Explainable AI and susceptibility to adversarial attacks: a case study in classification of breast ultrasound images

Ultrasound is a non-invasive imaging modality that can be conveniently used to classify suspicious breast nodules and potentially detect the onset of breast cancer. Recently, Convolutional Neural Networks (CNN) techniques have shown promising results in classifying ultrasound images of the breast into benign or malignant. However, CNN inference acts as a black-box model, and as such, its decision-making is not interpretable. Therefore, increasing effort has been dedicated to explaining this process, most notably through GRAD-CAM and other techniques that provide visual explanations into inner workings of CNNs. In addition to interpretation, these methods provide clinically important information, such as identifying the location for biopsy or treatment. In this work, we analyze how adversarial assaults that are practically undetectable may be devised to alter these importance maps dramatically. Furthermore, we will show that this change in the importance maps can come with or without altering the classification result, rendering them even harder to detect. As such, care must be taken when using these importance maps to shed light on the inner workings of deep learning. Finally, we utilize Multi-Task Learning (MTL) and propose a new network based on ResNet-50 to improve the classification accuracies. Our sensitivity and specificity is comparable to the state of the art results.

[1]  Bolei Zhou,et al.  Learning Deep Features for Discriminative Localization , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Amir Asif,et al.  Fast Multi-Focus Ultrasound Image Recovery Using Generative Adversarial Networks , 2020, IEEE Transactions on Computational Imaging.

[3]  Paulo Sergio Rodrigues,et al.  Breast Ultrasound Image , 2017 .

[4]  A. Boss,et al.  Classification of breast cancer in ultrasound imaging using a generic deep learning analysis software: a pilot study. , 2017, The British journal of radiology.

[5]  Abubakar Abid,et al.  Interpretation of Neural Networks is Fragile , 2017, AAAI.

[6]  Samy Bengio,et al.  Adversarial examples in the physical world , 2016, ICLR.

[7]  Purang Abolmaesumi,et al.  Ultrasound RF Time Series for Classification of Breast Lesions , 2015, IEEE Transactions on Medical Imaging.

[8]  Zhang Yi,et al.  Automated diagnosis of breast ultrasonography images using deep neural networks , 2019, Medical Image Anal..

[9]  Hassan Rivaz,et al.  Low Variance Estimation of Backscatter Quantitative Ultrasound Parameters Using Dynamic Programming , 2018, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control.

[10]  Michael Galperin,et al.  Breast mass classification in sonography with transfer learning using a deep convolutional neural network and color conversion , 2019, Medical physics.

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

[12]  Wagner Coelho A. Pereira,et al.  Analysis of Co-Occurrence Texture Statistics as a Function of Gray-Level Quantization for Classifying Breast Ultrasound , 2012, IEEE Transactions on Medical Imaging.

[13]  Hassan Rivaz,et al.  Ultrasound Scatterer Density Classification Using Convolutional Neural Networks and Patch Statistics , 2021, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control.

[14]  Jeon-Hor Chen,et al.  Robust Texture Analysis Using Multi-Resolution Gray-Scale Invariant Features for Breast Sonographic Tumor Diagnosis , 2013, IEEE Transactions on Medical Imaging.

[15]  Kok-Swee Sim,et al.  Convolutional neural network improvement for breast cancer classification , 2019, Expert Syst. Appl..

[16]  Abhishek Das,et al.  Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization , 2016, 2017 IEEE International Conference on Computer Vision (ICCV).

[17]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[18]  Walid Al-Dhabyani,et al.  Dataset of breast ultrasound images , 2019, Data in brief.

[19]  Ling Zhang,et al.  Automated breast cancer detection and classification using ultrasound images: A survey , 2015, Pattern Recognit..

[20]  Heng-Da Cheng,et al.  Breast Ultrasound Image Classification Based on Multiple-Instance Learning , 2012, Journal of Digital Imaging.

[21]  Sebastian Ruder,et al.  An Overview of Multi-Task Learning in Deep Neural Networks , 2017, ArXiv.

[22]  Ravinder Agarwal,et al.  Deep feature extraction and classification of breast ultrasound images , 2020, Multimedia Tools and Applications.

[23]  Hassan Rivaz,et al.  Two-stage ultrasound image segmentation using U-Net and test time augmentation , 2020, International Journal of Computer Assisted Radiology and Surgery.

[24]  Jonathon Shlens,et al.  Explaining and Harnessing Adversarial Examples , 2014, ICLR.

[25]  Il Dong Yun,et al.  Joint Weakly and Semi-Supervised Deep Learning for Localization and Classification of Masses in Breast Ultrasound Images , 2017, IEEE Transactions on Medical Imaging.

[26]  Reyer Zwiggelaar,et al.  Automated Breast Ultrasound Lesions Detection Using Convolutional Neural Networks , 2018, IEEE Journal of Biomedical and Health Informatics.

[27]  Leslie N. Smith,et al.  Cyclical Learning Rates for Training Neural Networks , 2015, 2017 IEEE Winter Conference on Applications of Computer Vision (WACV).

[28]  Seokmin Han,et al.  A deep learning framework for supporting the classification of breast lesions in ultrasound images , 2017, Physics in medicine and biology.

[29]  Hassan Rivaz,et al.  Deep classification of breast cancer in ultrasound images: more classes, better results with multi-task learning , 2021, Medical Imaging.

[30]  Sebastian Thrun,et al.  Dermatologist-level classification of skin cancer with deep neural networks , 2017, Nature.