The healthcare data firm Jvion, which utilizes artificial intelligence (AI) approaches in predictive analytics and machine learning, has create a vulnerabilty map based on 30 million de-identified patient records. The patient data was then modeled against respiratory viruses similar to COVID.
DETROITography 
Thanks for sharing this! Though it looks like they didn’t filter out some things, which makes certain parts of the map look more vulnerable than they actually are. For example, on Detroit’s lower east side, both Belle Isle and the Conner Creek plant are shaded in dark purple, signifying high vulnerability. However, since there are no residential units on those tracts, there is no vulnerability, so just looking at the colors without allowing for context is misleading.