Browsing by Author "Tu, Tse-Yi"

Now showing 1 - 20 of 25

A fast digital chip implementing a real-time noise-resistant algorithm for estimating blood pressure using a non-invasive, cuffless PPG sensor
(1970-01-01)
Wang, Jeremy H. -S.
;
Yeh, Ming-Hua
;
Chao, Paul C. -P.
;
Tu, Tse-Yi
;
Kao, Yung-Hua
A fast digital chip is designed and fabricated successfully to implement the algorithm for estimating blood pressure (BP) in real time by a non-invasive cuffless photoplethysmogram (PPG) Sensor. The fast response accomplished by this digital computation chip is made possible by one-time calibration only and the Taiwan Semiconductor Manufacturing (TSMC) T18 process to realize the chip in an area of 3.97 mm(2). The experimental result shows that the overall power consumption of the chip is 15.62 mW. Finally, the blood pressure measurement platform has been developed with a user interface (UI) for long-term continuous cuffless BP measurement. The non-invasive blood pressure sensor is applied to the wrist arteries of 44 subjects for sensing the PPG pulsation of the blood vessel, and then estimating BPs (blood pressures). Measurement results shows that the maximum error in the BP measurement is +/- 6 mmHg. Which is less than 8 mmHg, defined by the Advancement of Medical Instrumentation (AAMI) and British Hypertension Society (BHS) standard.
A fast-sensing readout circuit in 240 Hz enabled by code division multiple access (CDMA) implemented for an ultra-thin on-cell flexible capacitive touch panel
(1970-01-01)
Chang, Chia Yu
;
Chao, Paul C. -P.
;
Wang, Jeremy H. -S.
;
Su, Ying Cheng
;
Thakur, Smriti
A new fast readout circuit employing the known coding scheme of code division multiple access (CDMA) is successfully designed and applied to a 7-inch ultra-thin, flexible on-cell touch screen panel (TSP). The adopted CDMA is known originally as a coding scheme for data communication, which is applied in this study to address the sensing electrodes of the ultra-thin flexible touch panel. Due to the orthogonality between the driving signals to the touch panel coded by Walsh transform, one type of CDMA, the interference noises between sensing electrodes can be reduced effectively to render accurate touch sensing results. The electromagnetic interference from the flexible display can also be filtered out as baseline component in the output signal. And the frame time of touch reporting can be substantially shortened. Following the sensing electrode is a new readout designed of the switched-capacitor (SC) circuit, to avoid distributing sample signals from parasitic capacitance and also to enlarge the voltage changes due to the capacitance changes caused by touches. A 12-bit analog-to-digital converter (ADC) is orchestrated after the SC circuit to transform the front-end analog signal to digital codes. The digital part of the designed readout adopts a correction algorithm to eliminate the background signals from the display, and also a moving average algorithm to minimize the higher-frequency noises from the display and other electrodes. Experiments are conducted to validate the expected performance. It is evidenced that the Walsh code driving algorithm improves the quality of the readout output signal to be in 42 dB SNR, the report rate to a fast 240 Hz, and a power consumption of 0.39 mW by each sensing channel.
A low-power self-biased rail-to-rail preamplifier as a readout circuit for a capacitor-type microphone
(2013-09-01)
Tu, Tse-Yi
;
Chao, Paul C. -P.
