Medical Image Classification Through Deep Learning

The authors investigate the problem of image classification. Earlier, the task of image classification is accomplished by traditional machine learning techniques and other shallower neural network models. Later with the evolution of deeper networks, convolutional neural networks have gained without importance due to its outstanding accuracy in various domains. Unlike in real-world datasets for performing the classification of various images under different categories, the job of biomedical image classification of chest X-rays is quite tedious due to overlapping characteristics of X-ray images. The objective of this paper is to classify the images of chest X-rays and predict the pneumonia traces in lungs. Inception V3 model, with transfer learning is applied on this medical dataset. The model is implemented in Keras as front-end library with tensor flow framework. The training on this dataset to generate a custom model file on GTX 1070 video card consumed 30 min yielding 98% training accuracy and 99% validation accuracy.

[1]  Toby P. Breckon,et al.  Transfer learning using convolutional neural networks for object classification within X-ray baggage security imagery , 2016, 2016 IEEE International Conference on Image Processing (ICIP).

[2]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Zhiping Lin,et al.  Deep CNNs for microscopic image classification by exploiting transfer learning and feature concatenation , 2018, 2018 IEEE International Symposium on Circuits and Systems (ISCAS).

[4]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Sergey Ioffe,et al.  Rethinking the Inception Architecture for Computer Vision , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).