Software To Classify People For Ayurveda Treatment

It will help identify a person's constitution type, or "Prakriti". Ayurveda may have got its most futuristic upgrade yet. A multidisciplinary team of scientists, including statisticians, genomics experts, Ayurveda researchers and computer scientists, have developed a software — using Big Data and machine learning approaches — that can efficiently classify people into the three dominant types of constitutions that underlie the Ayurvedic treatment philosophy.

Unlike Western medicine that grades patients on the severity of their disease, Ayurveda classifies even the healthy. Every person, in this view, falls into seven broad constitution types, or "Prakriti," that is determined at the time of birth and remains invariant throughout one's life.

Vata (V), Pitta (P) and Kapha (K) are the most important divisions, decided on the basis of wide variety of features such as anatomy, appetite, skin-type, allergies, susceptibility of disease. The other four are permutations of these.

One result of this is that medicines that work for V, may not cut it for, say K, and therefore researchers — such as those at the CSIR's Institute of Genomics and Biology (CSIR-IGIB) — have previously proffered evidence that there are differences at the molecular level between these categories that may explain, for instance, why a drug like warfarin — used as a blood thinner — may require different dosages for patients with different Prakriti.

Determining a patient's Prakriti typically requires at least an hour-long interview by an Ayurveda practitioner. Now, reports the consortium of scientists in the peer-reviewed PLOS One, a computer has been taught to parse 'prakriti' using a database of 147 healthy individuals, whose prakriti was already prepared by Ayurveda practitioners. The researchers, through sifting data on patients, part of long-term investigations in "North and West" India, zeroed in on 65 key questions that can accurately parse Prakriti.

"The next step would be to develop an app or some such," Mitali Mukherjee, a senior scientist at the CSIR-IGIB told The Hindu. "And it is a useful tool if we want to classify large groups of patients." Bhavan Prasher, an Ayurveda researcher at the CSIR, and computation biologists at the Indian Statistical Institute, Kolkata, were critical to combing these data sets for preparing a questionnaire, Mr. Mukherjee added.