Title:Internet of medical things for non-invasive and non-contact dehydration monitoring away from the hospital: state-of-the-art, challenges and prospects

Abstract:Dehydration occurs when the body loses more water than it takes in. Mild dehydration can lead to fatigue, cognitive impairments, and physical complications, while severe dehydration can cause life-threatening conditions like heat stroke, kidney damage, and hypovolemic shock. Traditional bio chemistry-based clinical gold standard methods are expensive, time-consuming, and invasive. Thus, there is a pressing need to design novel non-invasive methods that could do in-situ, early and accurate detection of dehydration, which will in turn allow timely intervention. This article presents a methodological review of the literature on a range of innovative internet of medical things-based techniques for dehydration monitoring. We begin by briefly describing the pathophysiology of the dehydration problem, its clinical significance, and current clinical gold-standard methods for assessing hydration level. Subsequently, we critically examine a number of non-invasive and non-contact hydration assessment studies. We also discuss multi-modal sensing methods and assess the impact of dehydration among specific population groups (e.g., elderly, infants, athletes) and on different organs. We also provide a list of existing public and private datasets which make the backbone of machine learning-driven research on dehydration monitoring. Finally, we provide our opinion statement on the challenges and future prospects of non-invasive and non-contact hydration monitoring.