QuantMD

The cover art of the April issue of Annals of Biomedical Engineering (https://link.springer.com/journal/10439/48/4/page/2) is from my latest article on classifi...cation of abdominal aortic aneurysms using machine learning algorithms :) The full paper is here: https://link.springer.com/article/10.1007/s10439-020-02461-9

Meet new collaborators and learn Complexity Science by doing @ Complexity Weekend! Complexity Weekend is a virtual conference focused on developing innovative solutions that correlate with the COVID19 pandemic. This is a virtual call to action for our community of computational and data scientists. Registration is free! This event will feature Complexity Science-inspired lectures, discussions, and workshops on Friday night and Saturday. All attendees will then engage in a ...collective brainstorming and team formation process Saturday afternoon, followed by a facilitated hackathon experience with these teams on Sunday. Check it out here: https://www.complexityweekend.com #datascience #covid19 #hackathon # #workfromhome #healthcare #datascientist #machinelearning #googlecolab #computervision #ai #complexity

Always touching to receive appreciation from clients (especially when cupcakes are involved) #datascienceforgood #quantmd #covid19 #cares #ppp #customerappreciation

Loving the Copper3D open-source 3D printable ventilator mask. Here's one I printed at home on my MakerBot Replicator 2 Learn more at: https://copper3d.com/hackthepandemic/

Now curating a modeling-ready time-series for the city of Chicago, in addition to PA counties, by request from colleagues. Links: - PA county-wise modeling-re...ady time series data from PA dept of health (updated daily): https://github.com/menonpg/COVID19_PA) - Chicago, IL model-ready time-series data scraped from https://www.chicago.gov//si/covid-19/home/latest-data.html : https://github.com/menonpg/COVID19_Chicago Snippets of code to directly import these data into your Python scripts are in the repository's Readme . On other fronts, forecasts for Allegheny County seem to be going on track with my last output! Image attached. My Bayesian model for true cases in Allegheny County has us right around 497.5 cases per Million. Philadelphia is estimated at 1900+ per million presently.

Finally, a model to estimate the "True" number of infections in a given PA county, given the reported cases and deaths owing to COVID19. I employ a Beta-Binomi...al Hierarchical model and a Markov-Chain-Monte-Carlo simulation approach, assuming a current average availability of tests of 1000 People per Test, in order to obtain this result. The numbers depart relatively less further away than reported numbers as number of tests increase (i.e. people per test goes down from 1000 to something smaller). S. Korea is at ~5 people per test, for reference. So, in Allegheny County, as of 4/4/2020, we should have had 956 cases of confirmed COVID19, whereas 552 were reported, and so on. My data for this model is 2 days delayed but hopefully this provides a sense for the need for more testing and accessibility to home based testing in the US. For comparison, if we assume test accessibility of 100 People per Test, then Allegheny County's estimated true case-load would be 833, with 95% confidence ranging at (675, 1343) ...