Xiaolong Li

Prof. Dr. Xiaolong Li


Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

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

Representative Publications

View All Publications
1 July 2015 in Trends in Biotechnology
Verified
Journal Article
Closed access
XLXiaolong Li
Xiaolong Li
WSWouter A. Serdijn
Wouter A. Serdijn
WZWei Zheng
Wei Zheng
YTYubo Tian
Yubo Tian
BZBing Zhang
Bing Zhang

16 June 2024 in Electronics
Verified
Journal Article
Open access
PDF file
WZWeicheng Zhou
Weicheng Zhou
Wei ZhengWei Zheng
Wei Zheng
YFYoubing Feng
Youbing Feng
XLXiaolong Li
Xiaolong Li
Neonatal epilepsy is an early postnatal brain disorder, and automatic seizure detection is crucial for timely diagnosis and treatment to reduce potential brain damage. This work proposes a novel Lightweight Multi-Attention Network, LMA-EEGNet, for diagnosing neonatal epileptic seizures from multi-channel EEG signals employing dilated depthwise separable convolution (DDS Conv) for feature extraction and using pointwise convolution followed by global average pooling for classification. The proposed approach substantially reduces the model size, number of parameters, and computational complexity, which are crucial for real-time detection and clinical diagnosis of neonatal epileptic seizures. LMA-EEGNet integrates temporal and spectral features through distinct temporal and spectral branches. The temporal branch uses DDS Conv to extract temporal features, enhanced by a channel attention mechanism. The spectral branch utilizes similar convolutions alongside a spatial attention mechanism to highlight key frequency components. Outputs from both branches are merged and processed through a pointwise convolution layer and a global average pooling layer for efficient neonatal seizure detection. Experimental results show that our model, with only 2471 parameters and a size of 23 KB, achieves an accuracy of 95.71% and an AUC of 0.9862, demonstrating its potential for practical deployment. This study provides an effective deep learning solution for the early detection of neonatal epileptic seizures, improving diagnostic accuracy and timeliness.

27 September 2021 in International Journal of Antennas and Propagation
Verified
Journal Article
Open access
PDF file
XLXiaolong Li
Xiaolong Li
YXYi Xing
Yi Xing
ZZZhenkai Zhang
Zhenkai Zhang
Target localization plays an important role in the application of radar, sonar, and wireless sensor networks. In order to improve the localization performance using only two stations, a hybrid localization method based on angle of arrival (AOA) and time difference of arrival (TDOA) measurements is proposed in this paper. Firstly, the optimization model for localization based on AOA and TDOA are built, respectively, in the sensor network. Secondly,the majorization-minimization (MM) method is employed to create surrogate functions for solving the multiple objective optimization problem. Next, the hybrid localization problem is solved by the projected gradient decent (PGD) method. Finally, the Cramer–Rao lower bound (CRLB) for the joint AOA and TDOA method is derived for the comparison. Simulations proved that the proposed method has improved localization performance using AOA and TDOA measurements from only two base stations.

7 September 2022 in IEEE Circuits and Systems Magazine
Verified
Journal Article
Closed access
XLXiaolong Li
Xiaolong Li
MYMing Yu
Ming Yu
JZJin Zhu
Jin Zhu
WZWei Zheng
Wei Zheng
GMGuojun Ma
Guojun Ma
WSWouter A. Serdijn
Wouter A. Serdijn
Closed-loop vagus nerve stimulators (VNSs) are used to treat refractory epilepsy (RE) and provide an alternative treatment modality with fewer side effects for RE. This work proposes a closed-loop VNS that senses the scalp EEG via a wearable EEG recorder (WER) and stimulates the vagus nerve via an implantable VNS (IVNS), the combination of which is referred to as a “HybridVNS”. First, the architecture and system control (by an App) of the HybridVNS are presented. Next, discrete component implementations of its key building blocks including the multichannel preamplifier, pulse generator and wireless power link are provided. Further, a 4-channel preamplifier magnifies the recorded EEG signals, and each channel consists of a capacitively coupled instrumentation amplifier (CCIA) cascaded by a 2nd-order low pass filter. The measured gain and the input-referred noise (IRN) of the preamplifier are 45.8 dB and 4.9 ${\mu}$ V rms (1 Hz–200 Hz), respectively. We propose a switched-resistor array pulse generator for constant-current stimulation, and a novel distance-frequency adaptive inductive coupling link (DFAICL) to automatically tune the resonance frequency. A maximum wireless power transmission efficiency of 17% is achieved in the DFAICL when the supply power is 30dBm and the distance is 20 mm at 2 MHz. Finally, we demonstrate that in the WER, the onset of an epileptic seizure can be detected by a threshold of coastline length (294 ${\mu}$ V) combined with a threshold of slope (7.3 ${\mu}$ V/ms) within a unit time segment (780 ms), which provides an average accuracy of 88.2%. The feasibility of the IVNS was tested in an acute rodent epilepsy experiment. However, the closed-loop experiment was not conducted.

23 February 2019 in Micromachines
Verified
Journal Article
Open access
PDF file
CMCheng Ma
Cheng Ma
JZJin Zhu
Jin Zhu
XLXiaolong Li
Xiaolong Li
Wei ZhengWei Zheng
Wei Zheng
Quartz crystal microbalance (QCM) is still a new high-precision surface detection technique. However, the adsorption quality detected by the QCM currently contains a solvent-coupling quality and cannot separate the actual biomolecular mass. Local surface plasmon resonance (LSPR) can detect the mass of biomolecules, but requires a certain contrast between the solvent of the surrounding medium and the refractive index of the adsorbed layer. The sensor chip, combining two compatible technologies, can realize the simultaneous detection of biomolecules and improve the refractive index sensitivity. The structure of our chip is to prepare the ring-shaped gold electrode on the upper surface of the quartz crystal, the circular gold electrode on the bottom surface, and the spherical gold nanoparticles arrays in the center region of the ring electrode to form a QCM/LSPR dual-technology chip. Through simulation, we finally get the size of the best energy trap by the two electrodes on the upper surface and the lower surface: the ring-top electrode with a thickness of 100 nm, an inner diameter of 4 mm, and an outer diameter of 8 mm; and the bottom electrode with a thickness of 100 nm and a radius of 6 mm. By comparing the refractive index sensitivity, we chose a spherical gold nanoparticle with a radius of 30 nm and a refractive sensitivity of 61.34 nm/RIU to design the LSPR sensor chip.

19 November 2024 in Atmosphere
Verified
Journal Article
Open access
PDF file
JZJinbo Zhang
Jinbo Zhang
JNJiawei Niu
Jiawei Niu
ZXZhibin Xie
Zhibin Xie
YWYajun Wang
Yajun Wang
XLXiaolong Li
Xiaolong Li
QZQilin Zhang
Qilin Zhang
Sprite halos are transient luminous phenomena in the lower ionosphere triggered by tropospheric lightning. The effect of removed charge distributions on sprite halos has not been sufficiently discussed. A three-dimensional (3D) quasi-electrostatic (QES) field model was developed in this paper, including the ionospheric nonlinear effect and optical emissions. Simulation results show that, for a total charge of 150 C removed within 1 ms with different spatial distributions, higher altitudes of charge removal lead to stronger electric fields and increase sprite halos’ emission intensities. The non-axisymmetric horizontal distribution of charge affects mesospheric electric fields, and the corresponding scales and intensities of emissions vary with observation orientations. Considering the tilted dipole charge structure due to wind shear, the generated electric field and the corresponding position of sprite halos shift accordingly with the tropospheric removed charge, providing an explanation for the horizontal displacement between sprite halos and the parent lightning.