Transfer Learning Approach for Classification of Histopathology Whole Slide Images.

Shakil Ahmed,Maheen Bakhtyar,Mesfer Alrizq,Hani Alshahrani,Junaid Baber,Abdullah Alghamdi,Asadullah Shaikh

doi:10.3390/s21165361

Abstract

The classification of whole slide images (WSIs) provides physicians with an accurate analysis of diseases and also helps them to treat patients effectively. The classification can be linked to further detailed analysis and diagnosis. Deep learning (DL) has made significant advances in the medical industry, including the use of magnetic resonance imaging (MRI) scans, computerized tomography (CT) scans, and electrocardiograms (ECGs) to detect life-threatening diseases, including heart disease, cancer, and brain tumors. However, more advancement in the field of pathology is needed, but the main hurdle causing the slow progress is the shortage of large-labeled datasets of histopathology images to train the models. The Kimia Path24 dataset was particularly created for the classification and retrieval of histopathology images. It contains 23,916 histopathology patches with 24 tissue texture classes. A transfer learning-based framework is proposed and evaluated on two famous DL models, Inception-V3 and VGG-16. To improve the productivity of Inception-V3 and VGG-16, we used their pre-trained weights and concatenated these with an image vector, which is used as input for the training of the same architecture. Experiments show that the proposed innovation improves the accuracy of both famous models. The patch-to-scan accuracy of VGG-16 is improved from 0.65 to 0.77, and for the Inception-V3, it is improved from 0.74 to 0.79.

Highlights

In the field of medical science, automatic analysis of histological images has created great convenience for doctors and scientists
The digital scan of the sample on the slide is called a whole slide image (WSI) that enables the storage of the sample digitally on the computer in the shape of a digital image
The pre-trained models were trained on very large datasets; we provided the image as input to that layer and extracted the features from the n-1 layer, that feature was concatenated with the unit normalized image

Summary

Introduction

In the field of medical science, automatic analysis of histological images has created great convenience for doctors and scientists. Experts from different fields of computing and machine learning are able to contribute to medical science due to the availability of labeled data and technology that can digitize the data used in everyday analysis. In the field of pathology, it has become technologically easy to digitally scan the sample on the slides that are used for microscopy analysis and use it for computer-aided analysis and diagnosis. The WSI can be used for detailed analysis and diagnosis by experts remotely or as a reference for future predictions. The saved WSI can be shared with experts in entirely different corners of the world for their swift analysis of the image

Methods

Results

Conclusion