Attentive Symmetric Autoencoder for Brain MRI Segmentation

Huang, Junjia; Li, Haofeng; Li, Guanbin; Wan, Xiang

doi:10.1007/978-3-031-16443-9_20

Cited by 10 publications

(8 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, sensitivity and accuracy results are also compromised when using complementary modalities, but specificity seems insensitive to MRI modality selection. Compared to the baseline results from the classic U‐Net, both Attention U‐Net and Cascaded U‐Net models obtain higher overall Dice coefficients for all three segmentation targets, which are in line with the state‐of‐the‐art segmentation performance reported in the recent BraTS Challenges 37,59 . The sensitivity, specificity, and accuracy results from these two models are comparable to the classic U‐Net baseline results.…”

Section: Resultssupporting

confidence: 79%

“…Compared to the baseline results from the classic U-Net, both Attention U-Net and Cascaded U-Net models obtain higher overall Dice coefficients for all three segmentation targets, which are in line with the state-of -the-art segmentation performance reported in the recent BraTS Challenges. 37,59 The sensitivity, specificity, and accuracy results from these two models are comparable to the classic U-Net baseline results. F I G U R E 7 (a)-(c) presents the 𝜇 k t as a function of t for ET, TC, and WT segmentation, respectively.…”

Section: Figure 4a Provides An Example Of Four Image Flowsmentioning

confidence: 78%

See 1 more Smart Citation

A neural ordinary differential equation model for visualizing deep neural network behaviors in multi‐parametric MRI‐based glioma segmentation

Yang

et al. 2023

Medical Physics

View full text Add to dashboard Cite

PurposeTo develop a neural ordinary differential equation (ODE) model for visualizing deep neural network behavior during multi‐parametric MRI‐based glioma segmentation as a method to enhance deep learning explainability.MethodsBy hypothesizing that deep feature extraction can be modeled as a spatiotemporally continuous process, we implemented a novel deep learning model, Neural ODE, in which deep feature extraction was governed by an ODE parameterized by a neural network. The dynamics of (1) MR images after interactions with the deep neural network and (2) segmentation formation can thus be visualized after solving the ODE. An accumulative contribution curve (ACC) was designed to quantitatively evaluate each MR image's utilization by the deep neural network toward the final segmentation results.The proposed Neural ODE model was demonstrated using 369 glioma patients with a 4‐modality multi‐parametric MRI protocol: T1, contrast‐enhanced T1 (T1‐Ce), T2, and FLAIR. Three Neural ODE models were trained to segment enhancing tumor (ET), tumor core (TC), and whole tumor (WT), respectively. The key MRI modalities with significant utilization by deep neural networks were identified based on ACC analysis. Segmentation results by deep neural networks using only the key MRI modalities were compared to those using all four MRI modalities in terms of Dice coefficient, accuracy, sensitivity, and specificity.ResultsAll Neural ODE models successfully illustrated image dynamics as expected. ACC analysis identified T1‐Ce as the only key modality in ET and TC segmentations, while both FLAIR and T2 were key modalities in WT segmentation. Compared to the U‐Net results using all four MRI modalities, the Dice coefficient of ET (0.784→0.775), TC (0.760→0.758), and WT (0.841→0.837) using the key modalities only had minimal differences without significance. Accuracy, sensitivity, and specificity results demonstrated the same patterns.ConclusionThe Neural ODE model offers a new tool for optimizing the deep learning model inputs with enhanced explainability. The presented methodology can be generalized to other medical image‐related deep‐learning applications.

show abstract

Section: Resultssupporting

confidence: 79%

Section: Figure 4a Provides An Example Of Four Image Flowsmentioning

confidence: 78%

A neural ordinary differential equation model for visualizing deep neural network behaviors in multi‐parametric MRI‐based glioma segmentation

Yang

et al. 2023

Medical Physics

View full text Add to dashboard Cite

show abstract

“…Subsequently, I and W are transformed into patches represented as

and

, respectively. Here, n signifies the quantity of patches, and the patch size is configured at 8, a choice consistent with previous studies ( 35 ). This configuration leads to n = 16 × 16 × 16, aligning with the concept of vision transformers ( 12 ) splitting the 2D image into 16 × 16 tokens.…”

Section: Methodsmentioning

confidence: 99%

“…We adopt a shifted window vision transformer, known as SW-ViT ( 35 ), as the transformer encoder in API-MAE. As shown in Figures 3C , D , the multi-head self-attention (MSA) in the original transformer block is replaced with linear window-based multi-head self-attention (LW-MSA) and shifted linear window-based multi-head self-attention (SLW-MSA) in the Swin transformer block.…”

Section: Methodsmentioning

confidence: 99%

“…In the context of medical images, anatomical knowledge could help improve the pre-training. Huang et al ( 35 ) incorporated the symmetry characteristics of brain structures into the pre-training by constructing symmetric positional encodings. However, few studies have integrated the more precise brain atlas ( 36 ) into the masking strategy.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improving brain tumor segmentation with anatomical prior-informed pre-training

Wang,

Li,

Wang

et al. 2023

Front. Med.

View full text Add to dashboard Cite

IntroductionPrecise delineation of glioblastoma in multi-parameter magnetic resonance images is pivotal for neurosurgery and subsequent treatment monitoring. Transformer models have shown promise in brain tumor segmentation, but their efficacy heavily depends on a substantial amount of annotated data. To address the scarcity of annotated data and improve model robustness, self-supervised learning methods using masked autoencoders have been devised. Nevertheless, these methods have not incorporated the anatomical priors of brain structures.MethodsThis study proposed an anatomical prior-informed masking strategy to enhance the pre-training of masked autoencoders, which combines data-driven reconstruction with anatomical knowledge. We investigate the likelihood of tumor presence in various brain structures, and this information is then utilized to guide the masking procedure.ResultsCompared with random masking, our method enables the pre-training to concentrate on regions that are more pertinent to downstream segmentation. Experiments conducted on the BraTS21 dataset demonstrate that our proposed method surpasses the performance of state-of-the-art self-supervised learning techniques. It enhances brain tumor segmentation in terms of both accuracy and data efficiency.DiscussionTailored mechanisms designed to extract valuable information from extensive data could enhance computational efficiency and performance, resulting in increased precision. It's still promising to integrate anatomical priors and vision approaches.

show abstract