Big Data Solutions to Improve Patient Care and Enhance Efficiency

It is estimated that worldwide healthcare data is expected to grow to 50 times the current total. A McKinsey Global Institute Analysis has identified the following four distinct Healthcare big data pools and recommended that the integration of these provides bigger opportunities for healthcare players.

• Clinical, Providers
• Patient behavior & Sentiment
• Claims data from Payers
• Pharma R&D

Characteristics of Healthcare Big Data

• Volume: In healthcare, data growth comes from digitizing existing data (clinical trials, radiology, EHR/EMR, FDA submissions etc.) and from generating new forms of data. Newer forms of big byte data, such as 3D imaging, genomics and biometric sensor readings are fuelling this exponential growth.
• Variety: The potential of Big Data lies in combining traditional data with new forms of data. The different forms and varieties in healthcare data (Structured, Unstructured and Semi-structured) coming from wearable devices, social media, genetics/genomics, research and other sources.
• Velocity: Traditionally, most healthcare data (X-Ray, Scripts, and Paper files) has been static. However, in some scenarios patients’ real time data viz., bedside heart monitors, ICUs data, and operating room monitors need to be monitored and analyzed with utmost caution.

Healthcare Big Data Use Case − Using Big Data for Prognosis

Big data can be used to build a smart bio-medical image analysis and classification platform to diagnose and carry out prognosis study for diseases like breast cancer. A solution like this provides distributed storage, analysis and computing capabilities to process huge mammogram images and generate results of, analysis, segmentation and prognosis in parallel.

The benefits for Hospitals and Radiologists are:

• Storing massive mammogram/FNA images analysis and classification
• Classifying the images into Benign/Malignant through Naive Bayes classification algorithm
• Predicting the behavior of the tumor nuclei through prognosis and historical image analysis
• Detection of tumor characteristics/micro-calcifications
• Providing “Analytics as a Service” for Radiologists and Physicians