Improving health services of rural areas in Indian healthcare system through Machine Learning approaches
One of the challenging issues in healthcare domain is the preterm birth(PTB). PTB babies face serious problems like very low catch-up growth rate, low-birth-weight, often lifetime-disability, visual and hearing problems. Survivors have high chance of brain paralysis, breathing difficulty, and so on. For carrying out any fruitful research on PTB for maternal health as well as early warning system, we need complete, correct, up-to-date and consistent data which comes from digitized database. In practice, data related to pregnant women are usually collected from paper-form or often collected using google form. This research introduces a computerized database of PTB and non-PTB cases, where paper-form data related to pregnant women who delivered pre-term or full-term birth, are collected from Bokaro hospital of Jharkhand (India) and converted first manually to digitized form. Statistical analysis over this dataset is carried out to identify the leading parameters causing PTB. It also reveals the tribal areas to pay attention on maternal health. The research next focuses to design regression-based medium-term forecasting model and machine-learning based early prediction model for PTB cases. At present, a computerized data-collection tool (named as Birth-Curator) to store data of pregnant women at scale in the database is developed. The aim is to increase data availability, completeness, correctness, currency and concordance, and to reduce data loss, research time and cost. Afterall, users (or the persons concerned), practitioners and the researchers will be benefitted in their own interest from the database.