;
Mai, Yueh-Teng
;
影像與生醫光電研究所
;
電機工程學系
This study is dedicated to design a readout circuit to extract the output signal of a capacitive-type MEMS microphone. A low-power preamplifier is forged for a capacitive-type microphone to meet the demands on cell phone applications. The design starts with modeling the electro-mechanical behavior of a biased capacitive-type microphone using combination of AV and DC voltages sources. This aims to enable co-simulation of the microphone and the readout circuit design. The output signal of a capacitance-to-voltage converter is normally small and may cause substantial noise in the output signals. Therefore, a preamplifier is designed and applied to amplify the signal to an acceptable level for the convenience of ensuing signal processing. The designed readout circuit consists of two main sub-circuits, responsible for functions in two different stages. One is a newly-designed self-bias circuit in the structure of a capacitance-to-voltage converter at the first stage, while another is a low-voltage low-power two-stage or rail-to-rail MOS operational amplifier at the second stage. The proposed circuit is implemented by using the technology of TSMC 0.35 A mu m Mixed-Signal MODE (2P4M, 3.3 V/5 V) POLYCIDE. The supply voltage is fixed at low voltage (3 V), the total power dissipation is merely 200 mu W. Experiments are finally conducted to validate the performance of the designed readout circuit from 20 to 20k Hz with appropriate add-on high- and low-pass filters.
A New Cuffless Optical Sensor for Blood Pressure Measuring with Self-Adaptive Signal Processing
(2016-01-01)
Kao, Yung-Hua
;
Chao, Paul C-P
;
Tu, Tse-Yi
;
Chiang, Keng-Yueh
;
Wey, Chin-Long
A new cuffless, wireless optical blood pressure (BP) sensor is successfully developed by this study. This device employs the technique of photoplethysmograph (PPG) to read out in real time the intravascular blood volume change and then to calculate BP. The system utilizes a red-light LED with a wavelengths of 660 nm. The readout circuit includes a pre-amplifier, a band-pass filter, a programmable gain amplifier (PGA), an Arduino unit for calculation and a Bluetooth module for wireless communication. A notebook or a mobile phone is also used to display continuous BPs and conducts statistical analysis/results. The passband is from 0.3 to 3.4 Hz. The range of adjustable gain are from 60 to 80 dB. Measured PPGs of different subjects can be auto-adjusted with varied gain by a PGA with a pre-designed digital circuit. In results, the signal-to-noise ratio (SNR) is improved for accurate estimates on BPs. 8 subjects participated in the experimental validation, in which the obtained BPs are compared with the results from a commercial blood pressure monitor by OMRON. The maximum error of experimental results is +/- 6 mmHg, which is less than 8 mmHg conforming to the requirement by the Advancement of Medical Instrumentation (AAMI). This study pioneers to report stable, accurate estimates on BPs by a cuffless optical hand-held device.
A new hybrid fabrication process for a high sensitivity MEMS microphone
(2013-09-01)
Chao, Paul C. -P.
;
Tsai, Chun-Yin
;
Chiu, Chi-Wei
;
Tsai, Che-Hung
;
Tu, Tse-Yi
A novel multi-layer stacking capacitive type microphone is designed in this study based on theoretical analysis and numerical simulations, while fabricated via two standard stable silicon-based MEMS processes-PolyMUMPs and SOIMUMPs. The adoption of two standardized processes helps greatly to increase yield rate. The sensitivity of the microphone is first determined by an analytical model based on an equivalent circuit, which is followed by finite element (FEM) analyses on the capacitance value, static pull-in voltage and dynamic characteristics. Based on the developed analytical model, varied dimensions of the microphone are optimized and then the performance is validated by analytical simulations. In the next step, micro-fabrication of the microphone is accomplished using two standard processes, PolyMUMPs and SOIMUMPs provided by MEMSCAP. Experiments are conducted to acquire the information of pull-in voltage for safe operation and frequency response in sensitivity for performance evaluation. In the static case, experimental results show a good agreement with the analytical results with 90 Mpa residual stress assumed. As for dynamic validation, the frequency response is measured in an anechoic room adopting the exciting frequency as the audible range from 100-10 kHz. The measured sensitivity is as around 0.78 and 1.7 mV/Pa from 100 to 10 kHz, under the biases of 2 and 4.5 V, respectively. Within the audible frequency range, the proposed device maintains the loss as less as 2.7 dB (ref. to V/Pa), under 3 dB-the commonly acceptable drop within audible frequency range.
A new LCL-lens array with electrodes of interlaced structure to be applied for auto-stereoscopic 3D displays
(2014-08-01)
Kao, Yung-Yuan
;
Chao, Paul C-P.
