Standardising well being data with AI, deep studying – Instances of India

Kiran Kalakuntla, Srikanth Samudrala and Noel Coutinho began out in 2015-16 to digitise medical data to assist NRIs preserve a test on the well being of their members of the family in India. In the present day, the Hyderabadbased enterprise, ekincare, has metamorphosed into an AI-powered well being advantages platform, catering to over 500 MNCs and Indian corporations, and their staff.
ekincare has a patented healthcare data standardisation and personalisation instrument that makes use of AI, ML, deep studying and analytics. Kalakuntla, who’s CEO, says an enormous downside in India is that diagnostic centres and hospitals use totally different nomenclature for numerous exams. ekincare’s patented algorithm helps convert studies from a number of labs into structured knowledge, which in flip helps generate well being analytics for people, and inhabitants analytics for different clients.
Each time a take a look at report is shared with ekincare within the PDF format, it’s put by means of OCR (optical character recognition). This converts the PDF into textual content, from which all of the related data like take a look at element names, outcomes and ranges are extracted utilizing the proprietary mannequin. The mannequin assigns a novel identification code known as LOINC (Logical Remark Identifiers Names and Codes), which is a common customary used within the US to establish medical take a look at parts no matter the identify utilized by diagnostic centres. The outcomes are then color coded into inexperienced, amber and crimson robotically from the ranges offered, making it simpler for sufferers and medical doctors to rapidly zero in on irregular outcomes. It additionally extracts notes written by medical doctors in x-ray, USG and ECG studies, and shows them in a separate part.
Kalakuntla says three extra patents are pending. The enterprise has raised $40 million thus far, and has 260 staff.