You are on page 1of 14

OBJECTIVE

To design a method for segmenting


nuclei from microscopy images of
arbitrary types (scalar, color,
fluorescence, different staining, etc.).
INTRODUCTION
A persistent difficulty in evaluating digitized
pathologic specimens is its subjective nature.

An automatic, high throughput image analysis


system usually requires accurate and robust nucleus
segmentation as the pre-requisite step.

However, it is difficult to separate nuclei from each


other a well as the background due to nucleus
occlusion or touching, shape variation, intra-nucleus
inhomogeneity, background clutter, and poor
contrast
LITERATURE SURVEY
An Integrated Region-, Boundary-, Shape-Based
Active Contour for Multiple Object Overlap
Resolution in Histological Imagery-2012

Synergistic boundary and region-based active contour


model that incorporates shape priors in a level set
formulation with automated initialization based on
watershed.

METHODS/TECHNIQUES

PRINCIPAL COMPONENT ANALYSIS (PCA)


MERITS

to segment all the overlapping and non overlapping


objects within an image simultaneously.

DEMERITS

variability in digital slide appearance due to


inconsistencies in histological staining,
poor image quality with tissue samples due to slide
digitization
tissue fixation.
LITERATURE SURVEY
Partitioning Histopathological Images: An
Integrated Framework for Supervised Color-
Texture Segmentation and Cell Splitting -2011

An efficient LFT extraction algorithm based on image


shifting and image integral. For the splitting part, given
a connected component of the segmentation map, we
initially differentiate whether it is a touching-cell clump
or a single non touching cell.

METHODS/TECHNIQUES

LOCAL FOURIER TRASNFORM(LFT)


MERITS

The advantage of LFT segmentation


over the others cannot be fully reflected
by the marginal improvement if only
based on the segmentation accuracy.

DEMERITS

contours containing multiple


overlapping objects pose
LITERATURE SURVEY
Hypercolumns for Object Segmentation and
Fine-grained Localization-2015

Recognition algorithms based on convolutional


networks (CNNs) typically use the output of the last
layer as a feature representation.

METHODS/TECHNIQUES

CONVOLUTIONAL NETWORKS(CNN)
MERITS

The high resolution of the lower layers and


up sample the higher layers instead.

DEMERITS

The first problem is simultaneous detection


and segmentation (SDS).

The Second problem deals with detecting an


object and localizing its parts.
Problem statement
The choice of marker detection algorithms for
watershed segmentation highly depends on the
specific applications, and it is difficult to apply
a general rule.

Meanwhile, the aforementioned detection


methods are not robust to image noise or
require careful parameter selection, and thus
might need further processing to merge or split
falsely segmented regions from watershed.
Proposed system
Computer-aided image analysis of histopathology
specimens could potentially provide support for early
detection and improved characterization of diseases such
as brain tumor, pancreatic neuroendocrine tumor (NET).

a learning-based framework for accurate and automatic


nucleus segmentation with shape preservation.

the proposed algorithm on three large-scale pathology


image datasets using a range of different tissue and stain
preparations, and the comparative experiments with
recent state of the arts demonstrate the superior
performance of the proposed algorithm.
The positive probability map, which
denotes how likely each pixel to be
positive, to remove those pixels with
probabilities lower than threshold q1
before further analysis. In addition, there
may be some small noise caused regions.
We use the region size as a criterion and
eliminate those regions with size less
than a threshold before performing the
subsequent shape initialization.
Proposed diagram
Thank you

You might also like