Improving 3D convolutional neural network comprehensibility via interactive visualization of relevance maps: evaluation in Alzheimer’s disease

@article{Dyrba2020Improving3C,
  title={Improving 3D convolutional neural network comprehensibility via interactive visualization of relevance maps: evaluation in Alzheimer’s disease},
  author={Martin Dyrba and Moritz Hanzig and Slawek Altenstein and Sebastian Bader and Tommaso Ballarini and Frederic Brosseron and Katharina Buerger and Daniel Cantr{\'e} and Peter Dechent and Laura Dobisch and Emrah D{\"u}zel and Michael Ewers and Klaus Fliessbach and Wenzel Glanz and John-Dylan Haynes and Michael T. Heneka and Daniel Janowitz and Deniz Baris Keles and Ingo Kilimann and Christoph Laske and Franziska Maier and Coraline D. Metzger and Matthias H. J. Munk and Robert Perneczky and Oliver Peters and Lukas Preis and Josef Priller and Boris Stephan Rauchmann and Nina Roy and Klaus Scheffler and Anja Schneider and Bj{\"o}rn H. Schott and Annika Spottke and Eike J. Spruth and Marc-Andr{\'e} Weber and Birgit B. Ertl-Wagner and Michael Wagner and Jens Wiltfang and Frank Jessen and Stefan J. Teipel},
  journal={Alzheimer's Research \& Therapy},
  year={2020},
  volume={13}
}
Background Although convolutional neural networks (CNNs) achieve high diagnostic accuracy for detecting Alzheimer’s disease (AD) dementia based on magnetic resonance imaging (MRI) scans, they are not yet applied in clinical routine. One important reason for this is a lack of model comprehensibility. Recently developed visualization methods for deriving CNN relevance maps may help to fill this gap as they allow the visualization of key input image features that drive the decision of the model… 

A review of the application of three-dimensional convolutional neural networks for the diagnosis of Alzheimer’s disease using neuroimaging

The aim of this article is to present the current state of the art in the diagnosis of AD using 3D CNN models and neuroimaging modalities, focusing on the 3DCNN architectures and classification methods used, and to highlight potential future research topics.

Diagnostic accuracy study of automated stratification of Alzheimer’s disease and mild cognitive impairment via deep learning based on MRI

Using a deep CNN and iterated RF architecture, it is shown that brain image stratification is a promising means for evaluating AD, and examining the underlying etiology of the disease, by applying computer and medical images to achieve the early auxiliary diagnosis of AD.

Deep Learning-Based Diagnosis of Alzheimer’s Disease

The current state-of-the-art in AD diagnosis using deep learning is reviewed, including the most recent trends and findings using a thorough literature review.

Automated Evaluation of Conventional Clock-Drawing Test Using Deep Neural Network: Potential as a Mass Screening Tool to Detect Individuals With Cognitive Decline

The feasibility of implementing conventional CDT to be automatically evaluated by DNN with a fair performance in a larger scale than ever is demonstrated, suggesting its potential as a mass screening test for ruling-in or ruling-out those with executive dysfunction or with probable dementia.

Structural and Functional MRI Data Differentially Predict Chronological Age and Behavioral Memory Performance

Abstract Human cognitive abilities decline with increasing chronological age, with decreased explicit memory performance being most strongly affected. However, some older adults show “successful

Feature visualization for convolutional neural network models trained on neuroimaging data

This study shows, for the first time, results using feature visualization of neuroimaging CNNs for explainability, and reveals the learned concepts of the artificial lesions, including their shapes, but remain hard to interpret for abstract features in the sex classi fication task.

Individualized Gaussian Process-based Prediction of Memory Performance and Biomarker Status in Ageing and Alzheimer’s disease

The highest accuracy for memory performance prediction was obtained for a combination of neuroimaging markers, demographics, genetic information (ApoE4) and CSF-biomarkers explaining 57% of outcome variance in out of sample predictions.

References

SHOWING 1-10 OF 62 REFERENCES

On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation

This work proposes a general solution to the problem of understanding classification decisions by pixel-wise decomposition of nonlinear classifiers by introducing a methodology that allows to visualize the contributions of single pixels to predictions for kernel-based classifiers over Bag of Words features and for multilayered neural networks.

Methods for interpreting and understanding deep neural networks

Testing the robustness of attribution methods for convolutional neural networks in MRI-based Alzheimer's disease classification

It is shown that visual comparison is not sufficient and that some widely used attribution methods produce highly inconsistent outcomes, including gradient*input, guided backpropagation, layer-wise relevance propagation and occlusion.

Convolutional Neural Networks for Classification of Alzheimer's Disease: Overview and Reproducible Evaluation

Advancing research diagnostic criteria for Alzheimer's disease: the IWG-2 criteria

Multimodal imaging in Alzheimer's disease: validity and usefulness for early detection

Design and first baseline data of the DZNE multicenter observational study on predementia Alzheimer’s disease (DELCODE)

The initial baseline data for DELCODE support the approach of using SCD in patients recruited through memory clinics as an enrichment strategy for late-stage preclinical AD.

Visualizing evidence for Alzheimer's disease in deep neural networks trained on structural MRI data

It is concluded that LRP provides a powerful tool for assisting clinicians in finding evidence for AD (and potentially other diseases) in structural MRI data.

Automated Anatomical Labeling of Activations in SPM Using a Macroscopic Anatomical Parcellation of the MNI MRI Single-Subject Brain

An anatomical parcellation of the spatially normalized single-subject high-resolution T1 volume provided by the Montreal Neurological Institute was performed and it is believed that this tool is an improvement for the macroscopical labeling of activated area compared to labeling assessed using the Talairach atlas brain.
...