**1. Introduction**

Arrhythmia is the presence of abnormal cardiac rhythms. In 2018, more than 500,000 American deaths included arrhythmia as a contributing factor, demonstrating its deleterious impact on patient health [1]. Furthermore, the lifetime risk of atrial fibrillation in the United States is estimated to be one in three among Caucasians and one in five among African Americans [2]. Arrhythmias occur when the electrical pulses of the heart are not functioning properly, causing the heart to beat either too fast, too slow, or skip beats. Impulse-production arrhythmias can be grouped into six categories: premature beats, non-sinus rhythm, fibrillation, tachycardias, bradycardias, and flutter. Premature beats are abnormally timed beats that occur before the sinus rhythm and are caused by the heart being unable to fill with the appropriate amount of blood [3]. Atrial fibrillation, the most common arrhythmia, occurs when the electrical pulses between the upper chambers of the heart, the atria, do not sync with the pulses in the lower chambers of the heart, the ventricles. Ventricular fibrillation, on the other hand, occurs when there is a mismatch between the right and left atria, which makes the heart unable to pump blood to the

**Citation:** Guess, M.; Zavanelli, N.; Yeo, W.-H. Recent Advances in Materials and Flexible Sensors for Arrhythmia Detection. *Materials* **2022**, *15*, 724. https://doi.org/10.3390/ ma15030724

Academic Editor: Fabrizio Roccaforte

Received: 1 December 2021 Accepted: 16 January 2022 Published: 18 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

body [3]. Tachycardias occur when the heart is beating too fast, generally more than 100 beats per minute, and bradycardias occur when the heart is beating too slow, generally less than 40 beats per minute. In addition, impulse-conduction arrhythmia types include atrioventricular block, bundle branch block, Wolff—Parkinson—White syndrome, and escape beats [4]. An atrioventricular block occurs when the impulses between the atria and ventricles become blocked due to a failure in the heart's conduction system. Bundle branch block occurs as a result of blockages in the pathways in the heart. Wolff—Parkinson—White syndrome occurs when additional electrical pathways are made between the atria and ventricles, resulting in a rapid heartbeat [5].

Despite the clear need for early arrythmia detection to avoid these serious complications, existing detection mechanisms have proven insufficient. Arrhythmias have traditionally been diagnosed by medical professionals based on qualitative data, a patient's medical history, and clinical examinations. Electrocardiography (ECG) has proven instrumental in identifying arrhythmias. The importance of continuous monitoring for specific arrhythmias has been increasingly identified, as both asymptomatic arrhythmias and paroxysmal diseases remain difficult to detect through intermittent clinical ECG recordings [6,7]. The 12-lead Holter monitor has long been the clinical standard for detection and diagnosis of heart-rate diseases using long-term monitoring of ECG [8]. Though these devices are widely used, they are prone to poor patient compliance because of their bulkiness and reliance on wired leads [9]. In addition, these devices experience signal deterioration over time due to the drying of the conductive gels [10]. The advent of miniaturized, one-lead devices has offered an alternative to multi-lead ECG devices. However, they are susceptible to motion artifacts that disrupt data collection [11].

Flexible devices have emerged as alternatives to these rigid devices, eliminating motion artifacts by increasing sensor-to-skin adhesion. Recently, new areas of research have been developed to make these heart-rate monitoring devices cheaper and faster to manufacture, expanding accessibility for these previously costly devices. Table 1 shows the currently available flexible devices for arrhythmia monitoring. Additionally, new alternatives to ECG can provide information that ECG alone cannot. For instance, photoplethysmography, ultrasound, seismocardiography, and ballistocardiography can characterize the heart's electromechanics. Figure 1 shows these soft-sensor types, along with the desirable qualities of the devices. New arrhythmia-detection methodologies also offer more accurate, automatic information to patients for a low cost. For example, deep neural networks, which are capable of learning important features and patterns without extensive preprocessing or feature engineering, are becoming extremely accurate in predicting types of arrhythmia [12].

In this review, we summarize the types of sensors used to detec arrhythmias, with an emphasis on non-implantable devices and recent advances in the flexibility of previously rigid sensor types. Different sensing methods can offer low-cost alternatives to traditional sensing or provide information that is unobtainable with other sensors. We explore the materials needed to fabricate these flexible devices and discuss the mechanical, chemical, and electrical properties of these materials, as well as the effect of these properties on the detection of arrhythmias. Next, arrhythmia detection methodologies of various arrhythmiadetection devices are explored, along with their limitations. Finally, we comment on the current cutting-edge research in the field, the current problems and possible solutions, and the future development of wearable sensors for arrhythmia detection.

– – – – **Figure 1.** Examples of flexible sensors and functions for accurate arrhythmia detection [13–16]. (Figures are adapted or reprinted, clockwise, from the top–left: (1) *Sensors Actuators A Phys.* 2018, *272*, 92–101, Copyright 2019, Elsevier; (2) *Proc. Natl. Acad. Sci.* 2018, *115*, E11015–E11024, Copyright 2018, National Academy of Sciences; (3) Creative Common License by MDPI; (4) Creative Common License by Wiley.


