1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2015.2424979, IEEE Sensors Journal > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1 Abstract—Monitoring the photoplethysmogram (PPG) signal is essential for cardiovascular patients in a hospital or at home, as well as for those working in front of a personal computer (PC) at the office every day. Therefore, we developed a wireless PPG mouse that consists of a PC mouse, PPG sensor, and Bluetooth mote. The sensor is located within the PC mouse therefore the structure of the ordinary mouse is not changed. A user’s thumb can easily touch the surface of a sensor for PPG signal monitoring. However, it is challenging to process the signals collected from the PPG mouse, especially in cases where the mouse moves quickly or the user performs multiple actions on the mouse buttons. In this study, we propose a robust algorithm to detect the PPG peak under big motion artifact conditions. In the proposed algorithm, an adaptive method enables simultaneous detection of true peaks and eliminates fake peaks from the acquired PPG signal. Next, these detected error peaks can be corrected by a random error estimator. The combination of two sequential methods enhances the robustness of the algorithm for distinguishing irregular PPG patterns. The proposed algorithm presents an advantage for real-time applications and continuous heart rate monitoring systems using a wireless PPG sensor implemented in a PC mouse. Index Terms—Robust algorithm, photoplethysmogram (PPG), PPG mouse, real-time peak detection. I. INTRODUCTION he ability to collect and process vital or physiological signs from a simple, comfortable, and convenient non-invasive sensor at low cost during normal living conditions is the primary goal of this study. The PPG technique was developed in the 1930s for monitoring blood pulsing under the skin and this technique has offered a very common non-invasive method for diagnosing cardiovascular diseases since that time [1]. PPG is easy to setup, convenient, simple and economically efficient compared to ECG for diagnosing cardiovascular diseases. PPG has been used for analyzing hypoxia based on reflective oxygen saturation [2]. By using PPG waveforms without any additional sensor, continuous cuffless arterial blood pressure as well as Manuscript received February 5, 2015; accepted April 9, 2015. Date of current version April 19, 2015. This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2012R1A1B3004140). The authors are with the Department of Electronic Engineering, Pukyong National University, Busan, 608739, Korea (e-mail: tvthang@pknu.ac.kr, wychung@ pknu.ac.kr). extracting the respiratory activity can be estimated, in turn reducing the number of sensors connected to the objects [3], [4]. PPG signal is useful to study apnea diseases by detection of respiratory arousals [5] and analysis of pulse transit time (PTT) variability during decreases in the amplitude fluctuations of PPG waveforms [6]. Another study surveyed two groups of PPG signals from healthy and diabetic participants to diagnose diabetes [7]. By analyzing amplitude and peak-peak interval (PPI) of the PPG waveforms, various diseases such as cardiovascular diseases, mental stress, apnea disease, respiration activity, diabetes, and so on could be diagnosed ; therefore, PPG has probably become the most widespread method used in clinical monitoring [8]. On PPG technology, some previous studies have made PPG sensors more comfortable for users. For example, an eyeglass device can collect PPG signals from the nose bridge and transmit the data to mobile devices using Bluetooth standard [9]. And in a different work, a tiny sensor chip is integrated inside the ear-mold in the in-ear sensor system to provide a reliable method for collecting a stable PPG signals [10]. However, these wearable devices are sometimes uncomfortable due to the size and weight of an external RF module for wireless communication. Moreover, the signal captured by these wearable devices is not sufficiently strong. Nowadays, most people spend several hours each day on the computer with mouse, therefore, a mouse with PPG sensor could be a good choice to monitor the daily health conditions of the user. Until now, only one research group designed a PC mouse with PPG sensor for monitoring mental stress of daily life [11]. In this device, PPG signals are transmitted to PC by wired connection; the output signals are not sufficiently strong due to the infrared light that is used instead of typical red light. In this study, we suggest an alternative solution of a wireless PPG mouse, developed by using a tiny sensor, a Bluetooth module, and a PC mouse. In this solution, the tiny PPG sensor and Bluetooth module are integrated inside a mouse to capture the PPG signal. This device allows monitoring of the PPG waveform, heart rate, and heart rate variability while the mouse is used in common operations. There are two important data points used in analyzing the PPG signal; one is a maximum point or peak, and another is a minimum point or valley. The most critical PPG processing requires the detection of these peaks and valleys, because it supports the estimation of the heart rate, heart rate variability, A Robust Algorithm for Real-Time Peak Detection of Photoplethysmograms using a Personal Computer Mouse Thang Viet Tran, Student member, IEEE, and Wan-Young Chung, Member, IEEE T