Anupama Bhan

Parametric models for segmentation of Cardiac MRI database with geometrical interpretation

There are several medical image modalities which enable a proper diagnosis of Heart Diseases. Magnetic Resonance imaging is one of the most non invasive techniques used by radiologist for diagnosis nowadays. Segmentation of Cardiac MRI is basically dividing the heart into left and right ventricle, where left ventricle plays a vital role. In this paper the segmentation of Left Ventricle using parametric model is done which helps to calculate geometrical parameters. These parameters further helps to distinguish between an abnormal and normal heart. The segmentation is carried out on multi slice MRI frame. Parametric model used in this paper is Active Contour Model which is also called as Snake Model or Deformable model. The geometrical analysis is shown graphically to show the difference between different patient databases. The clinical diagnosis is also made on the physiological parameters like volume of blood and ejection fraction which will be the future extension to this paper.

Cervical cancer detection and classification using Independent Level sets and multi SVMs

Introduced in 1940, Pap smear test has proven to be an effective screening method to determine the different stages of cervical cancer. Identification and classification of Pap smear images to detect cervical cancer via manual screening is a challenging task for pathologists therefore increasing the chances of human error. In this paper, we propose an automatic method to detect and classify the grade of cervical cancer using both geometric and texture features of Pap smear images and classifying accordingly using multi SVM. The geometric features are obtained through segmentation of nucleus and cytoplasm using independent level sets, detecting whether the cell is cancerous or normal, with reference to the ground truth. By extracting well defined GLCM texture features and using a combination of PCA and the best class of multi SVM, the images are classified with an accuracy of 95%.

Analysis of histogram based compound contrast enhancement with noise reduction method for endodontic therapy

Radiographs are essential to all phases of endodontic therapy. They inform the diagnosis and the various treatment phases and help evaluate the success or failure of treatment. Because root canal treatment relies on accurate radiographs, it is necessary to master radiographic techniques to achieve films of maximum diagnostic quality. Such mastery minimizes retaking of films and reduces the radiation exposed on patient due to which image quality is low contrast. Hence, image processing techniques are an acceptable technique that can be used to improve the quality of image to assist dentist for diagnosis. Digital dental radiograph images are often noisy, blur edges and low in contrast. We have proposed the combination of sharpening and enhancement method to overcome these problems. The impulse noise is reduced by using median filtering technique. Then filtered image is passed through the homomorphic filter to improve the image illumination. The processed dental images for root canal teeth are contrast enhanced by sharp contrast adaptive histogram equalization. The combination of median and homomorphic filter with SCLAHE is compound contrast enhancement method. The results are verified in terms of peak signal to noise ratio (PSNR) and entropy of image.

Image-based pixel clustering and connected component labeling in left ventricle segmentation of cardiac MR images

This research demonstrates a completely automated sub-second fast technique for left ventricle (LV) segmentation from clinical cardiac MRI images for the crucial assessment of left ventricular dysfunction as a measure of cardiac diseases. In this work left ventricle segmentation is achieved using the combination of fuzzy c-means which is a pixel based classification method and connected component labeling. This strategic combination obviates user intervention and problem of seed point initialization as it automatically segments the LV accurately on all frames in the complete cardiac cycle in multi-frame MRI. The both methods complement each other such that it achieves sub-second fast computational speed of 0.7 seconds on average per frame. Thus this techniques computational time for left ventricle segmentation is much faster than iteration based methods. The accuracy of the automatic segmentation technique was tested against manual segmentation on the basis of correlation coefficient. The value of correlation coefficient between the automatic and manually traced LV boundaries was 0.932 which can be considered clinically significant.

Fast fully automatic multiframe segmentation of left ventricle in cardiac MRI images using local adaptive k-means clustering and connected component labeling

This paper presents a sub-second fast fully automatic method for segmentation of the left ventricle (LV) from cardiac MRI images, which plays a vital role in the diagnosis of left ventricular function for the assessment of cardiac disease in a patient. In this paper the segmentation of the left ventricle using local adaptive k-means clustering and connected components is achieved fully automatically. The segmentation is carried out on multi frame MRI. Adaptive k-means is used to cluster the pixels into groups based on their intensities in order to separate the foreground (ventricle) pixels from the background pixels. Connected component labeling is used to group the pixels into regions based on their connectivity in order to segment the LV pixel region from the other regions of the MRI image. This novel combined method eliminates the problem of initialization and iteration and it segments the LV accurately on multi frame MRI with sub-second fast computational times in the range of 0.01-0.1 seconds per frame. Thus this method achieves left ventricle segmentation for one frame in sub-second duration, much less than the time required for a single iteration in deformable model methods such as level sets and active contours. The automatic segmentations accuracy was also validated on two frames as the correlation coefficient between the automatic and manually traced LV boundaries (0.992 for frame 1 and 0.993 for frame 2) was found to be higher than the correlation coefficient between two manually traced LV boundaries (0.984 for frame 1 and 0.900 for frame 2) for the same frame.

