Abstract
Breast cancer is one of the most prevalent types of cancer today in women. The main avenue of diagnosis is through manual examination of histopathology tissue slides. Such a process is often subjective and error-ridden, suffering from both inter- and intraobserver variability. Our objective is to develop an automatic algorithm for analysing histopathology slides free of human subjectivity. Here, we calculate the fractal dimension of images of numerous breast cancer slides, at magnifications of 40×, 100×, 200× and 400×. Using machine learning, specifically, the support vector machine (SVM) method, the F1 score for classification accuracy of the 40× slides was found to be 0.979. Multiclass classification on the 40× slides yielded an accuracy of 0.556. A reduction of the size and scope of the SVM training set gave an average F1 score of 0.964. Taken together, these results show great promise in the use of fractal dimension to predict tumour malignancy.
Highlights
Exponential increase in technological capability owing to the scope of information technology applications over the past few decades has revolutionized the way modern society functions, especially in terms of communication
We focus on further testing of fractal dimension as a viable image feature that can lead to a high level of confidence in the resultant classification
The y-axis corresponds to fractal dimension, and the x-axis corresponds to the arbitrary image numbering
Summary
Exponential increase in technological capability owing to the scope of information technology applications over the past few decades has revolutionized the way modern society functions, especially in terms of communication. The diagnosis of cancer is one area of medicine that is. There have been many attempts to automate the diagnosis of cancer from histology slides. Common to all is a three-step process: (i) preprocessing of the image, (ii) extraction of relevant features, and (iii) diagnosis from those features [6]. The goal here is to select features of an image that may be amenable to quantification and subsequent computational analysis, while at the same time being good measures of cancer severity. The hope is to correlate an image feature with either a diagnostic or a prognostic indicator, including, but not limited to, tumour malignancy and a related survival rate
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
