TENSORFLOW BASED PREDICTION MODEL FOR CLASSIFYING HUMAN BLOOD SMEAR MICROSCOPIC IMAGES AS INDICATING PRESENCE OF MALARIA PARASITE
Abstract
Malaria is now considered to be present in south Asian and African regions. Many countries are declared and certified as malaria free nations. But India is yet seen to be hosting with one or more indigenous cases. As per the World Health Organization (WHO) report India is one among the 25 nations expected to be malaria free by 2025. Currently the two variants out of four species of malaria parasite are prevalent in India. Falciparum and Vivax are the seen in several states of the country. The malaria detection is a manual procedure followed in the pathology laboratories. The human blood smears are collected and examined under the microscope. This process requires the experienced staff for better and accurate identification of the stages of the parasite detection. As per the standards specified for examination of a single slide is a minimum of 3 minutes and maximum of 5 minutes. But the process usually consumed by the pathology lab staff is minimum of 7 minutes with thick smears. Whereas the time duration required in case of thin smears is usually more. The proposed work aims at automating this process of detecting and identifying from thin blood smears without any compromise on the actual process adopted in detecting the parasite stages. The experiments are conducted on 200 images from Kaggle database, and the results obtained are encouraging.

Authors
Preethi N Patil1, Andhe Dharani2, S. Anupama Kumar3, Naganagouda Patil4
RV College of Engineering, India1,2,3, Sri Krishna Institute of Technology, India4

Keywords
Malaria, Microscopic Image, Blood Smear, Tensor Flow, Prediction
Yearly Full Views
JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecember
000000000000
Published By :
ICTACT
Published In :
ICTACT Journal on Soft Computing
( Volume: 13 , Issue: 3 , Pages: 2995 - 2997 )
Date of Publication :
April 2023
Page Views :
344
Full Text Views :
4

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.