Medical Image Analysis
"Brain Analysis", Biomedical Signal and Image Computing Laboratory (BiSICL), The University of British Columbia (UBC), Advisor: Dr. Rafeef Abugharbieh, (09/2013 - present)
"Medical Image Managing and Analysis System", Southeast University, 09/2010-11/2010
- Multimodal Human Brain Connectivity Analysis based on Graph Theory
- Brain Connectivity Analysis and Activation Detection based on functional and structural MRI
"Medical Image Managing and Analysis System", Southeast University, 09/2010-11/2010
Manage and analyze CT source Medical Images, including image processing functions:
- Linear transformation
- Pseudo color display
- Best window width and level display
- Histogram equalization
- Filtering and denoising
- Segmentation, etc.
Machine Learning
"Emotion Recognition based on the HRV Signals and Facial Expressions", SEU, Master’s thesis, Advisor: Dr. Feng Wang, 09/2011 - 06/2013
Award Received: Best grade in the class.
- Improved the Haar-like adaboosting cascades frame to better performance Human Face Detection, further realized fiducial points automatic detection and Gabor feature extraction.
- Proposed Supervised Divergence Linear Laplacian Eigenmap method to perform effective feature selection based on Manifold learning.
- Proposed a classifier based on synergetic prototype and similarity calculation has been to improve the recognition accuracy from 85.71% to 93.29%.
Award Received: Best grade in the class.
- Improved Human Facial Direction Recognition with higher accuracy of 90% to the selected database.
- Evaluation between Artificial Neural Networks, Ensemble Learning Adaboost, Genetic Algorithm and Support Vector Machine.
- Proposed an automated, patient-specific framework for Seizure Onset Detection.
- Feature selection from the temporal, spectral, entropy and spatial distribution components of EEG signal.
- Evaluated the accuracy, specificity and latency of two machine learning method, Support Vector Machine and Random Forest.
Signal Processing and Analysis
"Emotion Recognition experiment and analysis based on the Physiological Signals", SEU, Bachelor’s thesis, Advisor: Dr. Feng Wang, 12/2009 - 06/2010
- Designed 26 emotion inducing experiments using movie clips based on ECG and EEG signal.
- Proposed an adaptive wavelet-based de-noising algorithm improving SNR.
- Developed a linear regression model based on a weighted fuzzy feature selection method, increasing the accuracy from 71% to 80%.
"Remote ECG Monitoring System based on Non-contact Electrode", SEU, 01/2009-11/2009
Led a team of 8 members in "Challenge Cup" National Undergraduate academic contest.
Led a team of 8 members in "Challenge Cup" National Undergraduate academic contest.
- Created a novel sensor based on non-contact electrode, enables long-time monitoring free from issues like skin allergies to solve the heart attack detection suffered by 720,000 Americans per year.
- Developed a GUI platform incorporated with new de-noising method and automatic computer-aid diagnosis.
- Outstanding Award in Jiangsu Province and Second Prize in National level contest, whose project demonstration was broadcast in CCTV-1 channel.
"The research on Photoelectric pulse wave wireless measurement system", SEU, 05/2008 - 04/2010, Led a team of 3 members.
- Designed the PPG (Photoplethysmogram) Photoelectric pulse wave sensor, shaped and integrated the circuit and hardware using chip opt101 of TI (Texas Instruments).
- Programmed the microprocessor CC2430 for Data collection, management and wireless data transmission (Zigbee).
- Developed a GUI platform for data visualization and analysis using MFC and Matlab.
"The design of portable wireless ECG monitoring terminal", SEU, 06/2010 - 12/2010.
Signal Analysis System for Myocardial Infarction ECG, SEU, course project, 01/2009-04/2009
Voice recognition system based on wavelets, SEU, course project, 09/2008-12/2008
- Brainstormed, experimented and tested on the new textile electrode for ECG sensor with more comfort and higher signal quality.
- Algorithm exploration and ECG signal preprocessing and feature selection using Matlab.
- Designed & implemented the real-time ECG processing and Heart Rate calculation in MSP430 microcomputer using C language.
Signal Analysis System for Myocardial Infarction ECG, SEU, course project, 01/2009-04/2009
- Preprocess the raw ECG data.
- Detect QRS complex waves and P, T pulse.
- Disease related parameter calculation.
Voice recognition system based on wavelets, SEU, course project, 09/2008-12/2008
- Normalization and denoising of the audio signals.
- Automatic removal of the word interval and signal smoothing.
- Features extraction of area, amplitude and average energy, followed by password recognition using similarity matching function.
Computer Graphics
Three Dimensional Data Rendering using Ray-Tracing Algorithm, course project, 11/2008-12/2008
Implementation of the illumination model, reflection and shadow effect and the texture mapping.
Implementation of the illumination model, reflection and shadow effect and the texture mapping.