Abstract— Fluid retention, known medically as edema, is caused by the retention of fluid in the soft tissue of the lower extremities. This is most commonly found in the ankles and feet due to the effects of gravity. In this paper, we present a wearable device worn around the ankle that monitors edema in the legs and alerts the user of changes. We discuss the Edemeter system’s physical and functional design. We also present results from several experiments characterizing the use of flex sensors for measuring ankle swelling, as well as system component power consumption and its impact on battery life. I. INTRODUCTION Congestive heart failure is the most common cause of hospitalization and readmission in the United States, costing $39.2 billion dollars annually [1]. Monitoring fluid volume status in patients with chronic heart failure outside of the clinical setting is widespread. The current method for monitoring the condition involves weighing the person several times a day. This can be unreliable since weight gain and loss is affected by a range of factors, in addition to fluid retention. Also, patients can often forget to measure or misremember their weight. We present a wearable system to monitor lower extremity edema, a common sign of an impending heart failure exacerbation [2], as a way to prevent hospital admissions. Edema can be exacerbated by salt intake and prolonged standing; and it is associated with a number of conditions aside from heart failure, including venous insufficiency and deep vein thrombosis. We have developed the Edemeter prototype, composed of a Bluetooth-enabled sensored ankle cuff, leveraging data from a flex sensor. With the flex sensor, as the swelling progresses or regresses, the sensor will conform to the shape of the ankle. Following some local processing with a mini-microcontroller, the data from the ankle cuff is wirelessly transmitted to the user’s smart phone for storage and for interfacing with the user. Figure 1 provides an overview of the system hardware architecture. In our experimental results, we examine the feasibility and accuracy of a wearable and continuous approach for detecting edema. Specifically, we examine the sensitivity and accuracy of a variety of flex sensors in estimating small, on the order of a few millimeters, changes in the ankle size. We look at the consistency of the data over extended periods of time, up to an hour. We also examine the power consumption of various A. Resendiz is a Research Assistant with the Mobile Computing Group of the Computer Science Department at California State University, Northridge, CA 91330-8281, USA (e-mail: alex.resendiz.739@my.csun.edu). D. Odicho is a Research Assistant with the Mobile Computing Group of the Computer Science Department at California State University, Northridge, CA 91330-8281, USA (e-mail: dani.odicho.319@my.csun.edu). protocols for processing and transmitting the data from the ankle to a paired mobile device. Figure 1. Edemeter hardware architecture composed of a flex sensor affixed to a wearable cuff, an Arduino Mini Pro microcontroller, and a Bluetooth HC-06 module to transfer data to the user’s smart phone. The remainder of this paper provides an overview of the related work in lower extremity monitoring using wearable computing, along with heart failure-induced fluid build-up monitoring approaches. Our Edemeter prototype is described with elaboration regarding its hardware architecture. In depth experimental results and analysis of the Edemeter components in terms of accuracy, consistency, and power consumption are also provided. II. RELATED WORK Wearable computing and sensors placed on users’ shoes have been leveraged in gait and lower extremity monitoring in the research literature. For example, Hwang et al used the sound generated by the user’s footsteps to recognize the terrain on which the user is walking on [3]. Donkrajang et al [4] monitored footsteps using force sensors, leveraging a microcontroller and a ZigBee module to transmit gathered information. Sensored shoes, called smart shoes, have been developed for a variety of applications including detecting diabetic foot ulcers [5] and imbalance [6]. Customizable frameworks for monitoring systems using a variety of sensors have also been developed and explored [7][8]. Due to the significant risks and limited timeframe associated with heart failure exacerbation, research and commercial systems have been developed to create a self- monitoring environment for patient’s symptoms. Examples include continuous blood pressure [9], ECG [10], weight [11], and vital signs monitoring for heart failure patients [12][13]. V. Gabrielian is a Cardiologist with Foothill Cardiology/California Heart Medical Group in Pasadena, CA 91105, USA (email: vgabrielian@foothillcardiology.com) A. Nahapetian is an Associate Professor of Computer Science at California State University, Northridge, CA 91330-8281, USA and an Adjunct Assistant Professor with the Computer Science Department at University of California, Los Angeles, CA 90095-1596, USA (phone: 818- 677-3545; fax: 818-677-7208; e-mail: ani@csun.edu). Edemeter: Wearable and Continuous Fluid Retention Monitoring Alex Resendiz, Dani Odicho, Victor Gabrielian, and Ani Nahapetian, Member, IEEE 978-1-5090-3087-3/16/$31.00 ©2016 IEEE 153