Microcontroller based health care monitoring system using sensor network

2012 7th International Conference on Electrical and Computer Engineering 20-22 December, 2012, Dhaka, Bangladesh 272 Microcontroller Based Health Care Monitoring System Using Sensor Network Md. Manirul Islam1,*, Fida Hasan Md. Rafi1,**, Mohiuddin Ahmad1,***, and Abu Farzan Mitul1,****, T. M. N. Tunku Mansur2, M. A. Rashid2,***** 1 Department of Electrical and Electronic Engineering, Khulna University of Engineering and Technology Khulna-9203, Bangladesh 2 School of Electrical System Engineering, University Malaysia Perlis, Malaysia * mmislameee@gmail.com, **fidaeee@yahoo.com, ***ahmad@eee.kuet.ac.bd, ****mitulnaya@yahoo.com, ***** abdurrashid@unimap.edu.my Abstract— Non invasive and continuous measurement of blood pressure becomes popular now a day. This is very effective way for hypertensive patients to prevent cardiovascular problems and precisely regulate anti-hypertension cures. Blood pressure can be measured by conventional cuff-based sphygmomanometer or invasive instruments in non continuous mode. In this paper, we propose the microcontroller based continuous non invasive cuff less blood pressure measurement system with an alarm circuit for health care monitoring system. Light signal is used in sensor network section of this embedded system as light does not have any harmful effect on human body when it works in continuous mode. Pulse rate calculation and body temperature determination is also embedded in this system using sensor network. Accuracy of the system is found in acceptance range by comparing the results with the existing conventional systems. If BP reading, heart rate or body temperature exceeds the standard range for any patient, the system is able to notify using an alarming circuit. The whole system is controlled by microcontroller ATMEGA8L. The overall system is reliable, accurate, portable, trust worthy, user friendly and cost effective. Index Terms— non invasive BP, SP, DP, pulse rate, body temperature, PPG signal, ARS, health care, alarm circuit I. INTRODUCTION Health care monitoring system is an important means for hypertensive patients. Previously many authors worked on this concept using different methods. Our developed system can introduce a new approach in healthcare monitoring system by the use of volume oscillometric method and sensor network for continuous BP and temperature measurement. Non invasive continuous BP measurement process can be organised without calibration means using PWV methods [1]. Although the standard deviation (SD) and mean standard deviation (MSD) was in acceptable standard value, it can be further modified with calibration in continuous mode for higher accuracy. Instead of using sensor in particular organ, a wearable health-shirt can be used for 24 hours intensive monitoring using pulse transit time (PTT) method [2]. However the variation of the output from standard scale was high enough and not feasible for critical conditioned patients for 24 hours monitoring [2]. To achieve the advantage of oscillometry and vessel buckling method low cuff pressure process for BP measurement can be used [3]. Since, it was conflicted with the cuff less requirement for continuous monitoring in noninvasive means. Accurate beat to beat systolic blood pressure (SP) calculation using pulse wave transit time (PWTT) method is useful. It has a singular relation with SP. Diastolic blood 978-1-4673-1436-7/12/$31.00 ©2012 IEEE pressure (DP) was left outside of study in daily exercise case [4]. A PTT based BP model and computational algorithm was used to show dramatic variation of blood pressure in surgical stimulation. The considered velocity of blood is in a rigid medium artery [5]. But artery wall expands when blood volume increases. So volume concept can attain appropriate result than PTT. New parameter was introduced QdZ for cuff less non-invasive blood pressure measurement [7]. It is rather complicated and concerned only SP. Near Infra Red (NIR) CCD camera was introduced for continuous non-invasive blood pressure measurement with GPS notification in a compact system [8]. The overall system was costly and bulky for portable blood pressure monitoring means. An algorithm for the estimation of BP from pulse arrival time (PAT) and heart rate using body sensor network was proposed in [9]. PPG signal integrated with ECG signal for continuous non-invasive blood pressure measurement system is helpful in development of body sensor network [10]. From only PPG signal BP can be calculated. Nocturnal BP measurement and monitoring for patient’s during sleep condition was studied in [11]. In this