Tianxia Zhao

Tianxia Zhao

Auto-generated profile, based on publicly available information

Tianxia Zhao

The Key Laboratory of Integrated Microsystems, Peking University Shenzhen Graduate School, Shenzhen 518055, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2021 in Wireless Communications and Mobile Computing
Verified
Journal Article
Open access
PDF file
CQChangpei Qiu
Changpei Qiu
XWXin’an Wang
Xin’an Wang
TZTianxia Zhao
Tianxia Zhao
QLQiuping Li
Qiuping Li
BWBo Wang
Bo Wang
HWHu Wang
Hu Wang
In this paper, an FPGA‐based convolutional neural network coprocessor is proposed. The coprocessor has a 1D convolutional computation unit PE in row stationary (RS) streaming mode and a 3D convolutional computation unit PE chain in pulsating array structure. The coprocessor can flexibly control the number of PE array openings according to the number of output channels of the convolutional layer. In this paper, we design a storage system with multilevel cache, and the global cache uses multiple broadcasts to distribute data to local caches and propose an image segmentation method that is compatible with the hardware architecture. The proposed coprocessor implements the convolutional and pooling layers of the VGG16 neural network model, in which the activation value, weight value, and bias value are quantized using 16‐bit fixed‐point quantization, with a peak computational performance of 316.0 GOP/s and an average computational performance of 62.54 GOP/s at a clock frequency of 200 MHz and a power consumption of about 9.25 W.

3 March 2021 in Journal of Healthcare Engineering
Verified
Journal Article
Open access
PDF file
QLQiuping Li
Qiuping Li
TZTianxia Zhao
Tianxia Zhao
XWXin’an Wang
Xin’an Wang
CQChangpei Qiu
Changpei Qiu
BZBing Zhou
Bing Zhou
BWBo Wang
Bo Wang
In recent years, Traditional Chinese Medicine (TCM) has attracted more and more attention due to its good therapeutic effect, low cost, and convenience. This research is also a part of the goal of the modernization of TCM. Based on the meridian electric potential acquisition system independently developed by our project team, in this paper, we designed the human body's meridian electric potential acquisition scheme. We use principal component analysis (PCA) to prove that the meridional potential signal is derived from the ECG signal. Then, Inception ResNet V2 was used to classify acupoints and nonacupoints. Finally, the classification accuracy rate reached 86.59045265, and the F1 score = 0.72161642. This shows that acupoints and nonacupoints can be distinguished by their surface potential.

6 January 2021 in Journal of Healthcare Engineering
Verified
Journal Article
Open access
PDF file
CQChangpei Qiu
Changpei Qiu
TZTianxia Zhao
Tianxia Zhao
QLQiuping Li
Qiuping Li
XWXin’an Wang
Xin’an Wang
KXKanglin Xiao
Kanglin Xiao
BWBo Wang
Bo Wang
A low-power stable wideband current source for acupuncture point skin impedance measurements has been designed employing a differential architecture and negative feedback. The circuits extend bandwidth to 1MHz, reducing harmonic distortion to 0.24 at 1MHz. The output impedance is 37M at 100kHz and 11M at 1MHz. The stability of the output current of the current source when connected to different loads is below 0.1 at frequencies up to 500kHz and increases to 0.74 at 1MHz. The circuit was manufactured in a 0.13-μm CMOS technology and measured results are presented. The area of the current source is 0.09mm2 and its consumption is 1.2mW. It is intended for low-power acupuncture point skin impedance measurements.

18 January 2022 in Journal of Healthcare Engineering
Verified
Journal Article
Open access
PDF file
TZTianxia Zhao
Tianxia Zhao
XWXin’an Wang
Xin’an Wang
CQChangpei Qiu
Changpei Qiu
This study introduces a method to classify single-lead ECG signals by extracting features through traditional methods and deep neural network methods. At first step, the statistical type features of the ECG signals are exacted by traditional methods, including time domain features, frequency domain features, and medical domain features. And then, deep neural networks are used to extract the deeper features of the ECG signal. The database of ECG signals is from Cinc 17, which have 8528 samples of short-time ECG signal. The huge amount of data makes the classification and identification more accurate by atrial fibrillation, normal sinus rhythm, noise, and indiscernible. Compare the base model built by the classified data and the data collected by the ECG device of CareON to enable daily early screening and a remote alert function with WeChat app. This method can extend the prevention, detection, and diagnosis of heart disease to the family, company, and other out-of-hospital scenarios, thus enabling faster treatment of heart patients and saving medical resources.

25 April 2021 in Journal of Healthcare Engineering
Verified
Journal Article
Open access
PDF file
QLQiuping Li
Qiuping Li
XZXing Zhang
Xing Zhang
XWXin’an Wang
Xin’an Wang
TZTianxia Zhao
Tianxia Zhao
CQChangpei Qiu
Changpei Qiu
BZBing Zhou
Bing Zhou
As traditional Chinese medicine (TCM) has gained more and more recognition in the world, Chinese medicine has also played its important role. However, traditional Chinese medicine equipment is relatively deficient, with insufficient functions and low degree of digitalization. For example, existing auscultation equipment can obtain few human characteristic indicators, which is difficult to meet the needs of reference in traditional Chinese medicine diagnosis. Based on this, this paper designed a human body characteristic index detection system based on the principle of traditional Chinese medicine, which includes respiratory and heartbeat signal acquisition device, meridian and acupoint signal acquisition device, temperature signal acquisition device, pulse and blood pressure acquisition device, processing module, keyword module, and output module. The respiratory and heartbeat signal acquisition device is used to collect the respiratory and heartbeat signal of human body. Meridian acupoint signal acquisition device is used to collect human meridian acupoint radio signals. The temperature signal acquisition device is used to collect the infrared temperature light wave signal of human body. Pulse and blood pressure acquisition devices are used to collect pulse and blood pressure signals. The processing module is used to obtain one or more human body characteristic indicators according to one or more of the respiration and heartbeat signals, meridians and acupoints signals, temperature signals, pulse, and blood pressure, including Qi and blood characteristic indicators, viscera and six meridian characteristic indicators, and temperature characteristic indicators. The keyword corresponding module is used to obtain the corresponding keyword representing the physiological state information of human body according to the one or more human body characteristic indicators. The output module is used to output the human body characteristic index and the key words. It includes the key words of Qi and blood state information, the key words of viscera state information, the key words of Qi and blood state information, etc. The system can be used for serious disease screening, chronic disease management, and risk early warning.

22 April 2021
Verified
Conference Paper
Open access
CQChangpei Qiu
Changpei Qiu
XWXin'AN Wang
Xin'AN Wang
XZXing Zhang
Xing Zhang
Bo WangBo Wang
Bo Wang
TZTianxia Zhao
Tianxia Zhao
QLQiuping Li
Qiuping Li
In this paper, a high precision, high linearity 10 bit Digital-to-Time Converter (DTC) circuit was designed and implemented using a conventional open loop amplifier comparator and voltage reference source. The proposed Digital-to-Time Converter circuit (DTC) uses a large capacitor as the load to implement the "fixed-slope method" and controls the capacitor array with a "segmented coding method" to improve the nonlinearity of the DTC. The DTC mentioned has been implemented in standard 130nm CMOS. The DTC designed in this paper has high precision and linearity. The accuracy of the DTC designed in this paper is 315fs. The integral nonlinearity is below 0.52LSB, and the power consumption is 26.4μW.