About Me

Hi, and welcome to my page! My name is Morgan, and I'm a fourth year PhD candidate in the Bioinformatics and Integrative Genomics (BIG) program at Harvard Medical School advised by Dr. Arjun (Raj) Manrai. I received my B.S. in Biomedical Engineering from Johns Hopkins University and my B.M. in Vocal Performance from Peabody Conservatory in 2021.

Generally, I am interested in harnessing health records to improve healthcare. I think a lot about when and how machine-learning based technologies should be used in the clinical domain and how to effectively develop human-AI teams to ease clinician burden and improve patient experiences and care.

In my PhD research, I use statistics and machine learning to study and refine kidney function estimation, deriving insights from both national epidemiological data and longitudinal clinical data. Specifically, I have quantified the national implications of updated race-free eGFR equations on clinical decisions for oncology patients. I also hope to quantify national clinical implications of eGFR inaccuracy in the general population.

In my current primary project, I am building models which utilize multiple timepoints of blood lab biomarkers to predict kidney health trajectories in collaboration with Clalit Health Services. This project involves a combination of data engineering and survival analysis as I hope to derive insights from a comprehensive, diverse, longitudinal EHR dataset with over 5 million individuals.

As I approach the last couple of years of my PhD study, I am actively searching for opportunities to continue to build and study machine learning models for healthcare. I am in active pursuit of an internship and intend to start a full time position in the next 1-2 years and would love to chat about opportunities in industry and academia!

Education

B.S. in Biomedical Engineering
Johns Hopkins University
Baltimore, MD

Education

B.M. in Vocal Performance
Peabody Conservatory
Baltimore, MD

Education

Current PhD Student
Bioinformatics and Integrative Genomics
Harvard Medical School
Boston, MA

Recent Research Projects

AI-clinician collaboration via disagreement prediction: A decision pipeline and retrospective analysis of real-world radiologist-AI interactions

Clinical decision support tools can improve diagnostic performance or reduce variability, but they are also subject to post-deployment underperformance. Although using AI in an assistive setting offsets many concerns with autonomous AI in medicine, systems that present all predictions equivalently fail to protect against key AI safety concerns. We design a decision pipeline that supports the diagnostic model with an ecosystem of models, integrating disagreement prediction, clinical significance categorization, and prediction quality modeling to guide prediction presentation. We characterize disagreement using data from a deployed chest X-ray interpretation aid and compare clinician burden in this proposed pipeline to the diagnostic model in isolation. The average disagreement rate is 6.5%, and the expected burden reduction is 4.8%, even if 5% of disagreements on urgent findings receive a second read. We conclude that, in our production setting, we can adequately balance risk mitigation with clinician burden if disagreement false positives are reduced. Read More.

Cancer Treatment and Trial Eligibility Changes from Estimating Kidney Function without Race

In patients with cancer, kidney function assessment is instrumental for guiding treatment and determining trial eligibility. Many oncologists rely on serum creatinine or creatinine clearance, but nephrology guidelines recommend evaluation of glomerular filtration rate (GFR), with estimated GFR (eGFR) from serum creatinine as the first step. After calls to reconsider race in GFR estimation, the National Kidney Foundation and the American Society of Nephrology (NKF-ASN) recommended moving towards race-free eGFR equations, namely the 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine and creatinine/cystatin C combined eGFR equations. Many medical centers have made the shift to new equations, but the implications for both Black and non-Black cancer patients on treatment and trial eligibility are unclear. In this project, we estimate clinical implications of the shift to race-free equations on U.S. cancer patients.

Programming Languages

  • Python (adv)
  • R (int)
  • Java (int)
  • C/C++ (beg)
  • HTML/CSS (beg)
  • SQL (beg)
  • MATLAB (beg)
  • ROS (beg)

    • Skills

    • Cloud Computing
    • Collaborative Software Development
    • Neurocartography
    • Attention to Detail
    • Mentorship
    • Experiment Design
    • Creativity
    • Organization and Leadership

Select Coursework

Project-Based Courses

MIT.6.871: Machine Learning for Healthcare

Course Description: Introduction to machine learning in healthcare (i.e. the nature of clinical data, use of machine learning for risk stratification, disease progression modeling, precision medicine, diagnosis, subtype discovery, and improving clinical workflows, guest lectures and course projects with real clinical data)
Group Project: Augmenting Deep-Learning-Based Catheter Malpositioning Detection with Noisy Labels

JHU Precision Care Medicine

Group Project: Predicting Physiological Deterioration and Mortality in Mechanically Ventilated ICU Patients (Presented at 2021 Dept. of Medicine/Whiting School of Engineering (WSE) Research Retreat, WSE Research Award Finalist)

MIT.6.8610: Quantitative Natural Language Processing

Course Description: covers both modern and classic approaches to NLP with a particular focus on statistical methods Group Project: Speak Up! A Probabilistic and Practical Approach for Turn-Taking Prediction

JHU Neuro Data Design

Group Project: Implemented new split criteria in forked scikit-learn decision tree (in Cython) code. Shown to outperform all existing split criteria in the package in several nonlinear, multi-output simulations

Assessment-Based Courses

Mathematics
  • Linear Algebra (JHU)
  • Differential Equations (JHU)
  • Multivariable Calculus (JHU)
  • Probability and Statistics for Engineers (JHU)
Computational Biology
  • Biomedical Data Science (JHU)
  • Computational Medicine: Cardiology (JHU)
  • Biological Models and Simulations (JHU)
  • Computational and Functional Genomics (Harvard)
Computer Science
  • Computer Vision (JHU)
  • Machine Learning (JHU)
  • Data Structures (JHU)
  • Introduction to Data Science in Python (Coursera)
Biology & Chemistry
  • Organic Chemistry (JHU)
  • Biochemistry (JHU, Harvard)
  • Genetics (Harvard)
To learn more about my journey and introduction to research, check out my story.