Kim Mey Chew

Facial Expression Monitoring System Using PCA-Bayes Classifier

In order to endow a machine with an emotional intelligence is a challenging research issue and become one that has been of growing importance to those working in human-computer interaction. This study presents the framework of a special session to study and investigate the best techniques for emotion recognition, validation and analysis of expressivity in human-computer interaction, based on the common physiological background. A PCA-Bayes classifier (PCABC) was proposed in this study for facial recognition problem. The session is primarily concerned with visual emotion analysis; the analysis of physiological signals serves as a complement to this modality. Signal is taken from a different aspect of the physiology and visual. The signal will go through a process of elimination votes in order to extract better signal features. It is shown that the PCABC can perform much better than Least Mean Square (LMS) classifier. Psychological backgrounds will be studied to obtain good signal.

Reflection coefficient detection of simulation models for microwave imaging simulation system.

The study was conducted based on two objectives as framework. The first objective is to determine the point of microwave signal reflection while penetrating into the simulation models and, the second objective is to analyze the reflection pattern when the signal penetrate into the layers with different relative permittivity, εr. Thus, several microwave models were developed to make a close proximity of the in vivo human brain. The study proposed two different layers on two different characteristics models. The radii on the second layer and the corresponding antenna positions are the factors for both models. The radii for model 1 is 60 mm with an antenna position of 10 mm away, in contrast, model 2 is 10 mm larger in size with a closely adapted antenna without any gap. The layers of the models were developed with different combination of materials such as Oil, Sandy Soil, Brain, Glycerin and Water. Results show the combination of Glycerin + Brain and Brain + Sandy Soil are the best proximity of the in vivo human brain grey and white matter. The results could benefit subsequent studies for further enhancement and development of the models.

Development and measurement properties of prosthetics users' survey

The need to measure and evaluate prosthetics practice has received growing recognition in the past several years. Reliable and valid self-report instruments are needed that can help facilities evaluate patient outcomes. Evidence of construct validity is provided by hierarchies of item difficulty that are consistent with clinical experience. A survey with 16 completed survey returned from hospital and rehabilitation centre. The instruments demonstrate adequate internal consistency of mean 4.0 for satisfaction with services, mean 3.81 for satisfaction with devices and 16.67% for defects or drawbacks of devices. Documenting mechanical stability of the prosthesis is inadequate when reporting follow-up studies and a questionnaires assessing patient satisfaction should be added to provide a better picture of the outcome and results.

Prosthetics: Health quality of life effects of limb loss

This study investigates and compares clinical outcome and patient health related quality of life in Malaysia who still undergoing routine prosthetic care. It was developed to access the quality of life that can be used in hospitals and clinics. It aims to establish how widespread the quality of life effects of limb loss. Reliability and validity are needed that can help facilitate patient outcomes. A survey with 16 completed surveys was returned from rehabilitation centre and hospital. The instruments demonstrate adequate internal consistency of average mean 3.44 for health related quality of life. Two overseas sample data from America and India were then made comparison with the existing data to provide a better picture of the assessment of patient satisfaction with the devices. The impact that limb loss can have on patients and their lives should not be underestimated. In this study it affected 64.4% of the respondents.

Human Brain Phantom Modeling: Concentration and Temperature Effects on Relative Permittivity

This paper discusses on selecting the most appropriate material for developing human-like brain phantom. It aims to investigate the effect of concentration and temperature to relative permittivity of a sample. This phantom was developed as a human-like brain for tumour detection using microwave signal. Result shows that plant gelatine is a stable and appropriate material for building phantom than agar-agar. A few models were developed based on the proposed ratio of mixture.

Human Body and Body Part Movement Analysis Using Gyroscope, Accelerometer and Compass

This study investigates and acts as a trial clinical outcome for human motion and behaviour analysis in consensus of health related quality of life in Malaysia. It was developed to analyse and access the quality of human limbs motion that can be used in hospitals, clinics and human motion researches. An experiment was set up in a laboratory environment with conjunction of analysing human motion and its behaviour. The instruments demonstrate adequate internal consistency of results as below: 1. Compass sensor gives a better result with less standard deviation values especially in x-axis according descriptive statistical data. 2. Compass sensor gives a clearer scatter plot for better classification. 3. R2 (amount of variation explained) for sensor attached on arm is lower than hip and that means data collected from this site have a consistent trend. A

Human Hand Movement Analysis Using Principle Component Analysis Classifier

This study investigates and acts as a trial clinical outcome for human motion and behavior analysis in order to investigate human arm movement during jogging and walking. It was developed to analyze and access the quality of human motion that can be used in hospitals, clinics and human motion researches. It aims to establish how widespread the movement and motion of arm will bring to effect of human in life. An experiment was set up in a laboratory environment with conjunction of analyzing human motion and its behavior. The instruments demonstrate adequate internal consistency of optimum scatter plot in gyroscope and accelerometer for pattern classification. PCA used in this study was successfully differentiate and classify

Signal Processing of Microwave Imaging Brain Tumor Detection Using Superposition Windowing

This paper discusses the selection of window function for signal processing in microwave imaging brain tumor detection. Most of the window functions are non-negative bell-shaped curves. This paper proposed a superposition windowing function for better time series data analyses and enhancement. The performance of the selected five window functions (Hamming, Blackman-Harris, Parzen, Chebyshev and Bartlett-Hanning) and the proposed superposition window were compared and evaluated. The results show the superposition window function is potentially reduce the unwanted noise and preserve important information of the signals.

Comparison of Human Jogging and Walking Patterns Using Statistical Tabular, Scatter Distribution and Artificial Classifier

This study investigates and acts as a trial clinical outcome for human motion and behavior analysis in order to investigate human arm movement during jogging and walking. Three methods were proposed to differentiate and separate both of the jogging and walking data set, they are statistical tabular, scatter distribution and artificial classifier recognition. Linear decision boundary and radial basis function kernel (RBF) were proposed to perform the separation works for artificial classifier recognition section. It aims to establish how widespread the movement and motion of arm will bring to effect of human in life. An experiment was set up in a laboratory environment with conjunction of analyzing human motion and its behavior. The instruments demonstrate adequate internal consistency of optimum RBF kernel for jogging and walking pattern classification. RBF used in this study was successfully differentiate and classify the jogging and walking patterns of a human arm movement during performing these activities.

Left and right hand gesture classification using scatter diagram and Principle Component Analysis

This study investigates and acts as a trial clinical outcome for human motion and hand behaviour analysis in consensus of subjects habit related quality of life. It was developed to analyse and access the quality of human hands motion that can be used in hospitals, clinics and human motion researches. It aims to establish how widespread the quality of life effects of human motion. An experiment was set up in a laboratory environment with conjunction of analysing human hand motion and its habit. Sensors are attached on both wrists. The instruments demonstrate adequate internal consistency of findings: 1. it is hard for subject to draw a perfect circle whether using left or right hand and this is supported by descriptive statistical data and scatter diagram. 2. Subjects left hand unable to draw a perfect circle or square. These two drawings are looks alike and it is supported by PCA analysis. A simple and informative representation for statistical data, scatter diagram and PCA plot were developed to demonstrate the results.