Page 20 - Moving Forward 2020: Harold and Inge Marcus Department of Industrial and Manufacturing Engineering magazine

Department pivots to address covid-19
Modeling COVID-19 vulnerability
In an an effort to lessen COVID-19 spread Soundar Kumara Allen Allen E Pearce Pearce and Allen Allen M Pearce Pearce Professor of of Industrial Engineering brought together a a a a a a a team to to design an an an advanced artificial intelligence and machine learning model to assess vulnerability
counties in in in in the United States The team includes Kumara Paul Griffin professor of of of industrial engineering and interim director
of of of the Consortium to to to Combat Substance Abuse and and doctoral students Mihir Mehta and and Juxihong Julaiti The team gathered data from various public sources such as as the the U S Census Centers for Disease Control and and Prevention Global Heath Data Exchange and and The New York Times to obtain county level statistics for confirmed COVID-19 cases including age gender population density diabetes cancer hypertension and overall county disease mortality The model was designed to potentially assist intervention and and preparation for resource allocations health health policy and and health health care systems The team also designed the the the model to to be used at the the the recovery stage in in order to to assess the the the viability of re-opening counties and re-distributing resources Using artificial intelligence (AI) to allocate testing
Utilizing the the zip-code level of COVID-19 data from the the the Pennsylvania Department of Health researchers in the the Penn Penn State National Science Foundation Center for for Health Organization Transformation (Penn State CHOT) have developed an an an AI tool to to make the optimal allocation of diagnostic testing
possible According to to Runsang Liu a a a a a master’s student in in IME and and Hui Yang director
of of of Penn State CHOT and and associate professor of of of of industrial engineering the density of of of of of a a a a a population number of of of confirmed cases and size of of of vulnerable populations can provide critical information to help optimize the allocation of testing
testing
centers such as drive-thru testing
testing
sites however optimality depends greatly on on on a a a a a a a site’s accessibility location and region coverage Their AI tool aims to to help balance accessibility and testing
equity for each region in in in Pennsylvania allowing the flexibility to make site adjustments based on data “Artificial intelligence is is an indispensable tool to to help realize the full potential of data for decision support as as well as as to optimize accessibility and equity in health systems ” Yang said Modeling COVID-19 spread According to Yang simulation modeling of virus spread is essential for making health systems respond in in a a a a a a a a a fast and proactive manner to protect people from COVID-19 exposure and disease propagation risk Via a a a a a $200 000 grant for Rapid Response Research from the National Science Foundation (NSF) Yang gathered a a a a a a a a a Penn State CHOT team to develop a a a a a a a a a simulation model of human movement patterns and COVID-19 spread dynamics to predict real-time positions of infected populations in in a a a a a a a spatial network The provision of of a a a a a graphical user interface (GUI) software for the the simulation allows the the user to select different types of spatial networks provides flexibility to define the population size and and heterogeneity and and enables the prediction of real- time positions for unaffected infected recovered and deceased people in in a a a a a a spatial region Additionally the GUI tool enables “what-if” analysis that can possibly allow population centers at at at at any scale to dynamically adjust health health policies plan near-term health health care capacity and and control virus spread with rapid and and timely measures 20 IME NEWSLETTER • VOLUME 5 2020 Faculty News