The benefits of Eastern versus Western medicine have been debated for decades. But the only way to really test the difference is scientifically--it's a difficult task, although there are plenty of ongoing studies. IBM's joining that fray with a tool to analyze medical records in collaboration with the Guang Dong Hospital of Traditional Medicine (South China's largest traditional hospital).
Today at their Smarter Cities event in Shanghai, IBM's announcing that its Healthcare Information Warehouse (HIW) system will be used in a pilot study to work out how effective the Eastern and Western approaches to treating one specific condition--chronic kidney disease--are, in an experiment that, if successful, could be extended to comparing cure rates and techniques of thousands of other conditions and diseases.
The whole idea behind the experiment is that comparative medicinal studies have nearly always involved painstaking manual trawling through thousands of patient's medical notes, with all the associated human errors, time demands, and practical difficulty of keeping track of the data. Electronic medical records are boosting this sort of research, making it far easier to gain access to data and analyze many people's records. But there's still a large degree of human to-ing and fro-ing required since the records are in word-form, and subject to individual doctor's habits. IBM's warehousing system is designed to minimize that effect by storing and synthesizing the anonymized data of thousands of patients all in one place. It's smart enough to do much of the correlation analysis too, deciding if different patients' symptoms changed in the same (or statistically similar) ways in response to particular treatments, and it can automatically take account of limiting factors like age, gender and other medical conditions.
In the end, IBM is creating an almost automatic diagnostic tool, able to sift through a ton of data and help clinicians decide if particular combinations of Eastern or Western treatments for chronic kidney disease are most beneficial for cure or prevention. Decisions about treating a condition can be made based on real scientific analysis much more quickly than before, and aggregate the expertise of many different practicing doctors to determine the most effective solutions. Technically the system represents an extremely clever coding effort, since it needs to be able to sift through highly complex XML documents (the individual medical records) which may be written numerically or in natural language and intelligently gather the relevant information needed for analysis.