Convolutional neural networks to automate the screening of malaria in low-resource countries

Oliver S Zhao; Nikhil Kolluri; Anagata Anand; Nicholas Chu; Ravali Bhavaraju; Aditya Ojha; Sandhya Tiku; Dat Nguyen; Ryan Chen; Adriane Morales; Deepti Valliappan; Juhi P Patel; Kevin Nguyen

doi:10.7717/peerj.9674

Convolutional neural networks to automate the screening of malaria in low-resource countries

PeerJ. 2020 Aug 4:8:e9674. doi: 10.7717/peerj.9674. eCollection 2020.

Authors

Affiliations

¹ Department of Biomedical Engineering, The University of Texas at Austin, Austin, TX, United States of America.
² Department of Electrical & Computer Engineering, The University of Texas at Austin, Austin, TX, United States of America.
³ Department of Psychology, The University of Texas at Austin, Austin, TX, United States of America.

Abstract

Malaria is an infectious disease caused by Plasmodium parasites, transmitted through mosquito bites. Symptoms include fever, headache, and vomiting, and in severe cases, seizures and coma. The World Health Organization reports that there were 228 million cases and 405,000 deaths in 2018, with Africa representing 93% of total cases and 94% of total deaths. Rapid diagnosis and subsequent treatment are the most effective means to mitigate the progression into serious symptoms. However, many fatal cases have been attributed to poor access to healthcare resources for malaria screenings. In these low-resource settings, the use of light microscopy on a thin blood smear with Giemsa stain is used to examine the severity of infection, requiring tedious and manual counting by a trained technician. To address the malaria endemic in Africa and its coexisting socioeconomic constraints, we propose an automated, mobile phone-based screening process that takes advantage of already existing resources. Through the use of convolutional neural networks (CNNs), we utilize a SSD multibox object detection architecture that rapidly processes thin blood smears acquired via light microscopy to isolate images of individual red blood cells with 90.4% average precision. Then we implement a FSRCNN model that upscales 32 × 32 low-resolution images to 128 × 128 high-resolution images with a PSNR of 30.2, compared to a baseline PSNR of 24.2 through traditional bicubic interpolation. Lastly, we utilize a modified VGG16 CNN that classifies red blood cells as either infected or uninfected with an accuracy of 96.5% in a balanced class dataset. These sequential models create a streamlined screening platform, giving the healthcare provider the number of malaria-infected red blood cells in a given sample. Our deep learning platform is efficient enough to operate exclusively on low-tier smartphone hardware, eliminating the need for high-speed internet connection.

Keywords: Data science; Deep learning; Epidemiology; Global health; Infectious disease; Machine learning; Malaria; Neural networks; Public health; Red blood cell.

Grants and funding

The research was funded by donations provided to Texas Engineering World Health (TEWH), a student-chapter of the parent organization Engineering World Health, based at The University of Texas at Austin. Individual donors and other TEWH members that are not listed on the authorship list had no role in any part of the research or writing of the manuscript. There was no additional external funding received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.