AI Infrastructure: Technical Integration Testing
How do we ensure what we built actually works in production?
Recent Posts
AI Infrastructure Example: C. difficile Infection Risk
Discussion of the technical integration for an in-hospital infection risk stratification model.
Healthcare AI Infrastructure - To be deprecated
Infrastructure is king.
M-CURES: AI Infrastructure for Predicting COVID-19 Deterioration in Hospitals
A detailed exploration of the development and implementation of M-CURES, a COVID-19 in-hospital deterioration model, highlighting the integration of AI model...
Healthcare AI Implementation Infrastructure: Technical Tools for AI Model Integration
A guide to the technical infrastructure needed to integrate AI models into clinical workflows, covering both internal and external integration approaches and...
Healthcare AI Development Infrastructure: Tools and Data for Model Creation
An in-depth look at the tools and environments needed to develop AI models for healthcare, emphasizing the importance of integrating implementation considera...
Healthcare AI Infrastructure: Key Systems for Making & Using Clinical AI Models
Infrastructure is king. An intro to the foundational healthcare IT systems that power AI, focusing on how electronic medical records & technical integrat...
Healthcare AI Implementation: Steps for Successful Clinical Integration
A look at the crucial steps in AI implementation, covering technical integration, prospective validation, and model monitoring to ensure AI tools make a mean...
Healthcare AI Development: From Data Access to Model Validation
A deep dive into healthcare AI model development, from selecting tasks to preparing data and validating models, highlighting the challenges and opportunities...
An Introduction to the Healthcare AI Lifecycle: From Development to Implementation
Explore the healthcare AI lifecycle, a journey from development to implementation, unveiling how AI can transform patient care amidst the complexities of mod...
OUWB Med Education Week: Intro to AI for Medicine
Lessons from Across the Healthcare AI Lifecycle.
AI in Clinical Care: Stanford EM Faculty Development Series
AI in Clinical Care
A Primer on Al in Medicine
The three questions you should feel comfortable asking as a physician when confronted with an AI system.
Biomedical Waveforms
Unveiling the alpha version of BiomedWaveforms, an open-source JavaScript framework to generate and plot commonly used biomedical waveforms, like EKGs and ca...
Revamping GitHub Action for Post Concatenation
It’s never too early to fix a mistake you’ve made with continuous integration.
Playing Around with Programmatic LLM Access
Exploring the practicalities and nuances of interacting with Large Language Models (LLMs) programmatically.
QRS*: The Next Frontier in Simulated Cardiac Intelligence
Dive into the world of QRS*, where cardiac rhythms meet cutting-edge simulation. Forget Q*’s quest for AI supremacy – here, we’re revolutionizing how we view...
Understanding How Apple Watch Estimates VO2 Max: Workout Data Extraction
Continuing our journey to understand Apple’s VO2 max estimation algorithm, by getting workout data.
WPI Business Week: ML for Healthcare Talk
I had the distinct pleasure of joining the vibrant community at WPI Business School for a conversation that took us to the crossroads of technology and healt...
Hyponatremia Modeling
Can we build tools to help with the algorithmic way of assessing hyponatremia?
Understanding How Apple Watch Estimates VO2 Max: Introduction and Data Extraction
Leveraging personal HealthKit data to evaluate and understand Apple’s VO2 max estimation algorithm.
Airline Merger Data Visualization
Analyzing how airlines have come and gone over the past century.
Iframe Embedding: Why and How
Exploring the motivation behind using iframes OR how to seamlessly integrate the internet into your blog.
Toki Conference Timer
Perfect Timing for Talks!
IAMSE Artificial Intelligence: Preparing for the Next Paradigm Shift in Medical Education
Joined the International Association of Medical Science Educators (IAMSE) for 2023 webinar series on artificial intelligence in medical education. Dr. Cornel...
It’s Automation All the Way Down! How to Use GitHub Actions for Blogging Automation with LLMs
CI/CD automation isn’t just for large-scale projects; it’s a game-changer for individual programmers. I’ve started using the power of GitHub Actions to impro...
2023 Machine Learning for Healthcare Conference
Presentation at Machine Learning for Healthcare 2023 in New York on our work on rank-based compatibility. During the conference I presented a brief spotlight...
Hangman
A simple hangman game. Made with chatGPT.
INFORMS Healthcare: Rank-based Compatibility
Presentation at INFORMS Healthcare 2023 on our work on rank-based compatibility. You can find a link to the post about the upcoming paper here.
QR Code Generator
A simple QR code generator that you can use to make QR code embeded with the strings of your dreams! I made this for a series of presentations I gave. It ena...
Updating Clinical Risk Stratification Models Using Rank-Based Compatibility
As machine learning models become more integrated into clinical care, how can we update them without violating user expectations? We proposed a new rank-base...
Hello, World! 2.0
New blog, who dis?
Collaborative for Human Factors, Cognitive Load, and Well-being: AI Overview
Collaborative for Human Factors discussion on artificial intelligence, ChatGPT, and applicable research.
National Academy of Medicine: AI in Health Professions Education Workshop
Panel discussion on AI in health professions education.
