The importance of visualizing data is universally recognized. But, usually the data is passive input to some visualization tool and the users have to specify the precise graph they want to visualize. BeyondCore simplifies this process by automatically evaluating millions of variable combinations to determine which graphs are the most interesting, and then highlights these to users. In essence, BeyondCore automatically tells us the right questions to ask of our data.
In this video, Arijit Sengupta, CEO of BeyondCore, describes how public health data can be analyzed in real-time to discover anomalies and other intriguing relationships, making them readily accessible even to viewers without a statistical background. Arijit will be speaking at Strata Rx 2013 with Tim Darling of Objective Health, a McKinsey Solution for Healthcare Providers, on the topic of this post.
The whole video can be viewed below.
Key topics include:
- Introduction to data visualization at BeyondCore. [Discussed at the 0:00 mark]
- Briefings and how they can be more intelligent. [Discussed at the 1:06 mark]
- Demo of an animated briefing. Based on the data, BeyondCore can find relationships and interesting anomalies. BeyondCore automatically highlights what is most interesting in each graph, and the relationships between graphs. [Discussed at the 2:44 mark]
- Explanation of how the software can examine millions of combinations to find important patterns. Multiple techniques such as clustering, machine learning, and regression analysis are used to relieve the user of the need to manually do statistical analysis. [Discussed at the 5:18 mark]
- How the software statistically accounts for confounding factors and explains exactly what is driving the observed behavior. [Discussed at the 7:59 mark]
- How the real-time analysis is achieved using Hadoop-like clustering. [Discussed at the 9:47 mark]
- Applications: revealing an unnoticed cluster of hospital infections; turning up hidden variations in length of stay. [Discussed at the 10:46 mark]
- Accurate analysis in the absence of clean data: “There is no such thing a
s clean data.” [Discussed at the 11:51 mark]
- Hazards of using visualization without pairing it with statistical analysis. One can be distracted by visually interesting patterns in the data that actually mean nothing significant. [Discussed at the 12:51 mark]
- Making data simple for clinicians without a background in analysis and computing. [Discussed at the 16:34 mark]
- Instead of getting into “collection mode,” gathering huge amounts of data for no reason, start with a problem and try to collect the relevant data. Use iterative exploration to interpret data and collect more that fills in the unknowns. [Discussed at the 17:37 mark]
- BeyondCore’s vision where data is available not only to clinicians but to patients. Someday soon, any patient will be able to leverage data to determine which hospital to get the best treatment for their specific needs. [Discussed at the 19:07 mark]
See a related posting on this topic, The Next “Top 5%”: Identifying patients for additional care through micro-segmentation.