Computer Aided Diagnosis of Cervical Cancer Using HOG Features and Multi Classifiers

Cervical cancer is very common in women, and it is the most dreaded disease. Cervical cancer if detected early can be treated successfully. Cervical cancer occurs due to the uncontrolled growth of the cells present in the cervix of the female body, and it also occurs due to the virus human papilloma virus (HPV). Pathologists diagnose cervical cancer by a screening test called Papanicolaou test or Pap smear test. The pap smear test is not always 100% accurate but it helps in early detection of cancerous cells. In this paper, a method is proposed that helps in detection and classification of the cancer using HOG feature extraction and classifying it by the help of support vector machine (SVM), k-nearest neighboring (KNN), artificial neural network (ANN). The database was collected from Air Force Command Hospital, Bengaluru. A total of 66 pap smear images were collected that are 25 normal pap smear images and 41 abnormal pap smear images. Histogram of gradient (HOG) extracts features of the region of interest in the image as it converts pixel-based representation into gradient-based representation. The classification of cervical cells—abnormal cells and normal cells—is done with the help of multi-classifier. The accuracy attained after classification is 62.12, 65.15, and 95.5% for SVM, KNN, and ANN, respectively.

Patient-Specific Cardiac Computational Modeling Based on Left Ventricle Segmentation from Magnetic Resonance Images

This paper presents three-dimensional computational modeling of the heart’s left ventricle extracted and segmented from cardiac MRI. This work basically deals with the fusion of segmented left ventricle, which is segmented using region growing method, with the generic deformable template. The multi-frame cardiac MRI image data of heart patients is taken into account. The region-based segmentation is performed in ITK-SNAP. The left ventricle is segmented in all slices in multi-frame MRI data of the whole cardiac cycle for each patient. Various parameters like myocardial muscle thickness can be calculated, which are useful for assessing cardiac function and health of a patient’s heart by medical practitioners. With the left ventricle cavity and myocardium segmented, measurement of the average distance from the endocardium to the epicardium can be used to measure myocardial muscle thickness.

Automatic Image Processing Based Dental Image Analysis Using Automatic Gaussian Fitting Energy and Level Sets

Identification of the Root canal length is a major concern in the dentistry worldwide, which currently seeks the manual calculation in order to detect the measurement of the teeth. Intensity inhomogeneity often is a major problem in dental x-rays which causes considerable difficulties in segmentation. For better computer-aided diagnosis in dentistry, having a precise tooth segmentation is a critical task, as the cysts and inflammatory lesions generally occur around tooth root areas and these areas in radiographs are generally subject to noise, poor contrast, and very uneven illumination. This paper presents an effective segmentation method using a combinational approach of Local Gaussian Distribution fitting energy along with level sets. Here the local intensities of images are defined by Gaussian distributions which are combined with the level set function for accurate segmentations of teeth contour. The experimental results indicate that segmentation achieves the less number of iterations making it computationally fast and work in real time situation.

Detection and Grading Severity of Caries in Dental X-ray Images

It is significant to analyze the dental images in order to improve and quantify medical images for correct diagnosis. Caries or cavity is one of the most prevalent diseases of the teeth. Dentists are putting the best effort to identify the problem at an earlier stage. The proposed method used in this paper is focused on the challenges faced during the root canal edge extraction from dental radiographic images, which is a major problem besides cavity detection and extraction. The image processing techniques helps to identify the caries that provide dentists with the precise results of the area affected by the caries. The proposed methodology consists of preprocessing of bitewing radiographic images using top hat bottom hat transformation followed by the sharpening filter for edge enhancement. This combinational approach provides qualitative and quantitative assessment to dentists on the presence of cavity. The caries are extracted by some morphological tools to grade the severity on the basis of some metric values. Preparatory experiments show the significance of the proposed method to extract cavity and grade its effect on the tooth.

Feature Line Profile Based Automatic Detection of Dental Caries in Bitewing Radiography

Dental caries is a bacterial infection that causes tooth decay and is amongst the most common incessant maladies of individuals around the world. Teeth are defenseless to this infection all through their lifetime especially when care is not taken for proper oral hygiene. It is significant to analyze the dental images in order to improve and quantify medical images for correct diagnosis. Caries or cavity is one of the most prevalent diseases of the teeth. Dentists are putting the best effort to identify the problem at an earlier stage. The proposed method used in this paper is focused on the challenges faced during the cavity detection which sometimes is very tedious task due to small lesions not visible to human eye. The image processing techniques helps to identify the caries that provide dentists with the precise results of the area affected by the caries. The proposed methodology consists of preprocessing of bitewing radiographic images followed by edge recognition, thresholding and connected component labelling. This combinational approach provides qualitative and quantitative assessment to dentists on the presence of cavity. The caries are detected by connected component and mask overlap helps to highlight the affected area to grade the severity which is tested on the basis of line intensity profiles. Preparatory experiments show the significance of the proposed method to extract cavity and grade its effect on the tooth.