;
Tu, Tse-Yi
;
電機工程學系
;
Department of Electrical and Computer Engineering
A new design of the liquid crystal lenticular (LCL) lens array with interlaced electrodes is proposed to realize an auto-stereoscopic 3D display. The structure of interlaced electrodes aims to generate a desired non-uniform electric filed to improve the focusing ability of the LCL-lens array with lower voltage drives. This paper is organized as first to describe the design criteria of a lenticular lens sheet for wide extended graphics array LCD panel. Based on the designed lenticular lens array, an LCL-lens array with proposed interlaced electrodes is next designed. The fabrication and the detailed structure are given. A series of experiments are then conducted and successfully verify the focusing capability and desired 3D display functions, that is, to separate two images to right and left eyes. Also, the developed auto-stereoscopic 3D display is able to adapt to varied viewer locations with favorable imaging quality.
A new liquid crystal lens with axis-tunability via three sector electrodes
(2012-09-01)
Tu, Tse-Yi
;
Chao, Paul C. -P.
;
Lin, Chin-Teng
;
影像與生醫光電研究所
;
電機工程學系
A novel liquid crystal (LC) lens with an on-line tunability on focus length and optical axis is proposed in this study. The designed lens has a LC layer sandwiched by two ITO glasses, one of which is patterned with three sector electrodes. With varied sets of pre-designed voltages applied to these three electrodes, the LC lens can not only render focusing effects but also tunability on the optical axis of the lens to an arbitrary axis. A vector-form equation is developed to predict the direction of axis tuning. Simulations are next conducted to predict dynamics of the LCs in the lens and also the focusing and axis-tuning properties of the lens. Important sizes and materials and fabrication process of the lens are determined and optimized based on simulation results. The designed LC lens is fabricated, and then experiments are conducted to demonstrate the performance of the designed LC lens on axis tuning. It shows that the focusing axis of the LC lens can be effectively changed by pre-calculated combinations of three voltages. It is also shown that the average movement of the focal point per applied voltage reaches 4.778 mu m/V.
A New Non-Invasive Cuff-Less Blood Pressure Sensor
(2013)
Tu, Tse-Yi
;
Chao, Paul C. -P.
;
Lee, Yung-Pin
;
電機工程學系
;
Department of Electrical and Computer Engineering
A non-invasive, cuff-less, capable of continuous detection and low cost blood pressure sensor is proposed in this study. The proposed blood pressure sensor has a specially-designed strain sensor attached to a flexible plate and encapsulated by a gel module. The sensor is able to detect pulse signals of the artery vibration under the skin of a human wrist by a simple, flexible attachment pad. The strain sensor is expected to vibrate in a synchronized fashion with artery pulsation. The pulse vibration is expected to be transmitted to the strain sensor through the gel module. This module is made of highly bio-compatible materials like polyethylene (PE), polyurethane (PU), polytetrafluoroethylene (PTFE), polydimethylsiloxane (PDMS), or polymethylmethacrylate (PMMA). The gel module supports the strain sensor to a stability condition with top surface contacting tightly with the human skin while the bottom surface is under a low pressure exerted by a pressurizing wrist belt.
A Portable, Wireless Photoplethysomography Sensor for Assessing Health of Arteriovenous Fistula Using Class-Weighted Support Vector Machine
(2018-11-01)
Chao, Paul C. -P.
