Welcome to the MD+ resources page! The MD+ community has curated a set of online resources that members have found helpful for learning data science and AI. Whether you're a physician interested in learning how AI-assisted tools might impact your clinical practice or a student interested in using and building applications yourself, our curated resource list has something for everyone and is updated weekly.
Have feedback on linked resources or want to suggest others? We'd love to hear your thoughts via email or using this anonymous Google form.
MD+ is partnering with ML for MDs to present a webinar on evaluating healthcare AI companies and technologies. During the online event, we will introduce a worksheet on company evaluation and use it to walk through an actual healthcare AI pitch deck in real time. Check the Slack for the public Zoom link.
MD+ is partnering with Ambience Healthcare to host Meet AutoScribe, an exclusive virtual experience where attendees can explore the latest cutting-edge advances in fully automated AI medical scribing. Volunteers will roleplay as patient and provider in unscripted mock visits as AutoScribe listens ambiently and instantly generates all of the documentation. We'll review the generated notes together in real time and discuss the implications of AI technology on provider burnout, patient care, and health system efficiency. Register here.
Coming soon! Keep an eye on this page and our Slack channels #ai-med and #datascience-med.
Python is a beginner-friendly programming language used for virtually all data science and AI. If you're interested in learning how to build data science and AI applications and develop fundamental programming skills for research and data parsing, check out some of the resources below!
BE/Bi 103a is a fast-paced college-level programming course taught by Justin Bois at Caltech that assumes no background knowledge on how Python work, and teaches students how to work with biological data using Python.
This notebook is a quick-and-dirty tutoral to Python and machine learning from Srinidhi Bharadwaj from the MedML club at Emory. By the end of the tutorial, you'll be able build a program that analyzes ECGs!
Put together by Justin Johnson while at Stanford, this course is particularly notable for introducing students to essential Python packages and toolkits like NumPy, SciPy, and matplotlib that anyone working in ML will undoubtedly come across.
R is a programming language commonly used by the biostatistics, data science, and clinical research communities.
This blog post written by the developers of R features links to a bunch of helpful resources and guides for getting started with R.
This textbook will teach readers how to do datascience with R from the ground up assuming no prior programminge experience. You'll be able to learn how to explore, visualize, and extract insights from complex dataset.
For MD+ members interested in exploring how AI will impact clinical medicine and other sectors from a macro lens.
This a great lecture series by professors at MIT giving a broad overview of the field of ML for healthcare, discussing everything from clinical datasets to NLP, disease prediction, and model robustness.
This Coursera course is taught by Andrew Ng, one of the leading experts on AI research. This short 10-hour course introduces AI and its potential impacts on society and business intended for a non-technical audience.
For those with some Python experience, this notebook provides a comprehensive walkthrough on how to train baseline models for disease diagnosis from scratch. All the code and data is provided alongisde detailed explainations and expected results.
Meskó and Görög provide an excellent overview article on the state of AI and medicine from 2020 that can be easily read without the need for a technical background.
This is a great introductory resource for learners interested in getting started with statistical learning using R with a broad and less technical approach.
For MD+ members with a working Python foundation and want to dive into the technical aspects of AI to build and deploy their own models down the road.
This AI introductory course from Harvard is an in-depth, challenging course that teaches students how to build AI tools from the ground up using Python.
For those interested in building their own LLMs and learning NLP from 0-to-100, Hugging Face's NLP intro course is an excellent starting point.
For a more mathematical overview on how to evaluate and train different ML models for a variety of different applications, check out this guide. Distilled AI also has a number of other useful guides on model design and engineering here too.
Hugging Face also provides great documentation for building generative diffusion models. Lilian Weng also provides an excellent overview on diffusion models from a technical standpoint in her blog post here.
MONAI is an open-source framework archive for ML development in medical imaging. Their Github repository features great tutorials on training models to classify and segment real medical images.
A reference textbook for those interested in a deep technical dive into statistical methods and areas such as data mining, machine learning, and bioinformatics.
For MD+ members interested in learning about basic SQL skills and the opportunities and challenges with data science applications in clinical medicine.
Built by MDplus community members, Wrangle is an online practice tool with 100+ free practice problems for people interested in learning SQL, R, and other skillsets for data science. Problems range in difficulty from basic filters and group by's to multi-step transformations.
W3 Schools offers an excellent beginner-friendly tutorial on how to use SQL, the standard computer language that all data scientists use for storing, manipulating, and retrieving data in databases.
For those that learn by doing, this is an excellent resource that teaches you how to use SQL using a series of progressive examples that demonstrate key SQL concepts while allowing you hands-on practice.
LeetCode is a problem bank that allows you to gain some additional practice with SQL for problem solving. While it doesn't necessarily have instructional material, it is an excellent playground to build your SQL skills at any stage of training.
These resources are intended for MD+ members interested in diving into their own healthcare datasets headfirst while developing the technical skillset to extract key insights in the process.
Built by MDplus community members, Wrangle is an online practice tool with 100+ free practice problems for people interested in learning SQL, R, and other skillsets for data science. Problems range in difficulty from basic filters and group by's to multi-step transformations.
For those with an interest in learning to use Python for data science, check out this comprehensive tutorial from Codecademy! It even features a portfolio project that teaches you how to parse through US medical insurance data from scratch.
This textbook will teach readers how to do datascience with R from the ground up assuming no prior programminge experience. You'll be able to learn how to explore, visualize, and extract insights from complex dataset.
NEJM AI Ground Rounds is a monthly podcast that features insightful conversations with experts at the intersection of AI and medicine. A great listen for both students and clinicians.
FYI is a podcast run by the American Medical Informatics Association (AMIA) and discusses the AI and medicine from clinical, research, and societal perspectives.
Glass Box is a ML+medicine blog run by Dr. Rachel Draelos, CEO of Cydoc and Duke MD-PhD graduate with a doctorate in computer science. Glass Box covers everything from the fundamentals of medicine to how AI tools can be used to augment patient care.
Decoding Bio is a Substack by Dr. Patrick Malone, a physician-scientist investor at KdT Ventures (and MD+ community member!)
Doctor Penguin is a must-read weekly newletter sharing up-and-coming research showcasing applications of AI in different clinical specialties.
Import AI is a weekly newsletter on AI primarily intended for AI researchers. While you won't necessarily find a focus on ML and medicine, Jack Clark does a great job distilling new research advancements in AI.
Dr. Roxana Daneshjou is a Stanford dermatologist working at the intersection of AI and precision health.
For members interested in learning about AI for biotech and healthcare from a research perspective, check out Professor James Zou's work!
Dr. Dereck Paul is a physician turned CEO of Glass Health building LLM-empowered tools for clinician workflows and documentation.
Dr. Dan Hashimoto is a general surgeon at the University of Pennsylvania applying computer vision to surgical videos and sharing helpful resources for both trainees and physicians.
Professor Suchi Saria is a professor of AI at Johns Hopkins and experienced builder and founder in ML+medicine.