Format

Send to

Choose Destination
J Med Imaging (Bellingham). 2018 Jul;5(3):034501. doi: 10.1117/1.JMI.5.3.034501. Epub 2018 Jul 18.

Understanding the learned behavior of customized convolutional neural networks toward malaria parasite detection in thin blood smear images.

Author information

1
Lister Hill National Center for Biomedical Communications, National Library of Medicine, Bethesda, Maryland, United States.
2
Mahidol University, Mahidol Oxford Tropical Medicine Research Unit, Bangkok, Thailand.
3
Chittagong Medical Hospital, Department of Medicine, Chittagong, Bangladesh.
4
University of Missouri, MU Informatics Institute, Department of Pathology and Anatomical Science, Columbia, Missouri, United States.
5
University of Oxford, Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, Oxford, United Kingdom.

Abstract

Convolutional neural networks (CNNs) have become the architecture of choice for visual recognition tasks. However, these models are perceived as black boxes since there is a lack of understanding of the learned behavior from the underlying task of interest. This lack of transparency is a serious drawback, particularly in applications involving medical screening and diagnosis since poorly understood model behavior could adversely impact subsequent clinical decision-making. Recently, researchers have begun working on this issue and several methods have been proposed to visualize and understand the behavior of these models. We highlight the advantages offered through visualizing and understanding the weights, saliencies, class activation maps, and region of interest localizations in customized CNNs applied to the challenge of classifying parasitized and uninfected cells to aid in malaria screening. We provide an explanation for the models' classification decisions. We characterize, evaluate, and statistically validate the performance of different customized CNNs keeping every training subject's data separate from the validation set.

KEYWORDS:

blood smears; classification; computer-aided diagnosis; convolutional neural networks; deep learning; visualization

PMID:
30035153
PMCID:
PMC6050500
[Available on 2019-07-18]
DOI:
10.1117/1.JMI.5.3.034501

Supplemental Content

Loading ...
Support Center