RISE Virtual Talking Circle: Innovations in Machine Learning and Artificial Intelligence for Application in Education
University of Michigan Medical School RISE (Research. Innovation. Scholarship. Education) virtual talking circle discussion with Dr. Cornelius James.
STAT News: How medical schools are missing the mark on artificial intelligence
Discussed my recent perspective paper on incorporating AI into medical education with Dr. James Woolliscroft and Katie Palmer of STAT News. Check out the ful...
Teaching AI as a Fundamental Toolset of Medicine
New article out in Cell Reports Medicine. It is a perspective paper on incorporating AI into medical education with Drs. Cornelius A. James, Kimberly D. Lom...
Using natural language processing to determine factors associated with high‐quality feedback
Global Surgical Education. Can be found here.
UMich IOE Promo Video
Was featured in the University of Michigan Department of Industrial and Operations Engineering promotional video.
Prospective evaluation of data-driven models to predict daily risk of Clostridioides difficile infection at 2 large academic health centers
Infection Control and Hospital Epidemiology. Can be found here.
Helping people get back to work using deep learning in the occupational health system
Discussed our recent JAMIA paper on predicting return to work with Jessalyn Tamez. Check out the news brief here.
Dynamic prediction of work status for workers with occupational injuries: assessing the value of longitudinal observations
Journal of the American Medical Informatics Association manuscript, can be found here.
Augmented Reality Demo
This is an augmented reality (AR) demo using Apple’s Augmented Reality tools. The 3D asset is a USDZ file created by Apple (they own all rights to it). It is...
Doctor’s Notes Software Prototype
A project that was focused on examining and improving the way doctor’s notes are written.
Intro to Machine Learning Models for Physicians: Part I
This a foundational post that has two aims. The first is to demystify machine learning which I believe is key to enabling physicians and other clinicians to ...
Why Doctors Should Contact Tech Support
If you see something, say something.
Early identification of patients admitted to hospital for covid-19 at risk of clinical deterioration: model development and multisite external validation study
British Medical Journal. Can be found here.
Bigger Needle Smaller Number
This is going to be a short post because the last one about EHR front-ends was about 3 times longer than I had originally planned for it to be.
The Complicated Way You See Patient Data: A Discussion on EHR Front-Ends for Doctors
I have a love-hate relationship with electronic health records (EHRs). This relationship first started in the early 2000s at a high school sports physical an...
Development and Validation of Models to Predict Pathological Outcomes of Radical Prostatectomy in Regional and National Cohorts
The Journal of Urology article, read here.
What Can Healthcare IT Learn from the B-52?
A lot of aviation videos show up on my YouTube feed (thank you, DarkAero team). A video that popped up recently was about the Boeing B-52 Stratofortress (B-5...
Looking for Data
One of the nice things about being an MD-PhD student at a research institution with a large academic medical center is that you tend to have a lot of support...
Plutonium Pacemakers
This is a reformatted version of a twitter thread I had put together nearly a year ago. In a former life I worked on designing the manufacturing system for c...
#@!% Flipping Cursor!
ecently I came across some interesting behavior in Microsoft Word. While scrolling through a document I noticed that my pointer had flipped. Instead of seein...
Wordle Solver
<!DOCTYPE html> Wordle Solver by eotles
Solving Wordle
Let’s talk about Wordle. [1] You, like me, might have been drawn into this game recently, courtesy of those yellow and green squares on twitter. The rules ar...
Hello, World!
Hello, World!
Precision Health Webinar: What Clinicians Need to Know when Using AI
Panel discussion on what is important for clinicians to know and how confident they can be when using these AI tools. Conversation with Drs. Rada Mihalcea, M...
IOE Research Spotlight
Shared an overview of my research during the 2021 University of Michigan Department of Industrial and Operations Engineering recruiting weekend.
Comparative Assessment of a Machine Learning Model and Rectal Swab Surveillance to Predict Hospital Onset Clostridioides difficile
IDWeek Abstract. Can be found here.
Natural Language Processing to Estimate Clinical Competency Committee Ratings
Journal of Surgical Education. Can be found here.
Natural Language Processing and Assessment of Resident Feedback Quality
Journal of Surgical Education. Can be found here.
Quantification of Sepsis Model Alerts in 24 US Hospitals Before and During the COVID-19 Pandemic
JAMA Network Open. Can be found here.
Forbes: Trust The AI? You Decide
Arun Shashtri of Forbes covered our JAMA IM Epic Sepsis Model evaluation paper. Check out the article.
INFORMS: Dynamic Machine Learning for Medical Practice
INFORMS conference talk focused on dynamic machine learning for medicine. Based on Joint work with Jon Seymour, MD (Peers Health) and Brian Denton PhD (Unive...
Using Natural Language Processing to Automatically Assess Feedback Quality: Findings From 3 Surgical Residencies
Academic Medicine. Can be found here.
CHEPS Seminar: Engineering Machine Learning for Medicine
Invited to give a talk for the 2021 University of Michigan Center for Healthcare Engineering and Patient Safety (CHEPS) fall seminar series. Discussed engine...