;
Chiang, Pei-Yu
;
Kao, Yung-Hua
;
Tu, Tse-Yi
;
Yang, Chih-Yu
A portable, wireless photoplethysomography (PPG) sensor for assessing arteriovenous fistula (AVF) by using class-weighted support vector machines (SVM) was presented in this study. Nowadays, in hospital, AVF are assessed by ultrasound Doppler machines, which are bulky, expensive, complicated-to-operate, and time-consuming. In this study, new PPG sensors were proposed and developed successfully to provide portable and inexpensive solutions for AVF assessments. To develop the sensor, at first, by combining the dimensionless number analysis and the optical Beer Lambert's law, five input features were derived for the SVM classifier. In the next step, to increase the signal-noise ratio (SNR) of PPG signals, the front-end readout circuitries were designed to fully use the dynamic range of analog-digital converter (ADC) by controlling the circuitries gain and the light intensity of light emitted diode (LED). Digital signal processing algorithms were proposed next to check and fix signal anomalies. Finally, the class-weighted SVM classifiers employed five different kernel functions to assess AVF quality. The assessment results were provided to doctors for diagonosis and detemining ensuing proper treatments. The experimental results showed that the proposed PPG sensors successfully achieved an accuracy of 89.11% in assessing health of AVF and with a type II error of only 9.59%.
An efficient energy harvesting circuit for batteryless IoT devices
(2020-01-01)
Yang, Ching-Cheng
;
Pandey, Rajeev
;
Tu, Tse-Yi
;
Cheng, Yuan-Po
;
Chao, Paul C-P
A new energy harvesting circuit for battery-less IoT beacon tags is developed herein to maximize power conversion efficiency as well as high throughput power with a wide input-output range. This design energy harvest (EH) circuit incorporates a charge pump (CP) with shoot-through current suppression, a body selector circuit, a maximum power point tracking circuit (MPPT), a timing control circuit, a hysteresis control circuit and a low dropout regulator. Also in this MPPT circuit is a gated clock tuned in a self-adaptive fashion to match the input impedance of the EH circuit to the output impedance of the photovoltaic (PV) panel, thus achieving successfully maximum power point. The circuit is implemented in an integrated chip in an area of 1.2 mm(2) via the TSMC 0.18 process. Experiments on the chip are conducted and the results show that the input voltage range is allowed from 0.55 to 1.7 V to effectively harvest the solar power from a flexible dye-sensitized solar cell. The achieved peak power conversion efficiency (PCE) is 77% at the input power of 52 mu W. For a wide range of lighting luminance (300-1300 lx,) the achieved average PCE is more than 70%. The achieved wide input-output range and the maximum throughput power of 200 mu W is much larger than others reported, while the 77% of PCE is close to that best power conversion efficiency reported.
An LCL-Lens Array with Electrodes of Interlaced Structure in Applications of Auto-Stereoscopic
(2013-01-01)
Chao, Paul C. -P.
;
Tu, Tse-Yi
;
Kao, Yung-Yuan
;
影像與生醫光電研究所
;
電機工程學系
Continuous blood pressure measurement based on a neural network scheme applied with a cuffless sensor
(2018-11-01)
Tu, Tse-Yi
;
Chao, Paul C. -P.
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
A new computation method of a back-propagation neural network (BPNNs) is designed and expected to implement continuous measurement of blood pressures (BPs) by a noninvasive, cuffless, handheld strain-type BP sensor. The sensor is successfully designed to acquire pulsation signals at the wrist artery of a subject with a readout designed of a Wheatstone bridge, amplifier, filter, and a digital signal processor. To predict BP based on the obtained pulsation signals, 22 features are extracted to compute systolic blood pressure (SBP) and diastolic blood pressure (DBP) based an established BPNN. There are 22 input neurons, 30 hidden layers and 2 output neurons in BPNN model. The inputs are presented to the pulsation signal of time domain and frequency domain and outputs are presented to SBP and DBP. Experiments are conducted to show the validness of the developed sensor and BPNN. The data show that the prediction errors are within mmHg, respectively. SBP and DBP are 1.35 +/- 3.45 and 2.29 +/- 3.28mmHg, respectively. The errors of blood pressure pass the criteria for Association for the Advancement of Medical Instrumentation (AAMI) method 2 and the British Hypertension Society (BHS) Grade B.
