Apache Spark, Machine Learning and Healthcare

Loading Events

« All Events

  • This event has passed.

Apache Spark, Machine Learning and Healthcare

February 1, 2017 @ 7:00 pm EST

For this meetup, we’ll have an invited speaker from IBM, Asad Mahmood, M.D., who will talk about two health-related applications he developed using Spark, Machine Learning and other technologies from the IBM Open Source Analytics portfolio.

The first application, Virtual Doc, is a user-interactive, Open Source web application for clinical medicine built in IBM Bluemix. Using patient records and patient satisfaction datasets that have been loaded into dashDB, IBM’s analytic cloud-based data warehouse, Virtual Doc prompts the user to enter their symptoms, suggests additional symptoms that may be related to the user’s present condition and then invokes Spark Machine Learning to calculate and predict the most likely diagnoses in real-time. Finally, Virtual Doc recommends the highest rated physicians by specialty that are best suited for the treatment/management of each predicted diagnosis. 

The second application, Suicide Prevent, is an open source mobile application for the VA built as well in IBM Bluemix. Using patient medical and social media (Twitter, Facebook, LinkedIn) data that has been loaded into IBM Cloudant, Suicide Prevent invokes a standardized psychiatric scale in order to calculate a suicide risk score for each veteran loaded into the database. Based on the stratification of each score, Suicide Prevent then issues appropriate recommendations designed to continue preventative measures for low risk patients, schedule psychiatric consults for medium risk patients and contact emergency health services for hospitalization and rehabilitation of high risk patients.  

Asad Mahmood, M.D. is an Open Source Analytics Technical Evangelist with the IBM Federal Healthcare Team. He is a Biomedical Engineer, a Computer Scientist and a Medical Doctor. As a medical student researcher at Howard University Hospital, Asad analyzed data for over 15,000 patients admitted for diabetic complications and utilized Hadoop and NoSQL in order to store and process data while determining the validity of diagnosis for hospitalized diabetic patients. In addition, he has experience developing software applications in Java and Python, web-based applications in JavaScript and mobile applications in Objective C and Swift. Since joining IBM, Asad has developed an exploratory analysis demo in Bluemix of the Zika virus epidemic in the U.S. states and territories. He is currently working on a number of projects leveraging the Spark iPython Notebook as well as building web and mobile applications in Bluemix to perform exploratory and predictive analyses of data with dashDB, Open Source databases such as PostgreSQL in Compose, Spark SQL, Spark MLib, Spark GraphX and Python for Defense Health and Federal Healthcare agencies.

Food and drinks will be provided by IBM.

Details

Date:
February 1, 2017
Time:
7:00 pm
Website:
https://www.meetup.com/Miami-Hadoop-User-Group/events/237061386/

Organizer

Miami Hadoop User Group
Website:
https://www.meetup.com/Miami-Hadoop-User-Group/

Venue

FIU Modesto a Maidique Campus
11200 Sw 8th St
Miami, FL 33174 United States
+ Google Map