Precise segmentation of Pap smear cell nucleus is crucial for early diagnosis cervical cancer. This task particularly challenging because overlapping, inconsistent staining, poor contrast and other imaging artifacts. In this study, a novel method proposed to segment from overlapping images. The technique introduces circular shape function (CSF) increase the robustness using fuzzy c-means clustering. CSF imposes constrain over formed clusters, while improves boundary nucleus. helps...

The study addresses the challenging problem of automatic segmentation human anatomy needed for radiation dose calculations.Three-dimensional extensions two well-known state-of-the art techniques are proposed and tested usefulness on a set clinical CT images.The new 3D Statistical Region Merging (3D-SRM) Efficient Graph-based Segmentation (3D-EGS). Segmentations eight representative tissues (lungs, stomach, liver, heart, kidneys, spleen, bones spinal cord)were accuracy using Dice index,...

An automatic method for detection of mammographic masses is presented which utilizes statistical region merging segmentation (SRM) and linear discriminant analysis (LDA) classification. The performance the scheme was evaluated on 36 images selected from local database mammograms 48 taken Digital Database Screening Mammography (DDSM). Az value (area under ROC curve) classifying each 0.90 dataset 0.96 DDSM. Results indicate that SRM can form part an robust efficient basis mammograms.

Mass segmentation in mammograms is a challenging task if the mass located local dense background. It can be due to similarity of intensities between overlapped normal breast tissue and mass. In this paper, self- adjusted mammogram contrast enhancement solution called Adaptive Clip Limit CLAHE (ACL-CLAHE) developed, aiming improve regions mammograms. An optimization algorithm based on entropy used optimize clip limit window size standard CLAHE. The proposed method tested 89 images with 41...

Cervical nuclei contain important diagnostic characteristics useful for identifying abnormality in cervical cells. Therefore, an accurate segmentation of is the primary step computer-aided diagnosis. However, cell overlapping, uneven staining, poor contrast, and presence debris elements make this task challenging. A novel method presented paper to detect segment from overlapping smear images. The proposed framework segments by merging superpixels generated statistical region (SRM) algorithm...

A technique utilizing an entropy measure is developed for automatically tuning the segmentation of screening mammograms by minimum spanning trees (MST). The lack such has been a major obstacle in previous work to segment registration and applying mass detection algorithms. proposed method tested on two sets mammograms: set 55 chosen from publicly available Mini-MIAS database, 37 selected local database. performance evaluated conjunction with three different preprocessing filters: gaussian,...

A method based on sublevel sets is presented for refining segmentation of screening mammograms. Initial provided by an adaptive pyramid (AP) scheme which viewed as seeding the final sets. Performance tested with and without prior anisotropic smoothing compared to refinement component merging. The combination smoothing, AP found outperform other combinations.

Image segmentation based on minimum spanning trees (MST) is used to identify the pectoral muscle in screening mammograms. The found using MST initialise an active contour for finding anatomically reasonable estimate of boundary muscle. error reported terms number in-correctly assigned pixels. Out 83 images, 25 images have rates less than 5 percent and 56 10 percent. nature errors encountered indicates that accuracy computer algorithms this task approaching its practical limit.

A framework to detect and segment nuclei from cervical cytology images is proposed in this study. Poor contrast, spurious edges, degree of overlap, intensity inhomogeneity make the segmentation task more complex overlapping cell images. The technique segments by merging over-segmented SLIC superpixel regions using a novel region criteria based on pairwise regional contrast image gradient contour evaluations. was evaluated first challenge - ISBI 2014 dataset. result shows that outperforms...

The paper studies the feasibility of using 3D extensions two state-of-the-art segmentation techniques, Statistical Region Merging (SRM) method and Efficient Graph-based Segmentation (EGS) technique, for automatic anatomy on clinical CT images.

A graph based segmentation approach is proposed in this study to segment nucleus from cytology images. This utilizes a novel method applying weighted circular shape prior adaptively efficient image segmentation. The was evaluated by segmenting two public Pap smear datasets: ISBI 2014 challenge dataset (945 images) and DTU/Herlev intermediate squamous cell (70 images). Segmentation results of the outperformed standard one terms Dice similarity coefficient, pixel-based precision recall,...

Computer-assisted techniques for cytological analysis and abnormality detection, can help to early diagnose anomalies in cervical smear images. Cell nuclei carry substantial evidence of pre-cancerous changes, thus morphological properties are important accurate diagnostic decision. A novel nucleus feature-based cell classification framework is proposed this study. Prior guided segmentation algorithms employed accurately detect segment nucleus. Fuzzy entropy based feature selection technique...

Finding masses in dense background is a difficult task for even experienced radiologist. It due to the similarity of intensity between and overlapped normal tissues. A novel method classification localised breast proposed. Nine structured superpixel patterns were generated using local binary pattern technique on superpixels. Analysis these nine revealed most prominent ones, allowing successful malignant regions. Two mammographic databases used evaluate proposed approach: publicly available...

Variability in the color appearance H and E stained histopathological images are typically observed. Color normalization has been found useful standardizing of prior to quantitative analysis with machine learning (using handcrafted features). However, its usefulness not previously studied when deep convolutional neural networks (CNNs) used classifying breast cancer images. In this paper, we have adopted a representative CNN for evaluated benefit/necessity normalisation using commonly...

The performance of two image segmentation methods are compared according to robustness the distortion. This criterion is crucial for temporal analysis screening mammograms where natural changes in breast plus inherent deformation soft tissue during acquisition result severe registration problems. A method based on minimum spanning trees (MST) found be more robust distortions studied than a adaptive pyramids (AP). Although leads great differences distorted images many components low saliency,...