Norliza Mohd Noor

Regression Methods to Investigate the Relationship Between Facial Measurements and Widths of the Maxillary Anterior Teeth

STATEMENT OF PROBLEM In rehabilitating edentulous patients, selecting appropriately sized teeth in the absence of preextraction records is problematic. PURPOSE The purpose of this study was to investigate the relationships between some facial dimensions and widths of the maxillary anterior teeth to potentially provide a guide for tooth selection. MATERIAL AND METHODS Sixty full dentate Malaysian adults (18-36 years) representing 2 ethnic groups (Malay and Chinese), with well aligned maxillary anterior teeth and minimal attrition, participated in this study. Standardized digital images of the face, viewed frontally, were recorded. Using image analyzing software, the images were used to determine the interpupillary distance (IPD), inner canthal distance (ICD), and interalar width (IA). Widths of the 6 maxillary anterior teeth were measured directly from casts of the subjects using digital calipers. Regression analyses were conducted to measure the strength of the associations between the variables (alpha=.10). RESULTS The means (standard deviations) of IPD, IA, and ICD of the subjects were 62.28 (2.47), 39.36 (3.12), and 34.36 (2.15) mm, respectively. The mesiodistal diameters of the maxillary central incisors, lateral incisors, and canines were 8.54 (0.50), 7.09 (0.48), and 7.94 (0.40) mm, respectively. The width of the central incisors was highly correlated to the IPD (r=0.99), while the widths of the lateral incisors and canines were highly correlated to a combination of IPD and IA (r=0.99 and 0.94, respectively). CONCLUSIONS Using regression methods, the widths of the anterior teeth within the population tested may be predicted by a combination of the facial dimensions studied.

A discrimination method for the detection of pneumonia using chest radiograph.

This paper presents a statistical method for the detection of lobar pneumonia when using digitized chest X-ray films. Each region of interest was represented by a vector of wavelet texture measures which is then multiplied by the orthogonal matrix Q(2). The first two elements of the transformed vectors were shown to have a bivariate normal distribution. Misclassification probabilities were estimated using probability ellipsoids and discriminant functions. The result of this study recommends the detection of pneumonia by constructing probability ellipsoids or discriminant function using maximum energy and maximum column sum energy texture measures where misclassification probabilities were less than 0.15.

Contrast resolution enhancement based on wavelet shrinkage and gray level mapping technique

The speckle pattern produced by the use of a coherent transducer in forming the B-mode ultrasound images is also a source of contrast resolution-degradation. We have presented the soft thresholding rule to relieve artifacts from the discrete wavelet transform of an image. Our study begins with the estimation of the required determinants to obtain the desired noise/resolution trade off, extended to the gray level mapping technique (GLM) application. The GLM application is used to further enhance the ultrasound images in different contrast levels. Experiments have shown that the proposed methods have significantly improved PSNR in the specular and reflectance area. Both techniques are translated into a PC-based tool that could be transferred into ultrasound system that can be implemented for real time image processing.

Segmentation of the Lung Anatomy for High Resolution Computed Tomography (HRCT) Thorax Images

In diagnosing interstitial lung disease (ILD) using HRCT Thorax images, the radiologists required to view large volume of images (30 slices scanned at 10 mm interval or 300 slices scanned at 1 mm interval). However, in the development of scoring index to assess the severity of the disease, viewing 3 to 5 slices at the predetermined levels of the lung is suffice for the radiologist. To develop an algorithm to determine the severity of the ILD, it is important for the computer aided system to capture the main anatomy of the chest, namely the lung and heart at these 5 predetermined levels. In this paper, an automatic segmentation algorithm is proposed to obtain the shape of the heart and lung. In determine the quality of the segmentation, ground truth or manual tracing of the lung and heart boundary done by senior radiologist was compared with the result from the proposed automatic segmentation. This paper discussed five segmentation quality measurements that are used to measure the performance of the proposed segmentation algorithm, namely, the volume overlap error rate (VOE), relative volumetric agreement (RVA), average symmetric surface distance (ASSD), root mean square surface distance (RMSD) and Hausdorff distance (HD). The results showed that the proposed segmentation algorithm produced good quality segmentation for both right and left lung and may be used in the development of computer aided system application.

Determining features for discriminating PTB and normal lungs using phase congruency model

The appearance of the infected zone on the digital chest X-ray image for pulmonary tuberculosis (PTB) does not conform to standard shape, size or configuration. This study uses phase congruency (PC(x)) values to gather information from transition of adjacent pixel values that may be used as features to represent known disease type. The feature vector consisting of the average, variance, coefficient of variation and maximum PC(x)-values was found to be able to detect PTB with high accuracy.

