What is the Optimal Attribution Method for Explainable Ophthalmic Disease Classification?

Singh, Amitojdeep; Sengupta, Sourya; Balaji, J. Jothi; Mohammed, Abdul Rasheed; Faruq, Ibrahim; Jayakumar, Varadharajan; Zelek, John; Lakshminarayanan, Vasudevan

doi:10.1007/978-3-030-63419-3_3

Cited by 19 publications

(26 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In contrast to this evidence by us and others [5,69] in favor of Guided Backprob in a clinical setting, Guided Backprob has been shown to be insensitive to the object classes in ImageNet [63,55]. This likely happens because the algorithm exploits local connections in convolutional layers, which extract a series of hierarchical feature representations from a given image, and the final dense layers, where class label assignments are made, have less impact on saliency maps [55].…”

Section: Discussionmentioning

confidence: 95%

Clinical Validation of Saliency Maps for Understanding Deep Neural Networks in Ophthalmology

Ayhan

Kümmerle

Kühlewein

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep neural networks (DNNs) have achieved physician-level accuracy on many imaging-based medical diagnostic tasks, for example classification of retinal images in ophthalmology. However, their decision mechanisms are often considered impenetrable leading to a lack of trust by clinicians and patients. To alleviate this issue, a range of explanation methods have been proposed to expose the inner workings of DNNs leading to their decisions. For imaging-based tasks, this is often achieved via saliency maps. The quality of these maps are typically evaluated via perturbation analysis without experts involved. To facilitate the adoption and success of such automated systems, however, it is crucial to validate saliency maps against clinicians. In this study, we used two different network architectures and developed ensembles of DNNs to detect diabetic retinopathy and neovascular age-related macular degeneration from retinal fundus images and optical coherence tomography scans, respectively. We used a variety of explanation methods and obtained a comprehensive set of saliency maps for explaining the ensemble-based diagnostic decisions. Then, we systematically validated saliency maps against clinicians through two main analyses --- a direct comparison of saliency maps with the expert annotations of disease-specific pathologies and perturbation analyses using also expert annotations as saliency maps. We found the choice of DNN architecture and explanation method to significantly influence the quality of saliency maps. Guided Backprop showed consistently good performance across disease scenarios and DNN architectures, suggesting that it provides a suitable starting point for explaining the decisions of DNNs on retinal images.

show abstract

Section: Discussionmentioning

confidence: 95%

Clinical Validation of Saliency Maps for Understanding Deep Neural Networks in Ophthalmology

Ayhan

Kümmerle

Kühlewein

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Therefore, their predictions lack clinical interpretation, despite their high accuracy. This black box nature of DCNNs is the major problem that makes them unsuitable for clinical application [ 86 , 152 , 153 ] and has made the topic of eXplainable AI (XAI) of major importance [ 153 ]. Recently, visualization techniques such as gradient-based XAI have been widely used for evaluating networks.…”

Section: Resultsmentioning

confidence: 99%

Automated Detection and Diagnosis of Diabetic Retinopathy: A Comprehensive Survey

et al. 2021

Self Cite

View full text Add to dashboard Cite

Diabetic Retinopathy (DR) is a leading cause of vision loss in the world. In the past few years, artificial intelligence (AI) based approaches have been used to detect and grade DR. Early detection enables appropriate treatment and thus prevents vision loss. For this purpose, both fundus and optical coherence tomography (OCT) images are used to image the retina. Next, Deep-learning (DL)-/machine-learning (ML)-based approaches make it possible to extract features from the images and to detect the presence of DR, grade its severity and segment associated lesions. This review covers the literature dealing with AI approaches to DR such as ML and DL in classification and segmentation that have been published in the open literature within six years (2016–2021). In addition, a comprehensive list of available DR datasets is reported. This list was constructed using both the PICO (P-Patient, I-Intervention, C-Control, O-Outcome) and Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) 2009 search strategies. We summarize a total of 114 published articles which conformed to the scope of the review. In addition, a list of 43 major datasets is presented.

show abstract

“…However, other areas of medicine, for example, ophthalmology, have shown that certain classifiers approach clinician-level performance. Of further importance is the development of explainable AI methods that have been applied to ophthalmology where correlations are made between areas of the image that the clinician uses to make decisions and the ones used by the algorithms to arrive at the result (i.e., the portions of the image that most heavily weigh the neural connections) [83,[225][226][227].…”

Section: Discussionmentioning

confidence: 99%

“…Figure 8 presents a very general timeline of the evolution of AI and the principal relevant facts related to the field of medicine. In addition, all applications of AI to medicine and health are not covered, e.g., ophthalmology, where AI has had tremendous success (see [82][83][84][85][86][87]). It shows also the relation of the initial concepts and their evolution in machine and deep learning.…”

Section: Machine Learning and Deep Learningmentioning

confidence: 99%

MR Images, Brain Lesions, and Deep Learning

2021

Self Cite

View full text Add to dashboard Cite

Medical brain image analysis is a necessary step in computer-assisted/computer-aided diagnosis (CAD) systems. Advancements in both hardware and software in the past few years have led to improved segmentation and classification of various diseases. In the present work, we review the published literature on systems and algorithms that allow for classification, identification, and detection of white matter hyperintensities (WMHs) of brain magnetic resonance (MR) images, specifically in cases of ischemic stroke and demyelinating diseases. For the selection criteria, we used bibliometric networks. Of a total of 140 documents, we selected 38 articles that deal with the main objectives of this study. Based on the analysis and discussion of the revised documents, there is constant growth in the research and development of new deep learning models to achieve the highest accuracy and reliability of the segmentation of ischemic and demyelinating lesions. Models with good performance metrics (e.g., Dice similarity coefficient, DSC: 0.99) were found; however, there is little practical application due to the use of small datasets and a lack of reproducibility. Therefore, the main conclusion is that there should be multidisciplinary research groups to overcome the gap between CAD developments and their deployment in the clinical environment.

show abstract

What is the Optimal Attribution Method for Explainable Ophthalmic Disease Classification?

Cited by 19 publications

References 24 publications

Clinical Validation of Saliency Maps for Understanding Deep Neural Networks in Ophthalmology

Clinical Validation of Saliency Maps for Understanding Deep Neural Networks in Ophthalmology

Automated Detection and Diagnosis of Diabetic Retinopathy: A Comprehensive Survey

MR Images, Brain Lesions, and Deep Learning

Contact Info

Product

Resources

About