IIM Visakhapatnam researchers develop machine learning-based MMR dashboard

IIM Visakhapatnam researchers develop machine learning-based MMR dashboard: Researchers at the Indian Institute of Management in Visakhapatnam have developed a dashboard/decision support system (DSS) for maternal mortality. Using NFHS (national family health survey)-4 and NFHS5 data as well as data from the Union government’s health management information system (HMIS), they constructed the interface by employing sophisticated, yet explicable machine learning models.

This will specifically recommend healthcare interventions tailored to administrative units/districts in order to control maternal mortality rates (MMR). As the DSS provides in-depth insights and evaluates the performance of an MMR-related administrative unit, IIMV has initiated discussions with the government to extend its potential benefits to the entire nation in the larger interest of society.

IIM Visakhapatnam researchers develop machine learning-based MMR dashboard

Dr. Shivshanker Singh Patel, creator of the dashboard and director of the Inter-disciplinary Decision Science and Analytics Lab (IDeAL) at IIM Vizag, told the Times of India that maternal health is an urgent global public health issue, particularly in developing nations. Significant disparities in maternal health service utilisation and maternal mortality exist within and between states in a populous nation like India. A more targeted, precise healthcare strategy is required to effectively address this challenge. Dr. Shivshant Singh Patel explained that the DSS is founded on approximately 30-35 variables, such as medical insurance, agroclimatic conditions, accessibility to healthcare, topography, nutrition, health infrastructure, literacy levels, etc.

Dr. Patel added that they employed several explainable AI models to generate insights and recommendations by analysing an exhaustive list of MMR-related factors. “One of the algorithms used to categorise districts based on lower and higher MMR classes is CART heuristics.

Shapley additive explanations, which incorporate machine learning techniques such as’support vector machines,’ ‘artificial neural networks,’ ‘boosting,’ and ‘random forest,’ are also utilised to identify regions with varying MMR levels. In addition, a ‘explainable boosting machine’ was utilised, and the results were meticulously compared to provide useful policy recommendations, according to IIM Vizag’s decision science faculty.

The combination of precision healthcare strategies and explainable machine learning techniques offers a promising opportunity to address the complexities of maternal health disparities. With a concentration on personalised interventions and evidence-based policymaking, the field can make significant strides in improving maternal health and reducing maternal mortality rates, according to Dr. Shivshanker Singh Patel, an expert in machine learning with over a decade of experience.

AIIMS New Delhi Invites Applications for 528 Sr Resident/Demonstrator Posts, Apply here

Major Highlights

• IIM Visakhapatnam has established the ‘Interdisciplinary Decision Science and Analytics Lab’ (IDeAL).
• IDeAL is an inter-disciplinary research centre of excellence.
• The first dashboard developed by IDeAL concerned the Covid-19 pandemic.
• Now, researchers at IDeAL have devised the maternal mortality dashboard/decision support system.
• The MMR dashboard utilised explainable machine learning models applied to NFHS4, NFHS5, and HMIS data.
• The dashboard recommends administrative unit-specific healthcare interventions.
• Before deciding on these interventions, a number of factors pertaining to the specific region will be considered.
• The variables may include healthcare infrastructure, geography, agroclimatic conditions, insurance, etc.
• Policymakers can use the DSS to modify ongoing interventions or to customise their targeted interventions.