Evaluating healthcare access in informal settlements using satellites and neural networks

Abstract

The right to affordable access to hospitals, clinics and dispensaries is one of the major priorities of government authorities. The constant migration towards cities and consequent growth of informal urban settlements make it imperative to consider the 'grid of influence' of such facilities. An attempt is made to understand this influence through satellite imagery and neural networking. Parameters such as distance from access roads and different types of informal settlements were analysed to understand their significance. These parameter models were then used as inputs for training a neural network in a case study area. Statistical plots helped understand the weight of the parameters, and accordingly, a spatial probability map was prepared. The results of the study could aid development officials in their various plans and proposals for their urban jurisdiction, helping increase accessibility to health facilities using modern predictive tools.