Malaria detection using custom Semantic Segmentation Neural Network Architecture

Malaria is a significant disease that affects both animals and humans. The four main Plasmodium species cause human malaria are falciparum, vivax, malariae, ovale. knowlesi, parasite typically infecting forest macaque monkeys, was recently revealed to be able transmitted by anophelines provoke in This provides an increasing risk of spreading the areas previously unaffected with it people during increasingly popular travels abroad. Microscopic examination remains one most often used methods for its laboratory confirmation. These tests, however, should performed immediately after receiving samples from firstcontact doctor allow immediate therapy. research presents novel, semantic segmentation neural network architecture designed quickly create classification mask, giving information about position, shape, possible affiliation detected elements. evaluation method based on light microscope imagery created overcome problems resulting diagnosis specifics. There 3 abstract classes containing healthy cells, cells background. outputted mask can later mapped more readable form inclusion contrasting colors, next original image quick validation. Such approach allows semi-automatic recognition disease, nevertheless still final verdict specialist. developed solution has achieved high accuracy 96.65%, while computer power requirements kept at minimum. proposed help reduce misclassification rates providing additional data speed up entire process early made deep learning model.