INTRODUCTION:

Malaria is caused by parasites of the species Plasmodium that are spread from person to person through the bites of infected mosquitoes. A parasite is an organism that lives off another organism. Animals can also get malaria, but malaria cannot be passed from humans to animals or from animals to humans.

Facts and Figures: Approximately, 40% of the world’s population, mostly those living in the world’s poorest countries, is at risk of malaria. A child dies of malaria every 30 seconds. Every year, more than 500 million people become severely ill with malaria¹. Between 300 million and 500 million people in Africa, India,

Southeast Asia, the Middle East, the South Pacific, and Central and South America have the disease. The worldwide annual economic burden of malaria, calculated to include spending on prevention and treatment as well as loss of productivity due to illness, was estimated at US$ 500 million in 2005².

Diagnosis of Malaria: The definitive diagnosis of malaria infection is done by searching for parasites in blood slides (films) through a microscope. In peripheral blood sample visual detection and recognition of Plasmodium spp is possible and efficient via a chemical process called (Giemsa) staining. The staining process slightly colorizes the red blood cells (RBCs) but highlights Plasmodium spp parasites, white blood cells (WBC), and platelets or artifacts. The detection of Plasmodium spp requires detection of the stained objects. However, to prevent false diagnosis the stained objects have to be analyzed further to determine if they are parasites or not.¹

In the fig.1:, There are four types of human malaria – Plasmodium falciparum, P. vivax, P. malariae, and P. ovale. P. falciparum and P. vivax are the most common. P. falciparum is by far the most deadly type of malaria infection

Fig1. a) Plasmodium falciparum

b) P. vivax

c) P. malariac

d) P. ovale

Goals: The biggest detraction of microscopy, namely its dependence on the skill, experience and motivation of a human technician, is to be removed. Used with an automated digital microscope, which would allow entire slides to be examined, it would allow the system to make diagnoses with a high degree of certainty. It would also constitute a diagnostic aid for the increasing number of cases of imported malaria in traditionally malaria-free areas, where practitioners lack experience of the disease.

Objectives: The objective of the project is to develop a fully automated image classification system to positively identify malaria parasites present in thin blood smears, and differentiate the species. The algorithm generated will be helpful in the area where the expert in microscopic analysis may not be available. The effort of the algorithm is to detect presence of parasite at any stage. One of the parasites grows in body for 7 to 8 days without any symptoms. So if this algorithm is incorporated in routine tests, the presence of malarial parasite can be detected

PROPOSED ALGORITHM

The design follows the same steps as that of a pattern recognition problem. But the best part of the algorithm is the usage of the most appropriate algorithm for each stage. The test algorithms illustrated above give an insight about the algorithm to be used for each stage. The process is given below.

1.) Image Acquisition and database collection

2.) Image Analysis

3.) Image Segmentation

4.) Feature Generation

5.) Classification of Parasite and result verification

Image acquisition and database collection:

Oil immersion views (10x1000), of Giemsa stained blood films were captured using a binocular microscope mounted with a digital camera. Captured images were 460 pixels X 307 pixels bitmap images. Fig 2. shows the sample slide of P. falciparum. The database consists of 110 images.

Fig2.Plasmodium falciparum

Image analysis:

In an image an edge is a Curve that follow a path of rapid change in image intensity, edge are often associated with the boundaries of object in a scene. The most powerful edge detection method that edges provides in the canny method. The canny method¹⁰ differs from the other edge detection methods that are two different threshold and infection and includes the weak edges in the output only if they are connected to strong edges. This Method is therefore less likely than the others to be tooled by noise, and more likely to detect true weak edges.

Image segmentation:

Dilation and erosion are two fundamental morphological operations. Dilation adds pixels to the boundaries of objects in an image, while erosion remove pixels on object boundaries. The number of pixels added or removed from the objects in an image depends on the size and shape of the structuring element used to process the image. In morphological dilation and erosion operations, the state of any given pixel in the output image is determined by applying a rule to the corresponding pixel and its neighbors in the input image. The rule used to process the pixels defines the operation as dilation or erosion. In the dilation the value of the output pixel is the maximum value of all the pixels in the input pixels neighborhood. In binary image, if any of the pixels is set to value 1, the output pixel is set to 1. Erosion is the value of the output pixel is the minimum value of all the pixels in the input pixel’s neighborhood. In a binary image, if any of the pixels is set to 0, the output pixel is set to 0.

FEATURE GENRATION AND CLASSIFICATION:

Feature generation:

Two sets of features are used for development. The first set will be based on image characteristics that have been used previously in biological cell classifiers, which include geometric features (shape and size), colour attributes and grey-level textures.

