AI in Healthcare · Luis Diaz Busquets
Medicine,MeasuredBy Machines.
Interactive tools on how AI is used, tested and trusted in healthcare, all running live in your browser.
See how medical AI gets things right, and wrong.
Each one is a working model you can push around. Change a number and watch the consequences, the way a clinician, a regulator or a data scientist would.
Test Accuracy
Sensitivity, specificity and how common a disease is decide what a positive result really means. A 1,000-person grid shows every true and false result.
Outbreak Simulator
The SEIR model epidemiologists use. Set how contagious a disease is, then try vaccination and distancing and watch the curve and the hospital load change.
Algorithm Bias
A risk model trained mostly on one group quietly fails another. Change the training data and see who the model misses.
Accuracy is not one number. It depends on who you test, how common the disease is, and who the model learned from.
AI in Healthcare, one topic at a time.
Built alongside the AI in Healthcare course at Bryant University. Topics with a working tool are live; the rest arrive as the semester covers them.
- IMG-01How AI Reads a ScanA small image model on a public research dataset, with a map of the pixels it relied on.Coming
- NLP-01De-identificationFind and mask names, dates and numbers in a made-up clinical note.Coming
- SCI 359ProjectsAdded as the course goes.Spring 2027
Everything, in one list.
| ID | Name | What it does | Topic | |
|---|---|---|---|---|
| DX-01 | Test Accuracy | Sensitivity, specificity and prevalence on a grid of 1,000 people; positive and negative predictive value; a second test. | Diagnosis | Open → |
| POP-01 | Outbreak Simulator | An SEIR model with R0, incubation, vaccination, distancing and a hospital capacity line. | Epidemiology | Open → |
| FAIR-01 | Algorithm Bias | A logistic risk model trained on unbalanced synthetic data, with error rates for each group. | Fairness | Open → |
Why a finance student builds health tools.
The same questions run through finance and medicine: how much to trust a model, what an error costs, and who carries that cost. This site works through them for healthcare as part of the Applied AI minor at Bryant University.
It is a sister site to the LDB Lab, which covers how machine learning works and how it is used in finance. Everything here runs in your browser; nothing you enter is sent anywhere.
The heart at the top of the page is real anatomy, drawn as dots: the heart walls, coronary vessels, aorta and pulmonary vessels from BodyParts3D, © The Database Center for Life Science, used under CC BY 4.0 and adapted (parts extracted, vessels trimmed, geometry merged).