Mankind has known for a long time that the human brain can glean understanding from pictures at a much faster rate than it can from processing a table of rows and columns. Yet up until recently, business intelligence (BI) software did not fully embrace this scientific axiom. While the spreadsheet has been by far the most important and widely used business application in history, its simple rows and columns do not harness the full analytical capabilities of the human brain in the same way as visual patterns and symbols can.
In order to best conceptualize and intuit business trends, leading/lagging indicators, progressional outliers and the like, business users need a pictorial representation and manifestation of rows and columns. While a picture may be worth a thousand words, in modern times it is worth potentially tens of thousands of rows. No matter how well an organization’s data has been integrated; no matter how high the quotient of data integrity may be; no matter how minimal the latency of tactical information, or all-encompassing its breadth of business knowledge, corporate performance management will have deficiencies if visual analysis capabilities are lacking.
Mashups: Providing A Shared Context
There are many industries that are still starved for decent data visualization applications. The glaring need for visually compelling BI analysis is especially prevalent in the healthcare industry, where medical decision making is still often conducted in a piecemeal and overly iterative manner. The Medical Industrial Complex of hospitals, clinics, doctors, research facilities and insurance companies is still, for the most part, clinging to old-school modes of statistical heuristics, probability analysis, and patient prognostication and diagnosis.
Standardized and extensible visual templates that provide a unified frame of reference are often absent. If such root constructs do exist, they will be decades old or circumspect. For example, while a simple X-ray photograph may provide doctors with a reliable and conclusive patient diagnosis, the results portrayed may be superficial and will not be extensible. Furthermore, the results may not be easily distributable to other medical personnel. For more complicated imaging applications such as MRIs, these problems will be compounded. Wouldn’t it be nice if doctors could better “mark up” such images in a collaborative fashion, drilling down and expanding on each other’s comments? How about the ability to overlay multiple image slices over a base exoskeleton picture in order to better define patient ailments and socialize them to peers in other wellness facilities?
All too often, when a patient has an abnormal MRI, a long process of consensual diagnosis and prognostication (among many doctors) will commence, leaving the patient waiting longer than he/she should for treatment. The reality is that it may take weeks for medical records to be disseminated to all important parties. Still more precious time may pass until test results are reviewed by doctors and comments/recommendations are collected and aggregated into a final diagnosis, which is ultimately returned (with more lag time) to the patient’s primary care provider. In cases where there is a greater sense of urgency—such as results from computerized axial tomographies (CAT scans)—quick inter-doctor collaboration becomes much more critical. Aside from the need for common visual contexts, a unified operational locus (a distributed medical dashboard or other collaborative workflow application) is a fundamental yet outstanding need.
All is not gloomy though. Microsoft has positioned SharePoint to be the software of choice for these types of concerns. While MS SharePoint has emerged as the premier collaborative workflow solution for small, medium and large-size firms across the globe, data visualization software vendors have kept pace by including APIs for the most viable SharePoint services in their product offerings, resulting in richly visible web applications that are highly shareable, scalable, portable and extendable. It is now possible to dynamically integrate information from various business silos with external cloud-based data to create composite and multi-layered mashups on-the-fly.
With this kind of medical technology, healthcare professionals would find themselves more productive and their services more value added. In New York City (where I live and work), doctors would be able to better share medical documents, tests and health records with one another electronically, as opposed to what is often the modus operandi: relying on couriers and bike messengers to deliver the bulky physical images (scan results) all over Manhattan. Support personnel could be trained on new technology and spend their time constructing mashed-up test results and posting them on a common portal or remotely enabled dashboard—leveraging mechanisms like SharePoint image libraries—instead of wasting their valuable time tracking and scheduling physical deliveries of all records.
For now, the integration of data from disparate medical devices remains stuck in the dark ages at even the most modern of medical institutions. However, data visualization technologies that help establish a shared context and situational awareness for doctors, especially when these doctors specialize in differing disciplines, can (and hopefully will) be a key driver for much-needed future healthcare automation. Shared contexts will be best realized through a new breed of data visualization applications, specifically those which rely on data mashups as the primary means of data synthesis and exploration of business knowledge.
GIS: The Most Recognizable Flavor Of Mashup And Common Contextual Reference Point
Geographic Information Systems (GIS) provide perhaps the most common shared context in information navigation today. Not only have users of electronic applications become used to seeing data in the context of a location, they are coming to expect this functionality in an increasing number of applications. Such localization intelligence can give an implicit shared context that lends itself very well to data mashups and resonates with both novice and advanced users of web-enabled applications.
The enormous popularity of Yahoo Maps, Google Earth, MapQuest and other GPS applications bears this out, where semiotics provide a hundred times more information than what is available in a static text-oriented view. The ability to overlay maps with geographical landmarks - and the ability to selectively filter those landmarks out of a mashed-up result set - has forever changed the way people and businesses view the world. All leading visual reporting software and visual BI toolkits now provide geocoding APIs that let developers build applications that utilize and interact with mapping services and applets provided by Microsoft, Google, Yahoo and many more.