Xiaodi Ding

Dr. Xiaodi Ding


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
19 November 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JLJinghui Li
Jinghui Li
LKLi Ke
Li Ke
QDQiang Du
Qiang Du
XDXiaodi Ding
Xiaodi Ding
XCXiangmin Chen
Xiangmin Chen
Because a cardiac function signal cannot reflect cardiac health in all directions, we propose a classification method using ECG and PCG signals based on BiLSTM-GoogLeNet-DS. The electrocardiogram (ECG) and phonocardiogram (PCG) signals used as research objects were collected synchronously. Firstly, the obtained ECG and PCG signals were filtered, and then the ECG and PCG signals were fused and classified by using a bi-directional long short-term memory network (BiLSTM). After that, the time-frequency processing was performed on the filtered ECG and PCG signals to obtain the time-frequency diagram of each signal; the one-dimensional signal was changed into a two-dimensional image signal, and the size of each image signal was adjusted to input the improved GoogLeNet network for classification. Then we obtained the two-channel classification results. The three-channel classification results, combined with the classification results of the above BiLSTM network, were finally obtained. The classification results of these three channels were used to make a decision via the fusion strategy of the improved D-S theory. Finally, we obtained the classification results. Taking 70% of the signals in the database as training data and 30% as test data, the obtained classification accuracy was 96.13%, the sensitivity was 98.48%, the specificity was 90.8%, and the F1 score was 97.24%. From the experimental results, the method proposed in this paper obtained high classification accuracy, and the classification effect was better than a cardiac function signal, which makes up for the low accuracy of the cardiac function signal for judging cardiac disease.

2 December 2019 in BioMedical Engineering OnLine
Verified
Journal Article
Open access
PDF file
JCjia chen
jia chen
LKLi Ke
Li Ke
QDQiang Du
Qiang Du
WZWanni Zu
Wanni Zu
XDXiaodi Ding
Xiaodi Ding
Background: Magnetic induction tomography (MIT) is a tomographic imaging technique, which has potential applications in security, industry, and medicine. Typically, sensors form a closed structure around the object. However, the measurement cannot be achieved using a closed sensor array in the process of severe brain trauma nursing and the neurosurgery operation. Results: The new sector sensor array magnetic induction tomography (SMIT) system is developed to realize real-time monitoring in the treatment of the brain. The functions of the drive coil and the sensor coil are separated in this system. The detection sensitivity of the imaging region boundary is analyzed through simulation. The sensor array locates on the high detection-sensitivity area, and the low sensitivity detection area is reserved for operation and clinical equipment. The sensor array received the energy of the signal accounts for reach 90% of the total energy. The integrity measuring data are obtained using a rotating scan in the system. In the experiment, we analyze the effects that system parameters have on the quality of imaging, for example, the scan step size, the number of sensors, the coverage angle of the sensor array and the scan angle. The experiment result provides a reference for the SMIT system design under a particular condition. In the complete measurement, the SMIT system reconstructs the images of center goal and margin goal, and the actual images have high peak signal-to-noise ratio. Conclusions: The SMIT system can rebuild the conductivity distribution of the imaging region using incomplete space. In rotation measurement, the system provides a working place for clinical care. The flexible design of the system based on the experiment result makes the different treatment for brain injury own matched SMIT equipment.

11 July 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
XDXiaodi Ding
Xiaodi Ding
AWAn Wang
An Wang
YYYizhao Yuan
Yizhao Yuan
Hypertension is a major risk factor for primary stroke; however, reliable noninvasive physiological biomarkers for early stroke warning in hypertensive patients remain elusive. To characterize the dynamic evolution of cerebral arterial blood conductivity—from the earliest hypertensive stages through to imminent stroke onset—we developed four physiologically grounded numerical models: normotension, borderline hypertension, established hypertension, and pre-stroke hypertension. These models were constructed based on empirically validated correlations between in vivo cerebral arterial blood conductivity and key hemodynamic parameters. All input physiological data were rigorously sourced from peer-reviewed clinical studies. Each model was coupled with standardized geometries constructed using population-averaged templates. Using the finite element method, we simulated blood conductivity across a full cardiac cycle under realistic in vivo physiological conditions. Crucially, we performed quantitative decomposition analysis to isolate and assess the individual contributions of specific physiological parameter changes—such as systolic pressure, dynamic viscosity of blood, hematocrit, and blood flow velocity—to the observed shifts in conductivity during hypertensive progression. Results revealed a biphasic, stage-dependent evolutionary pattern: conductivity initially rises slightly in borderline hypertension, then declines markedly as hypertension advances toward the pre-stroke stage. Relative to the normotensive baseline, conductivity in the pre-stroke model decreased by 17.61%—a reduction strongly associated with hypertension-induced reductions in cerebral blood flow velocity and concurrent increases in hematocrit. Comprehensive mesh independence and time-step sensitivity analyses confirmed the numerical robustness and computational reliability of all simulations. Collectively, these findings demonstrate that cerebral arterial blood conductivity undergoes quantifiable, pre-symptomatic deviation prior to stroke onset, establishing it as mechanistically grounded. The research findings demonstrate substantial theoretical significance and translational potential, grounded in a rigorously validated modeling framework and serving as a robust basis for generating and refining novel scientific hypotheses.