UXNet: Searching Multi-level Feature Aggregation for 3D Medical Image Segmentation

09/16/2020
by   Yuanfeng Ji, et al.
0

Aggregating multi-level feature representation plays a critical role in achieving robust volumetric medical image segmentation, which is important for the auxiliary diagnosis and treatment. Unlike the recent neural architecture search (NAS) methods that typically searched the optimal operators in each network layer, but missed a good strategy to search for feature aggregations, this paper proposes a novel NAS method for 3D medical image segmentation, named UXNet, which searches both the scale-wise feature aggregation strategies as well as the block-wise operators in the encoder-decoder network. UXNet has several appealing benefits. (1) It significantly improves flexibility of the classical UNet architecture, which only aggregates feature representations of encoder and decoder in equivalent resolution. (2) A continuous relaxation of UXNet is carefully designed, enabling its searching scheme performed in an efficient differentiable manner. (3) Extensive experiments demonstrate the effectiveness of UXNet compared with recent NAS methods for medical image segmentation. The architecture discovered by UXNet outperforms existing state-of-the-art models in terms of Dice on several public 3D medical image segmentation benchmarks, especially for the boundary locations and tiny tissues. The searching computational complexity of UXNet is cheap, enabling to search a network with the best performance less than 1.5 days on two TitanXP GPUs.

READ FULL TEXT
research
06/13/2019

Scalable Neural Architecture Search for 3D Medical Image Segmentation

In this paper, a neural architecture search (NAS) framework is proposed ...
research
02/23/2022

Mixed-Block Neural Architecture Search for Medical Image Segmentation

Deep Neural Networks (DNNs) have the potential for making various clinic...
research
09/09/2023

SSHNN: Semi-Supervised Hybrid NAS Network for Echocardiographic Image Segmentation

Accurate medical image segmentation especially for echocardiographic ima...
research
11/15/2022

Feedback Chain Network For Hippocampus Segmentation

The hippocampus plays a vital role in the diagnosis and treatment of man...
research
03/29/2021

DiNTS: Differentiable Neural Network Topology Search for 3D Medical Image Segmentation

Recently, neural architecture search (NAS) has been applied to automatic...
research
05/07/2019

Efficient Neural Architecture Search on Low-Dimensional Data for OCT Image Segmentation

Typically, deep learning architectures are handcrafted for their respect...
research
12/21/2021

Leveraging Image Complexity in Macro-Level Neural Network Design for Medical Image Segmentation

Recent progress in encoder-decoder neural network architecture design ha...

Please sign up or login with your details

Forgot password? Click here to reset