Design and implementation of a new high-accuracy interpolation encoder IC for magneto-resistive sensors
(1970-01-01)
Chen, Wen-Yu
;
Chang, I-Feng
;
Chao, Paul C-P
;
Thakur, Smriti
;
Tu, Tse-Yi
A new algorithm named the eight-section (ES) method for carrying out the interpolation for a magneto-resistive (MR) encoder is successfully developed and fabricated in a digital chip. It is known that a conventional magneto-resistive (MR) encoder employs the interpolation method, which converts incoming front-end analog signals in harmonics in sinusoids/co-sinusoids to moving displacement via calculating arc-tangents. This conventional interpolation requires divisions to be carried out for the displacement, which often leads to large noises while conducting digital computation, eventually undermining significantly the accuracy of the MR sensor. The proposed interpolation of eight-section (ES) is designed specially without divisions in the digital computation, leading to higher precision than the conventional interpolation conducting the computation of arc-tangents. The digital computation chip designed by this study consists of a cycle counter, two decimators for incoming analog signals of the MR sensor, a correcting circuit, and the proposal ES interpolation unit. The designed chip is successfully fabricated by TSMC 0.18-mu m CMOS process, the area of which is 1643 x 1676 mu m. The chip is than calibrated by a reference interferometer by experiments for further improving the measurement accuracy. The precision finally results in measuring displacement reaches as accurate as within 1.065 mu m, which is much favorable to the existing performance around 2 mu m by the conventional interpolation.
Design and validation of the readout circuitry for an in-cell active-matrix capacitive (IAMC) touch panel
(2018-11-01)
Ku, Tsung-Yen
;
Chao, Paul C. -P.
;
Tu, Tse-Yi
;
Huang, Chin-Hai
;
Liu, En-Chih
This study is dedicated to design an in-cell active-matrix capacitive (IAMC) touch panel and its readout circuitry. This IAMC touch panel is in a structure that embeds all components for touch panel in a standard active matrix display. In this way, the thickness of the display-touch combined module can be significantly reduced, but introducing much noise from display to touch module, leading to higher mis-judgements on touches. To solve the problem, this study proposes a new readout circuit and digital signal processing techniques to tackle well the aforementioned noise for much accurate touch sensing. The readout consists essentially of an analog front-end switched capacitor charge amplifier (SCCA), a 12-bit SAR ADC and a digital signal processing (DSP) unit for calibration and filtering. A field-programmable gate array (FPGA) board is adopted to implement the designed digital signal processing in practice. The performance of the readout is tested for a 5x12, 0.88-inch IAMC touch panel with a scanning frequency of 200Hz. Experimental data show clearly that the sensed capacitance was successfully converted to 200 digits in a full dynamic range of 12-bits and a SNR of 52dB, which facilitates well accurate determination of touches and locations on an IAMC touch panel.
EFFECTS OF MIS-POSITIONING A NEW CUFFLESS BLOOD PRESSURE SENSOR AND OPTIMAL DESIGN VIA A 3D FLUID-SOLID-ELECTRIC FINITE ELEMENT MODEL
(2016)
Tu, Tse-Yi
;
Kao, Yung-Hua
;
Chao, Paul C. -P.