paper the system shows reliable output for BP, heart rate and body temperature using cuffless and non-invasive continuous operating mode with negligible error. This health care monitoring system works in very efficient way to calculate blood pressure (BP), body temperature and pulse rate of a subject. A sensor network consisting of BP sensor and temperature sensor is used to get PPG signal and body temperature. It is shown in Fig.1. Output from the sensor is amplified using double stage amplifier constructed with LM324 Op-amp with variable gain controller. The amplified signal is sampled and quantized in ADC and used for the estimation of BP and heart rate using microcontroller ATMEGA8L. Calibration set and memory storage block are used for accurate data analysis and decision making. Monitoring circuit is also introduced in case of patient’s critical BP, heart rate and temperature condition. Temperature sensor Blood LED LDR Light Finger Fig. 1 Position of sensors in the sensor network. 273 The temperature sensor senses patient’s body temperature change in accordance with the change of patient’s BP and heart rate. Overall processed numerical data from microcontroller is displayed in a mini LCD of the system. This paper is organized as follows: Section II describes the methodology of BP measurement system. Section III explains the simulation results as well as calculation of real results and error. Finally, section IV concludes the entire paper. II. METHODOLOGY SP is the degree of force when the heart is pumping (contracting) and is the maximum blood pressure point. Similarly the DP is the degree of force when the hearts is relaxed (expanding) and is the minimum blood pressure point in Arterial BP curve as shown in Fig. 2. Arterial blood pressure 140 SP Pressure (mm Hg) 130 120 microcontroller ATMEGA8L. This microcontroller has 6 channel 10 bit built in ADC. So it makes the overall system a compact one. A preloaded algorithm processes the data and estimates BP and pulse rate of the patient. This data is stored in memory storage for future processing and to compare it with a predefined standard SP and DP of a patient. If any patient`s data exceeds the standard value then the alarm circuit will become activated and will notify patient or the doctor of the patient’s in such critical condition. A temperature sensor IC (LM35) is also used along with the arrangement to determine patient’s body temperature in accordance with the patient’s BP and heart rate. LM35 can detect temperature between -55°C to 155°C .This temperature is proportional to the BP of a patient. So, in hypertension condition patient sweats a lot. The processed digital data from the microcontroller attains numerical value and is displayed in a mini LCD. The mini LCD has 16×2 display section. Only two lines of results can be displayed. All results can be shown in it simultaneously. 110 100 Gain control MAP 90 DP 80 Heart rate and BP sensor 70 Sample and hold Amplifier 60 0 0.5 1 1.5 2 Time (Second) 2.5 Fig. 2 Typical arterial BP curve for human body. Mean arterial blood pressure (MAP) can be calculated from (1). (1) MAP = 13 SP + 23 DP LCD BP has a proportional relation with density of blood in the artery or vessel. The work done by the pulse wave can be expressed in terms of the kinetic energy of the wave and the gravitational potential energy: W = F × l = PB A × l = 12 mv 2 + mgh 2 constant ( g = 9.81m / s )and h is height of blood in the vessel. From (2) PB can be solved. [ ] Where, V is volume of blood (V=Al) and blood ( ρ= m V Sample and hold Temp. Sensor Alarm circuit Speaker Fig. 3 Complete block diagram of health care monitoring system. The overall system connection for BP measurement and monitoring is shown in Fig 4. Alarm circuit will be connected with microcontroller which is not shown in the Fig 4. 5V 10k 1uF 1k 10V density of 500k 50% 1k 100k 4.7k 7 1k ).Density of blood vessels increases in accordance with the increase of blood pressure [6]. This health care monitoring system can easily be described by block diagram as shown in Fig 3. One high bright LED is placed in one side of the patient’s finger for sensing any volume change of blood and a detector LDR is placed in opposite side of the finger to convert the variation of light into variation of resistance. BP is directly proportional to the resistance of LDR discussed in our previous paper [6]. The electrical signal is feed to a double stage amplifier circuit. The amplifier amplifies the low amplitude bio-signal to high amplitude one using gain control mechanism for further data analysis. To suppress dc component from weak bio signal Automatic reference selector (ARS) is used. Its function was discussed elaborately in [6]. The highly amplified signal is sampled. The sampled data is quantized for processing in the Calibration setup ADC (3) ρ is ADC Microcontroller (2) Where, PB is blood pressure (BP), W is work done by pulse wave, F is force, l is length, A is cross sectional area, m is mass of blood, v is velocity of blood, g is gravitational PB = 2mV v2 + Vm gh= ρ 12 v2 + gh Memory Storage Automatic Reference selector 26 LM324 4.7k 10k 1 22 21 20 10V 10k 40% LM324 1uF 14 15 10k 10k 10k 27 28 ATMEGA 8L 10V 500k 50% 5V 5V 25 10k 10V 10V 9 12MHZ SP control 16 17 18 19 10 500k 50% Start switch 8 500k 50% DP control 1k Gain setting LCD (16*2) 1 2 5V 3 4 5 6 7 8 9 10 11 12 13 14 15 16 5V Fig. 4 Complete connection diagram for BP measurement and monitoring. 274 1.00002V -3.00mV 1.00000V -3.05mV 0.99998V LCD2 2.0s 4.0s V(Vi) Time (a) Input weak signal D7 D5 D4 D3 D2 D6 14 13 12 11 D0 E D1 9 10 8 7 R W RS 5 6 VEE VDD 5V U4 14 15 16 17 18 19 9 10 -3.10mV 0s 4 BAT1 3 VSS LM032L 1 III. RESULT AND CALCULATION The double stage amplifier circuit from [6] is used here for extremely high gain mechanism as the detected bio signal is very low in magnitude even cannot be detected in oscilloscope. The highly amplified operation of a low magnitude signal from the amplifier is shown in Fig. 5 and Fig. 6. The developed device is shown in Fig. 9. After inserting finger into the sensor module “Do Calibration” instruction is automatically shown in LCD. In the device three knobs is used for calibration purpose (SP and DP setup) and gain control. It is discussed in [6] elaborately. Then it starts to measure BP continuously. 2 The output of double stage amplifier circuit is connected to pin 26 of the microcontroller. Pin 27 and 28 are used for DP and SP control for calibration means (tuning) [6]. A LCD display is interfaced with the ATMEGA8L. 0s 2.0s 4.0s V(Vo1) Time X1 21 20 PC0/ADC0 PC1/ADC1 PC2/ADC2 PC3/ADC3 PC4/ADC4/SDA PC5/ADC5/SCL PC6/RESET PD0/RXD PD1/TXD PD2/INT0 PD3/INT1 PD4/T0/XCK PD5/T1 PD6/AIN0 PD7/AIN1 AREF AVCC 23 24 25 26 27 28 1 2 3 4 5 6 11 12 13 U1 1 37.0 ATMEGA8 12MHz st PB0/ICP1 PB1/OC1A PB2/SS/OC1B PB3/MOSI/OC2 PB4/MISO PB5/SCK PB6/TOSC1/XTAL1 PB7/TOSC2/XTAL2 (b) 1 stage output of amplifier 2 VOUT 3 LM35 Fig. 5 Input and 1st stage output wave shape of amplifier. Fig. 8 Connection diagram of body temperature measurement circuit. 2.65V 2.60V 2.55V 0s 2.0s 4.0s V(Vo2) Time nd Fig. 6 2 stage output wave shape of amplifier. (a) High amplification can be clarified in oscilloscope where input is in channel 1 in mV range and output is in channel 2 in in V range. It is shown in Fig 7. (b) (c) Fig. 9 Complete device. (a) Displaying BP, (b) Calibration and gain setup and (c) Sensor attachment with finger. Using the device, several data of a patient has been recorded during his relaxed and after exercise condition. Pressure (mm Hg) 130 120 SP 110 100 90 80 70 (a) (b) 60 0 Fig. 7 (a) Weak input signal versus output signal at high gain and (b) Weak input signal versus clipped output signal due to saturation. 48 72 96 120 144 168 192 216 240 264 288 Time (Sec) Fig. 10 Measurement of blood pressure at relaxed condition. 135 130 Pressure (mm Hg) Fig.7 (b) shows that output is clipped due to its saturation effect which is not desirable. So, a gain control mechanism is used in the system to prevent saturation effect of the output. This amplified signal is further processed for BP and pulse rate calculation as discussed above. The temperature measurement section is simulated and it is shown in Fig. 8. In Fig. 8 IC LM35 is used as the temperature sensor. It is connected with ATMEGA8L to show the body temperature in LCD. The temperature is accurately measured as shown in Fig. 8. This verifies the proper working of the temperature sensor in the system. 24 SP 125 DP 120 MAP 115 110 105 100 95 90 0 24 48 72 96 120 144 168 192 216 240 264 288 312 Time (Sec) Fig. 11 Measurement of blood pressure after exercise. 275 Blood pressure of PPG system (mm Hg) The SP, DP and MAP of the patients show reasonable variation in the output for different conditions. The results are in acceptable standard for practical usage in daily life. Continuous BP curve of a subject at normal and excited condition is shown in Fig. 10 and Fig. 11 respectively. In fig. 12, Red dotted points on the straight line show zero deviation between cuff based and PPG based system. The separation between red dotted points and straight line indicates deviation. So after collecting data of several patients using the developed device and Sphygmomanometer, it is plotted in Fig 12. The points are almost in-line with the standard line which indicates less deviation. 