Recognition and Classification of Life cycle of the Malaria Parasite in the Microscopic Images
Ongoing (Sep 2015 _ Present)
Source of Support: Barkat Foundation
Total Amount:$ 200,000
E-health services are considered as modern services which have been very effective in enhancing health level of Iranians who live in the deprived parts of the country and have been associated with mental security and safety for patients of these areas. Relying on such services, patients can be in touch with their doctors using the communication systems and in any condition. Electronic health examinations are provided through 20 services and cover the following items: basic examination through intelligent beds in impassable areas, performing 20 routine medical tests such as blood sugar test, blood fat test, thyroid tests, liver enzymes test, blood tests and ECG, developing a connection between patients and medical centers, establishing a system to control vital signs in ambulances and proposing consultation services from diagnosis to issuing prescription through making connection with the call center of Barakat Medical Services center, etc. In this regard, electronic health services are offered by Barakat Tel Company, an affiliated company of Barakat Foundation, in five deprived counties of Iran, Paveh, Zehak, Kaleibar, Osko and Lendeh and as a result of such services a total of 350,000 Iranian are covered by the mentioned services.
Automatic Recognition and Classification of Life cycle of the Malaria Parasite in the Microscopic Images
Microscopy is among the conventional methods used for Malaria diagnosis. A blood sample of the patient is spread over a slide and examined under a microscope. For a highly trained professional, it takes hours to visually examine the slide and report the results. It is even more difficult to detect the different types of malaria parasite and their stages by the conventional methods. The proposed method in the present study involves acquisition of the thin blood smear microscopic image at 100x magnification, pre-processing by Homomorphic filter, and then separation of infected cell from the image. After that, the life cycle of the Malaria parasite needs to be calculated. The features which include area, number of nucleus and Elongation characteristic, are selected in a way that the independency would be preserved. Finally, after feature extraction, using a decision tree, life cycle of Malaria can be achieved as an unsupervised learning. The algorithm is applied on 21 digital images containing 735 objects which are provided from thin blood smear films. Applying the automatic identification of malaria on provided database showed a sensitivity of 89.2% and specificity of 99.2%.We compare the output with manual decision of the experts and results seems quiet acceptable.
Collaborators:
Assistant Professor of Electrical and Computer Engineering at The University of Tehran
Farnaz Mohammadi
BSc biomedical engineering student at the University of Tehran
