Cindy Zhang

PhD Student, Biomedical Informatics & Medical Education

University of Washington

Cindy Zhang

About

I am a PhD student with BIME, working in the Foy Lab. I received my Bachelor of Science in Biomedical Engineering and Computer Science from Johns Hopkins University. My research focuses on improving the interpretation of routine blood-based laboratory tests to advance personalized and preventive medicine.

Bearing in mind the dynamic human lifespan, my current work develops personalized, adaptive setpoint models from longitudinal clinical data using Bayesian methods. Through this approach, I aim to disentangle health, physiological normality, and disease emergence by characterizing personalized trajectories over time.

Previously, I was at PMAP as a data engineer and J&J surgical robotics as a systems engineer. Outside of research, I am a registered yoga teacher (RYT-200) certified through The Yoga Shala in Seattle. I also enjoy reading.

Download my CV (PDF)

News

  • Nov 2025 Presented "Modeling Physiologic Setpoints from Blood Tests to Quantify Human Regulation" at UW Scholars' Studio, Seattle.
  • Sep 2025 Started PhD in Biomedical Informatics & Medical Education at the University of Washington.
  • Feb 2025 Presented "Adaptive AI Models for Personalized Laboratory Reference Intervals" at UW IMDS Symposium, Seattle.
  • Feb 2025 OPTIC paper on clinical message triage published on arXiv.
  • Dec 2024 "Haematologic Setpoints" paper published in Nature.
  • 2024 Selected as Leader of Tomorrow at GapSummit Global Biotech Leadership Forum, Cambridge, UK.
  • May 2023 Graduated from Johns Hopkins University with dual BS degrees.

Publications

  • Machine Learning and Artificial Intelligence-Based Clinical Decision Support for Modern Hematology
    Machine Learning and Artificial Intelligence-Based Clinical Decision Support for Modern Hematology Zhang C, Lam BD, Lucas F, Foy BH Clinics in Laboratory Medicine, 2025 (In Press)

    A review of ML and AI approaches for clinical decision support in hematology, covering diagnostic algorithms, risk stratification, and integration into laboratory workflows.

  • OPTIC: Optimizing Patient-Provider Triaging & Improving Communications in Clinical Operations Using GPT-4 Data Labeling and Model Distillation
    OPTIC: Optimizing Patient-Provider Triaging & Improving Communications in Clinical Operations Using GPT-4 Data Labeling and Model Distillation Santamaria-Pang A, Zhang C, et al. arXiv, 2025 [arXiv]

    We use GPT-4 to label clinical messages and distill the knowledge into smaller models for real-time patient message triage, reducing physician cognitive load at scale.

  • Machine Learning Approach to Predict Emergent RV PV Loop Phenotypes in Pulmonary Hypertension
    Machine Learning Approach to Predict Emergent RV PV Loop Phenotypes in Pulmonary Hypertension Sivakumar N, Zhang C, et al. Pulmonary Circulation, 2025 (Accepted)

    We apply machine learning to predict right ventricular pressure-volume loop phenotypes, enabling earlier identification of hemodynamic patterns in pulmonary hypertension.

  • Haematologic Setpoints Are a Stable and Patient-Specific Deep Phenotype
    Haematologic Setpoints Are a Stable and Patient-Specific Deep Phenotype Foy BH, Zhang C, et al. Nature, 2024 [Paper]

    We show that individuals maintain stable, personalized hematologic setpoints over time, establishing a new framework for interpreting routine blood tests as deep phenotypes.

  • Assessing Associations Between COVID-19 Symptomology and Adverse Outcomes After Piloting Crowdsourced Data Collection
    Assessing Associations Between COVID-19 Symptomology and Adverse Outcomes After Piloting Crowdsourced Data Collection Flaks-Manov N, Zhang C, et al. JMIR Formative Research, 2022 [Paper]

    We crowdsourced COVID-19 symptom data via Amazon Mechanical Turk and assessed associations between symptom profiles and adverse clinical outcomes.