190 REFERENCES 170 [1] 150 130 [2] 110 90 70 [3] 50 50 70 90 110 130 150 170 190 Blood pressure of cuff system (mm Hg) Fig. 12 Deviation between cuff based and PPG based BP measurement. For six subjects, SP and DP are measured 3 times for each. Then MAP is calculated. Standard deviation (SD) of SP, DP and MAP are calculated and termed as SDSP, SDDP and SDMAP respectively. These values are shown in Table I. Sub. SP (mm Hg) DP (mm Hg) A 120 121 120 128 127 128 130 131 132 126 125 124 135 136 135 145 143 144 82 81 80 84 89 88 85 86 85 82 81 83 87 88 88 92 91 92 C D E F BP measurement MAP SDSP (mm (mm Hg) Hg) 94.67 94.33 93.33 98.67 101.67 101.33 100 101 100.67 96.67 95.67 96.67 103 104 103.67 109.67 108.33 109.33 [5] SDDP (mm Hg) SDMAP (mm Hg) [7] 0.47 0.82 0.57 [8] 0.47 2.16 1.34 0.82 0.47 0.42 0.82 0.82 0.47 0.47 0.47 0.42 0.82 0.47 0.57 [9] [10] [11] SD of SP, DP and MAP are very less. Therefore the developed BP measurement device can easily be used in noninvasively continuous mode in daily life for 24/7 hours BP monitoring with reliable accuracy. View publication stats [4] [6] TABLE I BP MEASUREMENT AND SD CALCULATION B IV. CONCLUSIONS BP measurement using microcontroller based PPG system for continuous mode is user friendly, flexible and reliable for patients. We developed the BP measurement device considering all of these facts and found our results to be accurate enough for clinical use and all other daily needs. Implementation of microcontroller and sensor network in the system has evolved a new approach of health care monitoring system. Moreover the alarming system notifies cardiovascular problems at initial stage to the hypertensive patients. The healthcare monitoring system continuously monitors BP, heart rate and body temperature with reliable accuracy and alarms the hypertensive patients about the risk of heart stroke. Y. Chen, C. Wen, G. Tao and M. Bi, “A New Methodology of Continuous and Noninvasive Blood Pressure Measurement by Pulse Wave Velocity’’, Proceedings of the 11th International Conference on ICARCV, pp. 1018-1023, December, 2010 . C. H. Chang and Y. T. Zhang, “Continuous and Long-term Arterial Blood Pressure Monitoring by using h-Shirt’’, Proceedings of the 5th International Conference on Information Technology and Application in Biomedicine, in conjunction with the 2nd International Symposium and Summer School on Biomedical and Health Engineering, pp. 267269, May,2008. C. Y. Ding, C. N. Huang, F. M. Yu and H. Y. Chung, “Continuous Blood Pressure Measurement System Based on Low Cuff Pressure Approach’’, Proceedings of the Annual Conference on SICE, pp. 779782, September, 2007 . J. Lass, K. Meigas, D. Karai, J. Kaik, M. Rossmann, “Continuous blood pressure monitoring during exercise using pulse wave transittime measurement’’, Proceedings of the 26th Annual International Conference of the IEEE EMBS, pp. 2239-2242, September, 2004. P. Fung, G. Dumont, C. Ries,C. Mott, M. Ansermino, “Continuous Noninvasive Blood Pressure Measurement by Pulse Transit Time”, Proceedings of the 26th Annual International Conference of the IEEE EMBS, pp. 738-741, September, 2004. Md. M. Islam, F. H. Md. Rafi, A. F. Mitul and M. Ahmad, “Development of a Noninvasive Continuous Blood pressure Measurement and Monitoring system’’, Proceedings of the International conference on ICIEV, pp. 1085-1090, May, 2012. M. Y. M. Wong, E. P. MacPhreson and Y. T. Zhang, “Impedance Cardiography for Cuffless and Non-invasive Measurement of Systolic Blood Pressure’’, Proceedings of the 31st Annual International Conference of the IEEE EMBS, pp. 800-802, September, 2009. O. Krejcar, Z. Slanina, J. Stambachr, P. Silber and R. Frischer, “Noninvasive Continuous Blood Pressure Measurement and GPS Position Monitoring of Patients’’, Proceedings of the 70th IEEE Vehicular Technology Conference Fall (VTC 2009-Fall) , pp. 1-5,2009. F. S. Cattivelli and H. Garudadri, “Noninvasive Cuffless Estimation of Blood Pressure from Pulse Arrival Time and Heart Rate with Adaptive Calibration’’, Proceedings of the 6th International Workshop on Wearable and Implantable Body Sensor Network, pp. 114-119, 2009. G. Fortino and V. Giampa, “PPG-based Methods for Non Invasive and Continuous Blood Pressure Measurement: an Overview and Development Issues in Body Sensor Networks’’, Proceedings of the IEEE International Workshop on Medical Measurements and Applications Proceedings (MeMeA), pp. 10-13, 2010. H. R. Dajani, R. S. T. Leung, “The Measurement of Blood Pressure During Sleep’’, Proceedings of the IEEE International Workshop on Medical Measurements and Application, pp. 120-123, May, 2008.