Page 4 - Read Online

Page 2 of 39 Yang et al. Soft Sci. 2025, 5, 46 https://dx.doi.org/10.20517/ss.2025.44

The unique advantages of flexible optical waveguides in sensing, especially in human physiological signal sensing,
are demonstrated through detailed theoretical analyses. Their specific applications in human health assistance are
summarized under each category. Finally, this review proposes evolution paths for flexible optical waveguides by
addressing current bottlenecks through material innovation (e.g., hybrids, metasurfaces), functional enhancement
(e.g., self-powered sensing), and system integration (e.g., miniaturization, Internet of Things platforms).

Keywords: Flexible optical waveguide, sensing principle, invasive biomedical therapy, physiological information
monitoring, interactive soft robot

INTRODUCTION
In recent years, the global attention paid to human health has significantly increased, and health needs have
become multidimensional and refined. The Fourteenth General Programme of Work (GPW14) 2025-2028,
published by the World Health Organization (WHO), states that chronic noncommunicable diseases have
become the leading threat to global health, causing approximately 41 million deaths annually, accounting
[1]
for 74% of all global deaths . Among these diseases, the incidence of cardiovascular diseases, diabetes, and
[2]
cancer continues to rise, with cardiovascular diseases alone causing 17.9 million deaths annually . The
management of these diseases is highly dependent on long-term, accurate monitoring of physiological
parameters, such as real-time tracking of blood glucose fluctuations in diabetic patients and dynamic
[3]
monitoring of blood pressure and hemodynamic changes in patients with cardiovascular disease . In terms
of clinical treatment, taking cancer as an example, common pharmacological treatments and electrical
stimulation therapies have had little effect, and cancer remains the second leading cause of death worldwide.
Clinical treatments would be much more efficient if they could be directed at a more fundamental cellular
level by causing or inhibiting neuronal activity in brain regions or even directly targeting diseased cells
without damaging healthy cells . In the face of this situation, there is a need for health assistance equipment
[4]
with high sensitivity, low cost, portability, real-time monitoring ability, and efficient diagnosis and
treatment characteristics to build a more resilient health insurance system.

As an important part of health assistance, health sensing, which involves monitoring physiological signals
from the human body for health condition monitoring and disease prevention, has gained considerable
momentum. Augustus Desiré Waller first captured human electrocardiographic signals (ECGs) via a
capillary galvanometer that occupied an entire room in 1887, and by 1903, Willem Einthoven, who won the
1924 Nobel Prize, had improved the string galvanometer, enabling the clinical use of the ECGs, which led to
the beginning of biosignal monitoring . The devices of that time were groundbreaking, but they were
[5,6]
bulky and had low sensitivity, making them less practical in reality. Since then, advancements in sensor
technology have witnessed persistent refinements and evolutions: transitioning from bulky, inherently rigid
sensors posing physical risks to the human body to highly integrated and miniaturized variants, and
ultimately progressing into the contemporary epoch of flexible sensors . To date, different kinds of flexible
[7]
sensors have been used for health monitoring; for example, flexible electrochemical sensors based on
detecting target analytes in tears [8-10] , saliva [11,12] , sweat [13,14] , and so on for physiological signal monitoring.
These sensors are often characterized by high performance, miniaturization, and low cost; however, they are
susceptible to electromagnetic interference and highly dependent on the electrolyte and the environmental
conditions of the electrolyte, such as the pH and temperature; thus, their long-term stability and
[15]
biocompatibility must be improved . Another commonly used flexible sensor is the strain sensor, which
can be divided into four types, namely, resistive [16-18] , capacitive [19,20] , piezoelectric , and friction electric ,
[21]
[22]
which realize sensing through the detection of deformation or pressure-induced electrical signals and are
commonly used to monitor physiological signals, such as joint and muscle movements, pulse rates, heart
[23]
rates (HRs), and respiratory rates . Similar to electrochemical sensors, these sensors are also susceptible to