A statistical interpretation of the chest radiograph for the detection of pulmonary tuberculosis

This paper presents a statistical interpretation of the chest radiograph for the detection of pulmonary tuberculosis (PTB). Each region of interest was represented by a vector of wavelet texture measures which is then multiplied by the orthogonal matrix Q. The first two elements of the transformed vectors were shown to have a bivariate normal distribution. Misclassification probabilities were estimated using probability ellipsoids and discriminant functions. The most important result of this study recommends the detection of pulmonary tuberculosis by constructing discriminant function using maximum column sum energy texture measures where the misclassification probabilities were less than 0.15. In the validation exercise, the proposed discriminant procedure yielded 94% correct classification rate.

Wavelet as features for tuberculosis (MTB) using standard X-ray film images

The success of eliminating the disease Mycobacterium Tuberculosis (MTB) depends on the detection capabilities of medical organizations. In Malaysia government hospitals perform the major part of this particular task. An important ingredient of the diagnostic process in government hospital is the visual interpretation of standard chest X-ray films. Problems associated with this type of visual interpretation are well known. In this study, we propose an objective method, involving wavelets, for the purpose of detecting MTB.

Segmentation and detection of media adventitia coronary artery boundary in medical imaging intravascular ultrasound using otsu thresholding

In this paper we present an automated segmentation method to detect the boundary between adventitia and media on the cross sectional view of the artery of patients who have plaques. The problem encounter is that the boundaries of the adventitia, media, intima and lumen are embedded when plaques exist. Moreover, the artery disease has damaged the tissue layers. This paper proposed a method in segmenting and detecting the outer boundary which is the media adventitia area of the artery using intravascular ultrasound (IVUS) images. The proposed method for segmentation is to use Otsu thresholding, followed by empirical thresholding and binary - morphological operation. The data used in this study was 10 samples from dataset B of IVUS images, courtesy of Simone Balocco (Training set, Computer Vision Center, Bellaterra, Universitat de Barcelona, Dept. Matemàtica Aplicada i Anàlisi, Barcelona). The proposed method shows promising result in detecting and segmenting the media adventitia boundary of the IVUS images.

Double segmentation method for brain region using FCM and graph cut for CT scan images

In the field of neuropsychiatric disorders, it is known that brain segmentation is important for both detection and diagnosis. The segmentation of the brain, which leads to the computation of brain volume proved to be vital in the detection of many brain pathology having Computed Tomography (CT) scan as the primary modality. Due to the fact that Fuzzy c-Means (FCM) proven to be robust, it is often used in data clustering and also in image segmentation. On the other hand, Graph cut is also a great segmentation algorithm for image segmentation as it allows the separation of the image into numerous partitions according to the similarity between each nodes in the image. In this paper, FCM was first used as global processing on CT scan images that separated the images into clusters based on pixel intensity. After that, local processing with graph cut algorithm was carried out on the automatically selected cluster from the FCM. Manual interaction is needed after the images were separated into partitions to select the appropriate partitions that best represent the brain region. The results showed that the images are less erroneous when they are clustered first with FCM before going through the graph cut algorithm.

Detection of the lumen boundary in the coronary artery disease

The intravascular ultrasound (IVUS) modality is used by the medical practitioner to detect the coronary artery disease called atherosclerosis, which is the hardening of the artery wall and subsequently narrow the blood vessel. In this paper, we present the segmentation method for detecting the lumen border of a coronary artery using IVUS images. The automated segmentation used is Otsu threshold, binary-morphological operation and empirical threshold. Thirty samples of IVUS images inclusive of the ground truth (manual tracings) were obtained from Computer Vision Centre, Bellaterra, Dept. Matematica Aplicada i Anàlisi, Universitat de Barcelona, Barcelona was used in this study. The result of the proposed automated segmentation is then compared with the ground truth provided. The segmentation performance of the proposed method is measured using Jaccard Index (JI), Dice Similarity Index (DI), Hausdorff Distance (HD), Area Overlapped Error (AOE) and Percentage Area Difference (PAD). The Bland Altman Plot is used to show the variation between the proposed automatic segmentation and ground truth. The results obtained show that the segmentation performance based on JI, DI, AOE and PAD of the proposed method is reasonably good when compared to other existing segmentation methods. However, further improvement is needed to obtain better HD value.