Giridharan Gurumoorthy is a Business Analytics Senior Consultant at Kavi Global LLC. He has more than a decade of experience in the healthcare industry and advanced analytics technology. He also has published several papers in data analytics and its applications.

Up to 10% of the pharmacy claims submitted to health plans and insurance companies are estimated to be f raudulent. A two-step process comprising of a combination of unsupervised and supervised learning techniques can be used to effectively identify the Fraud, Waste and Abuse in the Pharmacy industry. Claims submitted by the pharmacies to insurance companies/health plans provide data rich in valuable insights that enable prediction of fraud, waste and abuse. Suspicious activity can be identified by detecting anomalies in the data using unsupervised techniques like clustering, univariate and multi-variate outlier analysis, link analysis, simulated fraud signatures, etc.

With known fraud, waste and abuse data, supervised learning techniques like Random Forest, Neural Networks, etc., can be used to identify fraudulent transactions similar to historical fraud signatures. Combining outputs of these techniques Fraud scores are generated for each player [member, pharmacy and prescriber]. In this talk, we’re going to illustrate how machine learning [Spark MLLib and GraphX] was used to identify suspicious activity like co-conspiracies to commit fraud by pharmacies and prescribers[doctors] and others. We’re also going to demonstrate how fraud score was determined in this pharmacy claims fraud detection application.

Session hashtag:#DS9SAIS

Advanced machine vision is increasingly being used to investigate, diagnose, and identify potential remedies and their progressions for complex health issues. In this study, a behavioral neuroscientist at the University of Chicago and his colleagues have collaborated with Kavi Global to characterize 3D feeding behavior and its potential changes caused by neurological conditions such as ALS, Parkinson’s disease, and stroke, or oral environmental changes such as tooth extraction and dental implants.

Videos of rodents feeding on kibble are recorded by a high-speed biplanar videofluoroscopy technique (XROMM). Their feeding behavior is then analyzed by tracking radio-opaque fiducial markers implanted in their head region. The marker tracking process, until now, was manual and tedious, and was not designed to process massive amounts of longitudinal data. This session will highlight a near-automated, deep learning-based solution for detecting and tracking fiducial markers in the videos, resulting in a more efficient and robust process, with a 300+ times reduction in data processing time compared to a manual use of the existing software.

Our approach involved the following steps:(i) Marker Detection-Deep Learning algorithms were used to identify the pixels corresponding to markers within each frame; (ii) Marker Tracking-Kalman filtering along with Hungarian algorithm were used for tracking markers across frames; (iii) 2D to 3D Conversion- sequence matching of videos recorded by both cameras, and triangulating marker locations in 2D track coordinates to generate 3D marker locations. The features extracted from videos would be used to characterize behaviorally relevant kinematic features such as rhythmic chewing or swallowing. The solution involved the use of TensorFlow-Python APIs and Spark.

Session hashtag:#AISAIS14