;
Samadhiya, Durgesh
;
電子工程學系及電子研究所
The effects of mis-positioning a newly-designed non-invasive, cuffless blood pressure sensor are thoroughly investigated via simulation and analysis on a 3D fluid-solid electric finite element model. A subsequent optimal design of this blood pressure is conducted based on the aforementioned mis-positioning effects. A highly-accurate, non-invasive, cuffless blood pressure (BP) sensor was successfully developed recently for an effective personal monitoring device on blood pressures. This new small-sized, portable blood pressure sensor is able to offer continuous BP measurements. The availability of continuous blood pressures are important for monitoring and evaluating personal cardiovascular systems. The sensor contains a strain-sensitive electrode encapsulated by flexible polymer. As the sensor placed on the position right on the top of the center of the wrist pulsation area, the deflection of the sensor induces the resistance changes of the electrode. By measuring the changes in electrode resistance, the level of pulsation is successfully quantified. Subsequent calculation based in this measurement can lead to fair estimates on blood pressures. However, as the sensor is placed on the wrist area where pulsation occurs, the mis-positioning of the sensor to the desired location, the center of the pulsation area, is inevitable. This study is dedicated to investigate the effects of the mis-positioning via a 3D finite element model. A new 3D fluid-solid-electro coupling interaction finite element model of the wrist is built for predicting the vibration of radial artery and then diastolic and systolic blood pressures. The FEM includes sensor of gel capsule and strain-sensing electrodes, radial artery, blood, radius bones, tendon, muscles and the front-end readout circuit. The FEM is the multi physics FEM with fluid, solid and electric. The section of wrist is constructed from magnetic resonance imaging (MRI) and the length of the FEM is 40mm. The complete 3D FEM model successfully simulated the vibration of skin surface and the sensor module. The diastolic and systolic blood pressures can be accurately predicted by the simulated output resistance. The pulsation levels due to varied mis-positionings are simulated by the built model, and simulation results are successfully validated by experiments. It is found that due to the unsymmetrical geometry of the wrist, the pulsation levels are also varied in an un-symmetric fashion with the mis-positionings in different directions. The maximum output of the BP sensor occurs when the sensor is placed 3 mm away from the center of the pulsation area, while the sensor output remain valid for subsequent signal processing as the sensor is placed within 5 mm from the pulsation center. Considering the inevitable mis-positionings by all possible users in different genders and ages, the sizes of the sensors are successfully optimized for satisfactory average signal quality over all possible users.
Estimating Blood Pressure via Artificial Neural Networks Based on Measured Photoplethysmography Waveforms
(2018-01-01)
Priyanka, K. N. G.
;
Chao, Paul C. -P.
;
Tu, Tse-Yi
;
Kao, Yung-Hua
;
Yeh, Ming-Hua
A new approach for estimating blood pressure from photoplethysmography (PPG) signals is developed using artificial neural networks (ANNs). Blood Pressure is one of the most important parameters that can provide valuable information of personal healthcare. A reflective photoplethysmography (PPG) sensor module is developed for the cuffless, non-invasive blood pressure (BP) measurement based on PPG at wrist on radial artery. Blood Pressure is in a relation with the pulse duration of the PPG. In this paper, we propose to estimate blood pressure from PPG signal by using artificial neural networks approach. This is the first reported study to consider varied temporal periods of PPG waveforms as features for application of artificial neural networks (ANNs) to estimate blood pressure. We compared our results with those measured using a commercial cuff-based digital blood pressure measuring device and obtained encouraging results of overall SBP and DBP regression (R) as 0.99115.
Fuzzy control design of a magnetically actuated optical image stabilizer with hysteresis compensation
(2009-04-01)
Tu, Tse-Yi
;
Chao, Paul C. -P.
;
Chiu, Chi-Wei
;
Wang, Chun-Chieh
;
Huang, Jeng-Shen
A fuzzy controller (FC) is designed for a magnetically actuated optical image stabilizer (OIS) in order to suppress the vibrations caused by hand shakings and hysteresis. To this end, the dynamic model of the OIS with consideration of hysteresis is first established, along with assuming the hand-shaking vibration as sinusoidal excitations. It is clearly shown that with capability of continuing parameter tuning, the FC is superior to the conventional PID for vibration suppression. (C) 2009 American Institute of Physics. [DOI: 10.1063/1.3068542]
Investigation of post-annealing indium tin oxide for future electro-optical device application
(2011)
Ho, Ching-Yuan
;
Tu, Tse-Yi
;
Wang, Chun-Chieh
;
Kang, Yuan
;
電控工程研究所
The nanostructure transformation associated with electro-optical properties via post-annealing of indium tin oxide film (ITO) is investigated by increasing post-annealing temperature in ambient oxygen. Although oxygen vacancy and activation Sn ions contribute to conductivity of ITO film, the oxygen vacancy inevitably reduces during posting annealing, but Sn-O related bonds are oppositely increased with IR absorption at 790 cm(-1). Moreover, the sheet resistance of as-deposited ITO film 8.6 Omega/sq increases to 47 Omega/sq as the annealing temperature rises to 500 degrees C., the photoluminescence (PL) spectrum indicates that the oxygen vacancy plays a key role in dominating R-s in comparison with Sn-O bonds. The blue light transmittance of ITO film is slightly proportional to the annealing temperature due to phase crystallization enhancing band gap narrowing. Furthermore, as the annealing temperature rises beyond 500 degrees C, the transmittance is compromised between the Burstein-Moss effect of high carrier concentration and nanostructure crystallinity. The oxygen deficient vacancy, instead of Sn-O related bonds, is a major contribution for ITO conductivity and transmittance
Machine Learning Classification for Assessing the Degree of Stenosis and Blood Flow Volume at Arteriovenous Fistulas of Hemodialysis Patients Using a New Photoplethysmography Sensor Device
(2019-08-01)
Chiang, Pei-Yu
;
Chao, Paul C. -P.