It will be advantageous to apply expert, a priori knowledge to a classification problem. This will be done with the second set of features, where measures of parasite and infected erythrocyte morphology that are commonly used by technicians for manual microscopic diagnosis are used. It’s desirable to focus on these features, because it is already known that they are able to differentiate between species of malaria.

Feature Classification:

The final classification of an erythrocyte as infected with malaria or not, and if so, the species of the parasite, falls to the classifier. The classifier is a two-stage tree classifier, with an infection classified as positive or negative at the first node, and the species assigned at the second node.

Fig .3: Structure of the tree classifier

The design of a tree classifier has the following steps: the design of a tree structure (which has already been assigned), the selection of features to be used at every node, and the choice of decision rule at each node¹¹. The same type of classifier is used at both nodes. Taking into account the fact that there is no guarantee that the classes are linearly separable, back propagation feed forward (BFF) neural networks is selected. The features selected for the first classifier are those that describe the color and texture of the possible parasites. The features used by microscopists to differentiate malaria species are selected for the second classifier. The training goal is to minimize squared errors, and training is stopped when the error of a validation set increased. This is done to avoid overtraining.

RESULT:

Our project give good result in every steps, the first order of features provide the best accuracy for simple blood cell, and different types of parasite.

The algorithm has been tested on various malaria parasites, the results are:

(a)

(b)

(c)

(d)

(e)

(f)

( g)

fig3. a)Orignal image , b)Canny Output

c)Binary gradient mask , d) Dilated gradient mask

e)Cleared border image , f)Segmented image

g)Outlined orginal image

CONCLUSION:

The proposed algorithm parasite detection algorithm has many advantages compared to other diagnosis techniques. It avoids the problem associated with rapid methods, such as being species –specific and having high per-test costs, while retaining many of the traditional advantages of microscopy, viz. species differentiation, determination of parasite density, explicit diagnosis and low per-test costs.

Apart from overcoming the limitations of conventional methods of parasite detection, the proposed algorithm is optimized to overcome limitations of image processing algorithms used in the past. CANNY edge detection technique’ gives good edge detection than ‘SOBEL’ edge detection. Dilation and erosion methods provide better segmented object, in the feature generation.

REFERENCE:

1. World Health Organization.

What is malaria? Factssheetno.94. http://www.who.int/ mediacentre/factsheets/fs094/en/.

2. Foster S, Phillips M, Economics and its contribution to the fight against malaria. Ann Trop Med Parasitol 92:391–398, 1998.

3. Makler MT, Palmer CJ, Alger AL, A review of practical techniques for the diagnosis of malaria. Ann Trop Med Parasitol 92(4):419–433, 1998.

4. Bloland PB (2001) Drug resistance in malaria, WHO/CDS/CSR/DRS/ 2001.4. World Health Organization, Switzerland, 2001.

5. Gilles H.M. The differential diagnosis of malaria. Malaria. Principles and practice of malariology (Wernsdorfer W.H., McGregor I Eds), 769-779, 1998.

6. F. Castelli, G.Carosi, Diagnosis of malaria, Chapter 9, Institute of Infectious and Tropical Diseases, University of Brescia (Italy).

7. Anthony Moody, Rapid Diagnostic Tests for Malaria Parasites, Clinical Microbiology Reviews, 0893-8512/02/$04.00_0 DOI: 10.1128/CMR.15.1.66–78.2002, p. 66–78, Jan. 2002.

8. Brown A.E., Kain K.C., Pipithkul J., Webster H.K. Demonstration by the polymerase chain reaction of mixed Plasmodium falciparum and P. vivax infections undetected by conventionalmicroscopy. Transactions of the Royal Society of Tropical Medicine and Hygiene; 86: 609-612, 1992.

9. Silvia Halim, Timo R. Bretschneider, Yikun Li, Estimating Malaria Parasitaemia from Blood Smear Images. 1-4244-03421/06/$20.00 ©IEEE, ICARCV 2006.

10. Rafeal C. Gonzalez, Richard E.Woods, Digital Image Processing, 2^nd Edition, Prentice Hall, 2006.

11. Mui JK, Fu.K-S, Automated classification of nucleated blood cells using a binary tree classifier. IEEE Trans Pattern Anal Machine Intell 2(5):429–443, 1980.

Received on 06.02.2011 Accepted on 06.03.2011

Research J. Engineering and Tech. 2(2): April-June 2011 page 91-94