HNAS-reg: hierarchical neural architecture search for deformable medical image registration
This work addresses the need for efficient and accurate deformable medical image registration, which is incremental as it builds on existing NAS and registration techniques.
The paper tackled the problem of suboptimal manually designed network architectures for medical image registration by proposing HNAS-Reg, a hierarchical neural architecture search framework that identified optimal architectures, resulting in improved registration accuracy and reduced model size compared to state-of-the-art methods.
Convolutional neural networks (CNNs) have been widely used to build deep learning models for medical image registration, but manually designed network architectures are not necessarily optimal. This paper presents a hierarchical NAS framework (HNAS-Reg), consisting of both convolutional operation search and network topology search, to identify the optimal network architecture for deformable medical image registration. To mitigate the computational overhead and memory constraints, a partial channel strategy is utilized without losing optimization quality. Experiments on three datasets, consisting of 636 T1-weighted magnetic resonance images (MRIs), have demonstrated that the proposal method can build a deep learning model with improved image registration accuracy and reduced model size, compared with state-of-the-art image registration approaches, including one representative traditional approach and two unsupervised learning-based approaches.