top

Heart Project

Interpretable AI for Healthier “Blue-Green” European Cities

Since the industrial revolution, the global climate is affected by man-made activities, most of them with negative environmental impact. Climate change (CC) severely engenders opportunities for a variety of natural disasters, resulting in a severe threat to the environment. For instance, environmental deterioration poses significant risks to urban environments.

In addition, public health (PH) is affected by the disruptions of physical, biological, and ecological systems caused by the environmental burden in big cities. Therefore, the significance of city-level actions for enhancing CC mitigation and adaptation is being increasingly recognized. Most of the policy responses are translated to mitigation actions attempting to reduce environmental pollution and improve PH and well-being (WB). Our idea lies in the fact that explainable AI (XAI) is the “language” to translate the results of the AI model, in an easy-to-understand and interpretable way, to the urban planners, to support them in the decision-making process for health-centered urban planning decisions.

The main scope, of handling CC and building healthier cities using AI, is not to have powerful AI tools, but to propose smart solutions at where and how to use AI. Our tool is an XAI framework that embodies the consensus of various expert fields (health-related, environmental, urban planning, policymakers, etc.) and gives insights into the impact of these different factors on public health and well-being. This tool can be utilized by urban planners and policymakers and assist them to propose adverse-event measures and building a mitigation strategy to slow down the rising of the non-communicable disease burden (figure above).

This article is part of our scientific paper submitted to the Remote Sensing Journal: “Temenos, A., Tzortzis, I.N., Kaselimi, M., Rallis, I., Doulamis, A. and Doulamis, N., 2022. Novel Insights in Spatial Epidemiology Utilizing Explainable AI (XAI) and Remote Sensing. Remote Sensing14(13), p.3074.” for the needs of HEART project.

Skip to content