Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks


No description

sabrina atzori

on 14 January 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Ematologia

Automatic Analysis of
Microscopic Images in
Hematological Cytology
Applications Automatic Analysis Andrea Loddo &
Sabrina Atzori Blood smear Bone Marrow smear Enhancement Tonal domain (histogram equalization)
Spatial domain (unsharp filters)
Spectral domain (log-Fourier) Boundary based Segmentation Region based Segmentation Segmentation Cell level segmentation
Component level segmentation Segmentation Improvement Edge Detectors
Active Contours
explicit models (snakes)
implicit models (level sets) Thresholding
Pixel based Classification
Region Growing
Region Clustering
Watershed Transform Otsu Algorithm Pixel based Classification K-means Algorithm Features Extraction Based on Visual Descriptors Chromatic Features Shape Descriptors Geometrical Features

Structural Based Representations

Spectral Based Features Textural Descriptors Statistical Approaches
Model Based Approaches
Geometric Approaches Applications cell distinction

leukocytes classification

infection stages Histogram moments of segmented objects Application: sequential thinning to represent segmented malaria parasites Application: number of chromatin dots count to identify infection stage Median filter Normalization by Tek Leukocytes, Erythrocytes, Platelets
Classification Leukocytes Classification System Splitting of Clumped Cells Sub-image
Morphological operators
Distance Transform
Concavity Analysis
Template matching strategy Clumped splitting using distance transform and watershed segmentation Clumped cells Segmented image Euclidean distance transform Regional maxima of the distance transform Watershed segmentation using maxima points as markers Classification Features Selection Features Selection Concavity Analysis cut point detected by measuring the concavity depth (distance from convex hull border) Segmented image after the best split line selection Live stages of hemoparasite recognition Throphozoites identification Erythrocytes classification
for each live stage 1: liver 2: generic RBC 3: new copies healty infected Fourier Based Transforms: EFD
Wavelet Transform Application: leukocytes and megakaryocytes nuclei classification 3 types:
filter-based Genetic Algorithms

Features reduction: PCA k-NN : N samples mapped in M classes Application: Leukocyte Classification and Differentiation The end... Case of Application Phase 1:
Image Acquisition Phase 2:
Image Preprocessing Phase 3, part I
Erythrocyte Segmentation 1 2 3 Thanks to: nucleus area, perimeter and color ratios Phase 4:
Feature Extraction 4 5 Problem: Luminance Variations Original image Low-pass filtered
image Original image Pixel-Based Classification Filtering by inclusion-tree
representation Original clumped cells Splitting result cells Phase 3, part II:
Erythrocyte Segmentation 25 features extracted : 5 statistical moments from
5 histograms functions Problem: slight differences between consecutive life stages Phase 5:
Cell Classification Training set
Validation set
Test set Goal: estimate the infection level of each smear ANN : Neural Networks Application: Hierarchical Classification of Erythrocytes Regular Shapes
Irregular Shapes
Full transcript