Oung Qi Wei
Thumbnail Image
Preferred name
Oung Qi Wei
Official Name
Oung, Qi Wei
Alternative Name
Oung, Qi Wei
Oung, Q. W.
Wei, Oung Qi
Main Affiliation
Scopus Author ID
56405592700
Researcher ID
J-3470-2015

Research Output

Filters

2
2
1 1
Reset filters

Settings
Now showing 1 - 2 of 2
  • Publication
    Evaluation of multimodal sensors for classification of parkinson's disease with severity levels
    ( 2021)
    Oung Qi Wei
    Parkinson’s disease (PD) is a progressive neurodegenerative disorder that has affected a large part of the population. Several symptoms of PD include tremor, rigidity, slowness of movements and vocal impairments. Recently, the evolution of numerous tools for monitoring PD has stimulated the interest of researchers, leading to the possibility of utilising wearable technology to assist in monitoring patient with Parkinson (PWP) state and generate valuable clinical feedback. There have been various initiatives from researchers for PD detection, which includes using (1) Non-signal data acquisition protocols i.e. patients’ diaries, self-reports, clinical rating scales and imaging modalities or through (2) Signal analysis techniques i.e. Bioelectrical signals (EMG signals and EEG signals), motion data signals via wearable motion sensors and speech signals via audio sensor. Among all the initiatives, detecting PD progression using wearable motion sensors and audio sensor has raised significant attention by researchers due to its enormous potentials. However, current signal processing and classification algorithms have limited capability and performance especially involving information obtained from wearable motion sensors for classification of PWP with different severity levels. Additionally, even though there were many works for distinguishing PWP and healthy controls via audio sensor, there are limited researches focusing on multiclass classification of PD. Most of the previous works were conducted based on a binary classification, resulting in early PD stage and the advanced ones being treated equally. Therefore, in this research, a recognition system based on wearable motion sensors and audio sensor information as biomarkers for initial PD detection with the ability for multiclass classification was developed. Three classes of PD severity levels were considered – ‘mild’, ‘moderate’ and ‘severe’ from healthy controls to give the classification results more accurate for clinical use. First, the signals from wearable motion sensors and audio sensor were decomposed by the proposed algorithms known as empirical wavelet transform (EWT) and empirical wavelet packet transform (EWPT) respectively. Then, several features were extracted from the coefficients of all decomposed signals. The performance of the algorithm was validated using three classifiers – K-nearest neighbour (KNN), probabilistic neural network (PNN) and extreme learning machine (ELM). Validation of the proposed algorithms demonstrated the effectiveness of wavelet energy, entropy-based and Higher Order Spectra (HOS) - based features from the EWT/ EWPT decomposition for multiclass PD severity levels classification with the highest average accuracy of 96.95% from ELM RBF kernel classifier. Finally, a non-invasive screening system using the proposed algorithm for detection of PD severity levels was also developed using Matlab Graphical User Interface (GUI) Toolbox. It is hoped, the GUI system eases medical professionals to monitor the progression of PD.
  • Publication
    Understanding Domain Knowledge in Initialization Method for K-Mean Clustering Algorithm in Medical Images
    ( 2022-01-01)
    Tan X.J.
    ;
    Nazahah Mustafa
    ;
    Mohd Yusoff Mashor
    ;
    Ab Rahman K.S.
    ;
    Wan Zuki Azman Wan Muhamad
    ;
    Leow Wai Zhe
    ;
    Cheor Wai Loon
    ;
    Oung Qi Wei
    This work serves as a preliminary study to investigate and identify the applicability of domain knowledge as an initialization method for K-Mean (KM), typically in medical images. For this purpose, 20 breast histopathology images were used as data set and the evaluations are focused on the clustering of the hyperchromatic nucleus. The iteration numbers and clustering results (i.e., accuracy, over-segmentation, and under-segmentation) are benchmarked with KM++ and the conventional random initialization method. The domain knowledge initialization method is found promising by achieving lower iteration numbers (<9), higher percentage in accuracy (85.5% (±2.27)), and lower percentages in over-segmentation (8.25% (±2.23)), and under-segmentation (7.00% (±2.14)). From this study, we hypothesize that the domain knowledge initialization method has the potential to be implemented as an initialization method and is posited to overperform some established initialization methods, typically for clustering tasks in medical images.
      1