Scalable, Axiomatic Explanations of Deep Alzheimer's Diagnosis from Heterogeneous Data
Authors:
Sebastian Pölsterl,
Christina Aigner,
Christian Wachinger
Abstract:
Deep Neural Networks (DNNs) have an enormous potential to learn from complex biomedical data. In particular, DNNs have been used to seamlessly fuse heterogeneous information from neuroanatomy, genetics, biomarkers, and neuropsychological tests for highly accurate Alzheimer's disease diagnosis. On the other hand, their black-box nature is still a barrier for the adoption of such a system in the cli…
▽ More
Deep Neural Networks (DNNs) have an enormous potential to learn from complex biomedical data. In particular, DNNs have been used to seamlessly fuse heterogeneous information from neuroanatomy, genetics, biomarkers, and neuropsychological tests for highly accurate Alzheimer's disease diagnosis. On the other hand, their black-box nature is still a barrier for the adoption of such a system in the clinic, where interpretability is absolutely essential. We propose Shapley Value Explanation of Heterogeneous Neural Networks (SVEHNN) for explaining the Alzheimer's diagnosis made by a DNN from the 3D point cloud of the neuroanatomy and tabular biomarkers. Our explanations are based on the Shapley value, which is the unique method that satisfies all fundamental axioms for local explanations previously established in the literature. Thus, SVEHNN has many desirable characteristics that previous work on interpretability for medical decision making is lacking. To avoid the exponential time complexity of the Shapley value, we propose to transform a given DNN into a Lightweight Probabilistic Deep Network without re-training, thus achieving a complexity only quadratic in the number of features. In our experiments on synthetic and real data, we show that we can closely approximate the exact Shapley value with a dramatically reduced runtime and can reveal the hidden knowledge the network has learned from the data.
△ Less
Submitted 30 September, 2021; v1 submitted 13 July, 2021;
originally announced July 2021.
DeepControl: 2D RF pulses facilitating $B_1^+$ inhomogeneity and $B_0$ off-resonance compensation in vivo at 7T
Authors:
Mads Sloth Vinding,
Christoph Stefan Aigner,
Sebastian Schmitter,
Torben Ellegaard Lund
Abstract:
Purpose: Rapid 2D RF pulse design with subject specific $B_1^+$ inhomogeneity and $B_0$ off-resonance compensation at 7 T predicted from convolutional neural networks is presented.
Methods: The convolution neural network was trained on half a million single-channel transmit, 2D RF pulses optimized with an optimal control method using artificial 2D targets, $B_1^+$ and $B_0$ maps. Predicted pulse…
▽ More
Purpose: Rapid 2D RF pulse design with subject specific $B_1^+$ inhomogeneity and $B_0$ off-resonance compensation at 7 T predicted from convolutional neural networks is presented.
Methods: The convolution neural network was trained on half a million single-channel transmit, 2D RF pulses optimized with an optimal control method using artificial 2D targets, $B_1^+$ and $B_0$ maps. Predicted pulses were tested in a phantom and in vivo at 7 T with measured $B_1^+$ and $B_0$ maps from a high-resolution GRE sequence.
Results: Pulse prediction by the trained convolutional neural network was done on the fly during the MR session in approximately 9 ms for multiple hand drawn ROIs and the measured $B_1^+$ and $B_0$ maps. Compensation of $B_1^+$ inhomogeneity and $B_0$ off-resonances has been confirmed in the phantom and in vivo experiments. The reconstructed image data agrees well with the simulations using the acquired $B_1^+$ and $B_0$ maps and the 2D RF pulse predicted by the convolutional neural networks is as good as the conventional RF pulse obtained by optimal control.
Conclusion: The proposed convolutional neural network based 2D RF pulse design method predicts 2D RF pulses with an excellent excitation pattern and compensated $B_1^+$ and $B_0$ variations at 7 T. The rapid 2D RF pulse prediction (9 ms) enables subject-specific high-quality 2D RF pulses without the need to run lengthy optimizations.
△ Less
Submitted 25 September, 2020;
originally announced September 2020.
