Clinic; Mounir Ghogho, PhD - UM6P; Imon Banerjee, PhD - Arizona State U, Mayo Clinic;

---

Providing Peer Mentoring to Health Informatics Graduate Students for Career Development

Click to View Presentation

2025 Annual Symposium On Demand

Presentation Time: 04:15 PM - 04:30 PM

Abstract Keywords: Education and Training, Surveys and Needs Analysis, Workforce Development
Primary Track: Applications
Programmatic Theme: Academic Informatics / LIEAF

Health Informatics graduate students have highly diverse backgrounds. The consequence of this fact is that they do not have clear guidelines about their future career, which can be dramatically different for students with different backgrounds. Based on opinions collected from our students, we set up a mechanism for peer mentoring. We collected feedback from peer mentoring participants in a pilot study. All participants (both mentors and mentees) were highly satisfied with this peer mentoring program.

Speaker:
Leming Zhou, PhD, FAMIA
University of Pittsburgh

Authors:
Leming Zhou, PhD, FAMIA - University of Pittsburgh; Kimberley Peterson, PhD - University of Pittsburgh; Andi Saptono, PhD - University of Pittsburgh School of Health and Rehabilitation Services; Bambang Parmanto, PhD - University of Pittsburgh;

---

Artificial Intelligence in Health Informatics Education - Perception Among Master’s in Health Informatics Students

Click to View Presentation

2025 Annual Symposium On Demand

Presentation Time: 04:30 PM - 04:45 PM

Abstract Keywords: Education and Training, Large Language Models (LLMs), Teaching Innovation
Primary Track: Applications
Programmatic Theme: Academic Informatics / LIEAF

This study examined perceptions of Artificial Intelligence (AI) among 40 Master’s in Health Informatics students at Grand Valley State University. Findings revealed strong support for AI in improving data accuracy and predictions, with concerns about privacy and reduced human oversight. Students preferred AI training in predictive analytics, cybersecurity, and medical imaging. The results highlight the need for a balanced curriculum that equips students with both technical competencies and ethical understanding to ensure responsible integration of AI in health informatics.

Speaker:
Suhila Sawesi, PhD
GVSU

Authors:
Suhila Sawesi, PhD - GVSU; Mohamed Rashrash, Associate Prof/PhD - University of Charleston; Guenter Tusch, PhD - Grand Valley State University;

---