Fully-automated functional region annotation of liver via a 2.5D class-aware deep neural network with spatial adaptation

Tian, Yinli; Xue, Fei; Lambo, R.; He, Jiahui; An, Cheolhong; Xie, Yaoqin; Cao, Hailin; Qin, Wenjian

doi:10.1016/j.cmpb.2020.105818

Cited by 11 publications

(5 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several opensource neural networks exist that can perform automatic segmentation tasks, such as U-net [78] . A complete automated liver segmentation problem is not fully solved yet, although the situation has changed since 2009 [79] thanks to deep neural networks [80,81] . The paper by Ibtehaz et al [78] is different from the other reviewed articles, as it describes U-net's use in automatic volumetry of liver graft.…”

Section: Discussion (Perfect)mentioning

confidence: 99%

Preoperative planning in paediatric liver tumour surgery - a literature review

Kościuszko¹,

Niemirycz-Makurat²,

Anzelewicz³

et al. 2021

View full text Add to dashboard Cite

Radiological assessment is evolving rapidly, including in paediatric liver tumour surgery. There are advancements in the fields of conventional radiology, 3D imaging and preoperative planning. This article presents the current research in paediatric liver imaging for liver tumour surgery and shows the results of a systematic search in computer-aided liver surgery in children. Sixteen original papers were found. We summarise the progress made and offer the directions in which further research could go. Computer-assisted surgery is a promising field for research and clinical application.

show abstract

Section: Discussion (Perfect)mentioning

confidence: 99%

Preoperative planning in paediatric liver tumour surgery - a literature review

Kościuszko¹,

Niemirycz-Makurat²,

Anzelewicz³

et al. 2021

View full text Add to dashboard Cite

show abstract

“…DDRN, a four‐module dilated deep residual network, was introduced by Mourya et al [34]. An automated method for functional region annotation using the ResU‐Net architecture was developed by Tian et al [35]. Hong et al [36] proposed an unsupervised domain adaptation approach for liver segmentation.…”

Section: Related Workmentioning

confidence: 99%

Transformer Skip‐Fusion Based SwinUNet for Liver Segmentation From CT Images

Kumar,

Vinod Kumar

2024

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Liver segmentation is a crucial step in medical image analysis and is essential for diagnosing and treating liver diseases. However, manual segmentation is time‐consuming and subject to variability among observers. To address these challenges, a novel liver segmentation approach, SwinUNet with transformer skip‐fusion is proposed. This method harnesses the Swin Transformer's capacity to model long‐range dependencies efficiently, the U‐Net's ability to preserve fine spatial details, and the transformer skip‐fusion's effectiveness in enabling the decoder to learn intricate features from encoder feature maps. In experiments using the 3DIRCADb and CHAOS datasets, this technique outperformed traditional CNN‐based methods, achieving a mean DICE coefficient of 0.988% and a mean Jaccard coefficient of 0.973% by aggregating the results obtained from each dataset, signifying outstanding agreement with ground truth. This remarkable accuracy in liver segmentation holds significant promise for improving liver disease diagnosis and enhancing healthcare outcomes for patients with liver conditions.

show abstract

“…Because 2D DL-based segmentation models are less accurate while 3D ones are accurate but memory and computationally expensive, Tian et al (2021) automatically annotate functional regions of the liver using a 2.5D class-aware deep neural network (DNN) with spatial adaptation. This framework is based on analyzing abdominal images using a ResU-Net model, which (i) adequately selects a pile of adjacent CT slices as input; (ii) generates the center slice; and (iii) automatically annotates the liver functional regions.…”

Section: 5d Inputmentioning

confidence: 99%