Clipped DeepControl: deep neural network two-dimensional pulse design with an amplitude constraint layer
This work addresses a safety and reliability issue in MRI pulse design for medical imaging, but it is incremental as it builds on prior deep learning methods.
The authors tackled the problem of amplitude overshoots in deep learning-designed MRI pulses by extending a convolutional neural network with a custom clipping layer, which completely eliminated overshoots while maintaining the ability to compensate for field inhomogeneities.
Advanced radio-frequency pulse design used in magnetic resonance imaging has recently been demonstrated with deep learning of (convolutional) neural networks and reinforcement learning. For two-dimensionally selective radio-frequency pulses, the (convolutional) neural network pulse prediction time (few milliseconds) was in comparison more than three orders of magnitude faster than the conventional optimal control computation. The network pulses were from the supervised training capable of compensating scan-subject dependent inhomogeneities of B0 and B+1 fields. Unfortunately, the network presented with a non-negligible percentage of pulse amplitude overshoots in the test subset, despite the optimal control pulses used in training were fully constrained. Here, we have extended the convolutional neural network with a custom-made clipping layer that completely eliminates the risk of pulse amplitude overshoots, while preserving the ability to compensate the inhomogeneous field conditions.