4.4. Image classification using the original and augmented datasets

The original image dataset is split into two halves, as shown in Table 1. The first half (Dataset 1) was used for data augmentation. Using data augmentation methods discussed above, images of 400 infected cells were increased to 4000 cells, and images of 1000 normal cells were increased to 10,000 cells. Consequently, we created two datasets, one as the result of using spatial domain interpolation, the other as the result of using feature domain (via stacked autoencoders) interpolation, as shown in Table 2. Note that the samples for validation were randomly selected.

Figure 7. Result of interpolation using two example images. (a) Top row: two images of malaria-infected red blood cells used to generate a new image using interpolation. Middle row: 11 images obtained by interpolation using Eq. (1), where k is a weight varied between 0 and 1 with a step size of 0.1. Bottom row: 11 images obtained by interpolation in the feature domain using Eq. (2), where k is a weight varied between 0 and 1 with a step size of 0.1. (b) Similar to (a), images obtained by interpolation in the spatial (middle row) and the feature domain (bottom row).

We conducted various simulations based on the configuration shown in Table 3. For example, we used the augmented images in Dataset 3 to train the LeNet-5 convolutional neural network, and tested the original images in Dataset 2 using the trained network in order to classify the images into two categories: either infected or normal cells. Inversely, we trained the LeNet-5 using the original dataset and tested using the augmented datasets, in order to see how the trained classifier would perform on the augmented datasets.

Simulation results are shown in Figures 8 and 9.

4.3. Results of image interpolation

168 Machine Learning - Advanced Techniques and Emerging Applications

As a visual comparison of the effect of image interpolation in the spatial and feature domains,

The original image dataset is split into two halves, as shown in Table 1. The first half (Dataset 1) was used for data augmentation. Using data augmentation methods discussed above, images of 400 infected cells were increased to 4000 cells, and images of 1000 normal cells were increased to 10,000 cells. Consequently, we created two datasets, one as the result of using spatial domain interpolation, the other as the result of using feature domain (via stacked autoencoders) interpolation, as shown in Table 2. Note that the samples for validation were randomly selected.

Figure 7. Result of interpolation using two example images. (a) Top row: two images of malaria-infected red blood cells used to generate a new image using interpolation. Middle row: 11 images obtained by interpolation using Eq. (1), where k is a weight varied between 0 and 1 with a step size of 0.1. Bottom row: 11 images obtained by interpolation in the feature domain using Eq. (2), where k is a weight varied between 0 and 1 with a step size of 0.1. (b) Similar to (a), images obtained

by interpolation in the spatial (middle row) and the feature domain (bottom row).

Figure 7 shows the result of interpolating from two example red blood cell images.

4.4. Image classification using the original and augmented datasets

It can be seen in Figure 8 that training and validation using the augmented dataset provides fairly high accuracy (above 90%) when testing using the original dataset, implying the augmented data agree reasonably well statistically with the original data. Besides, feature domain interpolation seems to offer higher accuracy than spatial domain interpolation. Furthermore, the classification accuracies vary more significantly with the interpolation (mixing) coefficient k for spatial domain interpolation than for feature domain interpolation. For both interpolation


The first half (Dataset 1) was used for data augmentation. The second half (Dataset 2) was used for training and testing.

Table 1. The original image dataset is split into two halves.


Dataset 3 was obtained using the spatial domain interpolation and Dataset 4 was obtained using the feature domain (via stacked autoencoders) interpolation.

Table 2. The augmented image dataset from dataset 1 in Table 1.


Table 3. Several combinations of datasets used in training, validation and testing of the convolutional neural network.

Figure 8. Classification accuracies as a function of the interpolation coefficient k in Eqs. (1) and (2), when we trained the LeNet-5 using the augmented datasets and tested the classifier using the original dataset. The augmented datasets are dataset 3 with spatial domain interpolation and dataset 4 with feature domain interpolation, respectively.

methods, there exists an optimal coefficient such that the classification accuracy reaches its maximum.

5. Conclusions and further research

with spatial domain interpolation and dataset 4 with feature domain interpolation.

Malaria is a widespread disease that has claimed millions of lives all over the world. Automation of the diagnosis process will provide accurate diagnosis of the disease, which will benefit health-care to resource-scarce areas. We showed that the deep convolutional network based on LeNet-5 was capable of achieving very high classification accuracies for automated malaria diagnosis, by automatically learn the features from the input image data. We briefly described the workflow of classification of the red blood cell images, and discussed in details the data augmentation methods we proposed to deal with the issue with training deep convolutional neural networks with a small dataset. We then compared the classification accuracies associated with training, validating, and testing with various combinations of the original dataset and the significantly augmented datasets, which were obtained using direct interpolation in the spatial domain, as well as indirect interpolation using automatically extracted features provided by stacked autoencoders. This comparative study indicated that data augmentation in the feature domain seemed to be more robust in terms of preserving

Figure 9. Classification accuracies as a function of the interpolation coefficient k in Eqs. (1) and (2), when we trained the LeNet-5 using the original dataset (dataset 2) and tested the classifier using the augmented datasets, where dataset 3 is

Classification of Malaria-Infected Cells Using Deep Convolutional Neural Networks

http://dx.doi.org/10.5772/intechopen.72426

171

It can be seen in Figure 9 that by using the classifier trained on the original dataset, we might get very low (below 80%) classification accuracy on the input images in the augmented dataset (obtained using interpolation in the spatial domain). This highlights the importance of using data augmentation in order to attain a more balanced estimation of the generalization ability of the classifier. This generalization ability seems to depend heavily on the varying new image samples that were interpolated using a different interpolation (mixing) coefficient from the original dataset (e.g., when k ¼ 0:7, the accuracy can reach about 99%). Figure 9 also shows that Dataset 4 (feature domain interpolation) seems to be a less challenging dataset than Dataset 3 (spatial domain interpolation), in that all accuracies are above 95%, possibly suggesting that mixing images using their features extracted by the stacked autoencoder would generate less diverse images than directly mixing images in the spatial domain.

Figure 9. Classification accuracies as a function of the interpolation coefficient k in Eqs. (1) and (2), when we trained the LeNet-5 using the original dataset (dataset 2) and tested the classifier using the augmented datasets, where dataset 3 is with spatial domain interpolation and dataset 4 with feature domain interpolation.
