Explainable Classifier Supporting Decision-making for Breast Cancer Diagnosis from Histopathological Images

This paper presents an application of semantically explainable classifier, Cumulative Fuzzy Class Membership Criterion (CFCMC), in medical domain, specifically for breast cancer detection from histopathological images. This classifier, in contrast with commonly used classifiers for image classification, is able to provide additional information about its classification results in human-friendly form. In this paper, we proposed a means for presenting the additional semantical informations that is potentionally useful for decision-making in medical domain. First, the classifier provides semantic explanation, regarding the possibility of misclassification of the test sample. Alongside with semantics, it provides visualization of similar and non-similar samples of different class. The classification performance of CFCMC is compared against three commonly used classifiers for image classification, Convolutional Neural Network (CNN), Stacked Auto-encoder (SAE) and Deep Multi-layered Perceptron. The experimental result shows that the CFCMC is not necessarily the best classifier. However, the ability to provide semantic and visual explanation of classification result allows the classifier to be applied as a supporting tool for pathologists in diagnostic of breast cancer.

[1]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[2]  Guido Bologna,et al.  Characterization of Symbolic Rules Embedded in Deep DIMLP Networks: A Challenge to Transparency of Deep Learning , 2017, J. Artif. Intell. Soft Comput. Res..

[3]  P. Rousseeuw.  Silhouettes: a graphical aid to the interpretation and validation of cluster analysis , 1987 .

[4]  Scott M. Lundberg,et al.  Explainable machine-learning predictions for the prevention of hypoxaemia during surgery , 2018, Nature Biomedical Engineering.

[5]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[6]  Nico Karssemeijer,et al.  Context-aware stacked convolutional neural networks for classification of breast carcinomas in whole-slide histopathology images , 2017, Journal of medical imaging.

[7]  Richard Staba,et al.  Multimodal data and machine learning for surgery outcome prediction in complicated cases of mesial temporal lobe epilepsy , 2015, Comput. Biol. Medicine.

[8]  Peter Sincak,et al.  Cumulative fuzzy class membership criterion decision-based classifier , 2017, 2017 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[9]  Christian Biemann,et al.  What do we need to build explainable AI systems for the medical domain? , 2017, ArXiv.

[10]  K. Borgwardt,et al.  Machine Learning in Medicine , 2015, Mach. Learn. under Resour. Constraints Vol. 3.

[11]  Fabio A. González,et al.  Automatic detection of invasive ductal carcinoma in whole slide images with convolutional neural networks , 2014, Medical Imaging.

[12]  Robert Tibshirani,et al.  Estimating the number of clusters in a data set via the gap statistic , 2000 .

[13]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[14]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[15]  Donald W. Bouldin,et al.  A Cluster Separation Measure , 1979, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[16]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[17]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.