Data-Mining for the Analysis of Alternative Medicine: The Case of Osteopathy Diagnostic Methodology in Japan

Osteopathy is a relatively common form of alternative medicine modalities used to treat musculoskeletal complaints in Japan. However, the diagnostic and therapeutic manipulations used are very different and sometimes unique, compared to that of western medicine. One problem is the difficulties in verification of alternative medical practices through modern clinical scientific methods. We examined whether and how data-mining methodologies can be applied to overcome some of the problems. We were fortunate to obtain diagnostic records from a very popular osteopathy clinic in Osaka, Japan thatincluded over 30,000 patient visits over 6 years of practice. The data consists of some careful measurements of tissue electro-conductivity differences at 5 anatomical positions which as a whole are looked upon as an indication of the most appropriate approach/ location of the patient´s problem. These include the left and right side of the neck, the armpits, the wrists, the knees, and the ankles. Our research group received these records and applied data mining and knowledge discovery algorithms to look for any scientifically meaningful associations within the patient data elements recorded. This study might assist the osteopath in discovering potentially useful and valuable knowledge in the form of patterns/correlations within the data elements and more importantly to help us scientifically verify the diagnostic methodology adopted by the osteopath.