RT-DNAS: Real-time Constrained Differentiable Neural Architecture Search for 3D Cardiac Cine MRI Segmentation
This work addresses the need for fast and accurate segmentation in real-time MRI-guided cardiac interventions, representing an incremental improvement by adapting NAS to handle specific real-time constraints in a medical domain.
The paper tackled the problem of real-time 3D cardiac cine MRI segmentation by developing RT-DNAS, a differentiable neural architecture search method that directly incorporates non-differentiable real-time constraints, resulting in architectures that achieve better accuracy while meeting latency and throughput requirements compared to state-of-the-art manual and automated designs.
Accurately segmenting temporal frames of cine magnetic resonance imaging (MRI) is a crucial step in various real-time MRI guided cardiac interventions. To achieve fast and accurate visual assistance, there are strict requirements on the maximum latency and minimum throughput of the segmentation framework. State-of-the-art neural networks on this task are mostly hand-crafted to satisfy these constraints while achieving high accuracy. On the other hand, while existing literature have demonstrated the power of neural architecture search (NAS) in automatically identifying the best neural architectures for various medical applications, they are mostly guided by accuracy, sometimes with computation complexity, and the importance of real-time constraints are overlooked. A major challenge is that such constraints are non-differentiable and are thus not compatible with the widely used differentiable NAS frameworks. In this paper, we present a strategy that directly handles real-time constraints in a differentiable NAS framework named RT-DNAS. Experiments on extended 2017 MICCAI ACDC dataset show that compared with state-of-the-art manually and automatically designed architectures, RT-DNAS is able to identify ones with better accuracy while satisfying the real-time constraints.