RT-DNAS: Real-time Constrained Differentiable Neural Architecture Search for 3D Cardiac Cine MRI Segmentation

TL;DR

This paper introduces RT-DNAS, a differentiable neural architecture search method that explicitly incorporates real-time constraints, resulting in more accurate and faster segmentation models for 3D cardiac cine MRI.

Contribution

The paper proposes a novel differentiable NAS framework that directly handles non-differentiable real-time constraints, improving the design of neural networks for real-time medical imaging tasks.

Findings

RT-DNAS outperforms existing architectures in accuracy.

RT-DNAS satisfies strict real-time constraints.

Identified architectures are more efficient for 3D cardiac MRI segmentation.

Abstract

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.