Page 23 - Read Online

Page 2 of 44 Jung et al. Soft Sci 2024;4:15 https://dx.doi.org/10.20517/ss.2024.02

presented underscore the adaptability, versatility, and inherent efficacy of these sensors in addressing the
multifaceted challenges intrinsic to DM and its associated complications within academic discourse.

Keywords: Wearable electronics, electrochemical sensor, biosensor, diabetes mellitus, machine learning

INTRODUCTION
The global epidemic of diabetes has become evident with the advent of widespread industrialization and a
[1]
substantial escalation in obesity rates . According to the International Diabetes Federation, the global
prevalence of diabetes mellitus (DM) among individuals aged 20 to 79 was estimated at 537 million in 2021.
According to the International Diabetes Federation, this number is expected to grow to 643 million by 2030
and further escalate to 783 million by 2045 . DM, an incurable chronic disease, is identified as a metabolic
[2]
disorder marked by persistent hyperglycemia. Its pivotal characteristic involves as inherent insufficiency in
the secretion and/or action of insulin by pancreatic β cells, whether absolute or relative . In instances of
[3,4]
severe hyperglycemia, common clinical symptoms include increased thirst (polydipsia) and excessive
urination (polyuria). In extreme cases, DM can lead to coma. On the other hand, in mild cases of
hyperglycemia, patients with diabetes might not show any signs, and if ignored, this situation could result in
fatal outcomes . Also, diabetes is intricately associated with and actively contributes to a diverse range of
[5]
complications, encompassing cerebrovascular disease , cardiovascular symptoms [9-11] , diabetic kidney
[6-8]
disease (nephropathy) [12-14] , Pancreatogenic diabetes [15-17] , and other related conditions [18-21] . Given the
substantial health risks associated with DM, it is imperative for individuals to consistently monitor and
maintain optimal concentrations of key biomarkers pertinent to diabetes and its associated complications,
as outlined in Tables 1 and 2.

As a solution for point-of-care systems in managing DM, blood glucose monitoring has played a critical
part, and over the past three decades, with the introduction of continuous glucose monitoring (CGM)
devices, tremendous progress has been made toward commercialization [58-60] . CGM technologies leverage
electrochemical and optical methods for real-time glucose monitoring. Electrochemical CGMs, prevalent in
current use, operate based on enzyme reactions that produce an electrical signal correlating with glucose
concentrations. Despite their proven accuracy and capability for real-time monitoring, these devices
encounter limitations, including relatively short lifespans, the necessity for regular calibration, potential for
skin irritation, and their restriction to monitoring glucose alone. The optical CGM introduces a different
approach, utilizing light and fluorescent chemistry, which potentially extends the lifespan of sensors and
reduces the need for frequent replacements. However, it faces challenges similar to those of its
electrochemical counterparts, such as skin irritation, and is also limited to monitoring glucose exclusively.
Notably, users of the optical CGM are required to visit a physician for sensor setup and replacement every
three months . Furthermore, ongoing extensive efforts and research endeavors aim to foster the broad
[61]
acceptance of wearable sensor technology . This involves the implementation of flexible, non-invasive
[62]
wearable sensors designed for seamless integration into lifestyles of individuals, mitigating the discomfort
typically associated with conventional finger prick blood testing methods. Additionally, this technology
enables multiplexed sensing capabilities using millimeter-long microneedle arrays (MNAs) and other non-
invasive methods, which not only facilitate minimal tissue inflammation and rapid skin recovery but also
allow for the simultaneous measurement of multiple analytes .
[63]

These wearable sensors serve as pivotal tools in preserving optimal health for individuals with DM and
preventing associated complications. They enable real-time monitoring of key metabolites, including
glucose (a gold standard biomarker for DM) [64-66] , lactate (relevant to diabetic kidney disease) [67-69] , ketones
[associated with diabetic ketoacidosis (DKA) and potential coma] [70-72] , uric acid (linked to cardiovascular