Machine Learning for Healthcare Conference: Characterizing the Performance Gap
Jeeheh Oh and I presented our work on dataset shift at the 2021 Machine Learning for Healthcare Conference. This talk briefly summarizes our our conference p...
Evaluating a Widely Implemented Proprietary Deterioration Index Model among Hospitalized Patients with COVID-19
Annals of the American Thoracic Society. Can be found here.
Mind the Performance Gap: Dataset Shift During Prospective Validation
Our 2021 Machine Learning for Healthcare Conference paper! It discusses a special kind of dataset shift that is particularly pervasive and pernicious when de...
The Washington Post: Sepsis prediction tool used by hospitals misses many cases, study says. Firm that developed the tool disputes those findings.
Erin Blakemore of The Washington Post covered our JAMA IM Epic Sepsis Model evaluation paper. Check out the article.
The Verge: A hospital algorithm designed to predict a deadly condition misses most cases
Nicole Wetsman of The Verge covered our JAMA IM Epic Sepsis Model evaluation paper. Check out the article.
WIRED: An Algorithm That Predicts Deadly Infections Is Often Flawed
Tom Simonite of WIRED covered our JAMA IM Epic Sepsis Model evaluation paper. Check out the article.
STAT News: A popular algorithm to predict sepsis misses most cases and sends frequent false alarms, study finds
Casey Ross of STAT News covered our JAMA IM Epic Sepsis Model evaluation paper. Check out the article.
External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients
JAMA Internal Medicine. Can be found here.
UMich MSTP Promo Video
Was featured in the University of Michigan Medical Scientist Training Program recruiting video. The MSTP at Michigan prepares physician scientists for car...
UMich Precision Health Onboarding Session: Precision Health De-Identified RDW
Precision Health Data Analytics & IT workgroup held an onboarding session for Engineering students who could use Precision Health tools and resources for...
UMich Precision Health Symposium: Prediction & Prevention - Powering Precision Health
Virtual panel discussion on precison health. A video segment from the 2020 University of Michigan Precision Health Virtual Symposium.
It’s time to bring human factors to primary care policy and practice
Appeared in Applied Ergonomics. Link
Ann Arbor District Library - Friday Night AI: AI and COVID-19
Virtual panel discussion on how artificial intelligence could guide the response to the coronavirus outbreak. Hosted by the Ann Arbor District Library. Panel...
Faster than COVID: a computer model that predicts the disease’s next move
Michigan Engineering News covered our work on the M-CURES COVID deterioration model that I helped to develop and led the implementation of. Read the article ...
COVID-19 Analysis
Quick exploration of case spread and mortality rates of the novel coronavirus.
COVID Staffing Project: Three Medical Students’ Contributions
Kenneth Abbott, Alexandra Highet and I catalogued our contributions to the COVID Staffing project in a Dose of Reality Blog Post.
Return to Work After Injury: A Sequential Prediction & Decision Problem
Machine Learning for Healthcare Conference clinical abstract, can be found here.
AAFP’s Innovation Fellow Studies Tech, Digital Scribes
Discussed my work studying digital scribes with David Mitchell. Read the interview.
Immune Genomic Expression Correlates with Discharge Location and Poor Outcomes in Trauma Patients
Academic Surgical Congress abstract, can be found here.
M is for Medicine
I developed an an iMessage Sticker Pack for all those interested in medicine, health, and the human body. Download it from the Apple AppSore.
WHO Technical Series on Safer Primary Care: Human Factors
Human factors engineering can play an important role in improving the safety of primary care.
Cues for PE Diagnosis in the Emergency Department: A Sociotechnical Systems Approach for Clinical Decision Support
American Medical Informatics Association Annual Symposium abstract focused on pulmonary embolism diagnosis.
The Impact of ED Census on the Decision to Admit
Emergency Department disposition decisions are influenced not only by objective measures of a patient’s disease state, but also by census.
Quoted in INFORMS in the News
Starting a new INFORMS student chapter.
A community pharmacy initiative to decrease hospital readmissions by increasing patient adherence and competency of therapy
Pharmacy based interventions may help to reduce hospital admissions.
I-PrACTISE Colloquium Primary Care & Predictive Analytics
Exploring the potential benefits and challenges of integrating ML into primary care.
Forecasting ED Patient Admissions Utilizing ML
Using machine learning to predict the likelihood of inpatient admission at the time of patient triage in emergency departments.
Cherry Picking Patients: Examining the Interval Between Patient Rooming and Resident Self-assignment
Investigating the impact of chief complaints on resident physician self-assignment times in emergency departments.
Arena Simulation Modeling Course
Discrete event simulation and operations modeling for undergrads.
Predicting ED Patient Throughput Times Utilizing Machine Learning
Machine learning for emergency department throughput prediction.
Wisconsin Engineer: Optimizing the ER
Article discussing Flexsim-SHS competition win.
2015 FlexSim - SHS ED Modeling Competition
Modeling and optimizing the operations of an emergency department using discrete event simulation.
Send Me Specials Website
Developed a custom text message gateway for businesses to reach their customers.
I-PrACTISE White Paper
Findings from the first I-PrACTISE conference bringing together physicians and engineers.
I-PrACTISE
Improving Primary Care Through Industrial and Systems Engineering