;
Tu, Tse-Yi
;
Kao, Yung-Hua
;
Yang, Chih-Yu
The classifier of support vector machine (SVM) learning for assessing the quality of arteriovenous fistulae (AVFs) in hemodialysis (HD) patients using a new photoplethysmography (PPG) sensor device is presented in this work. In clinical practice, there are two important indices for assessing the quality of AVF: the blood flow volume (BFV) and the degree of stenosis (DOS). In hospitals, the BFV and DOS of AVFs are nowadays assessed using an ultrasound Doppler machine, which is bulky, expensive, hard to use, and time consuming. In this study, a newly-developed PPG sensor device was utilized to provide patients and doctors with an inexpensive and small-sized solution for ubiquitous AVF assessment. The readout in this sensor was custom-designed to increase the signal-to-noise ratio (SNR) and reduce the environment interference via maximizing successfully the full dynamic range of measured PPG entering an analog-digital converter (ADC) and effective filtering techniques. With quality PPG measurements obtained, machine learning classifiers including SVM were adopted to assess AVF quality, where the input features are determined based on optical Beer-Lambert's law and hemodynamic model, to ensure all the necessary features are considered. Finally, the clinical experiment results showed that the proposed PPG sensor device successfully achieved an accuracy of 87.84% based on SVM analysis in assessing DOS at AVF, while an accuracy of 88.61% was achieved for assessing BFV at AVF.
OPTIMAL DESIGN AND EXPERIMENTAL VALIDATION OF A NEW NO-CUFF BLOOD PRESSURE SENSOR BASED ON A NEW FINITE ELEMENT MODEL
(2014)
Tu, Tse-Yi
;
Chao, Paul C. -P.
;
Lee, Yung-Pin
;
Kao, Yung-Hua
;
電子工程學系及電子研究所
A new multi-physics finite element model (FEM) on the vibration of the radial artery on the wrist is built by this study to predict vibration of wrist artery vessel vibration and then diastolic and systolic blood pressures. The FEM includes the sensor of gel capsule and strain-sensing electrodes, skin, bones and muscles. The vibrations of skin surface and the sensor module are successfully simulated by the established FEM. The resulted vibratory deformation of the sensor electrodes are further transformed to resistance variations to mimic realistic electronics of the front-end readout circuit via establishing a cross-discipline sub-FEM model. The established FEM can is particularly customized to varied ages, weights, heights, genders, and special cardiovascular diseases of users. The customization can be performed in a fashion of one-time calibration by medical staff and/or medical staff. With the customized FEM in high accuracy level, the diastolic and systolic pressures can be accurately predicted by the simulated output resistance variation. Based on simulation FEMs, the design of the pulse sensor is successfully optimized via the processes of Taguchi and numerical methods. The measurements are also conducted to a dozen of subjects. It shows that the designed novel blood pressure sensor is capable of sensing the blood pressure to require accuracy level of the error less than 10% as compared to commercial counterparts with a cumbersome cuff.