An End to End Thyroid Nodule Segmentation Model based on Optimized U-Net Convolutional Neural Network

Liu, Mengya; Yuan, Xueguang; Zhang, Yangan; Chang, Kunliang; Deng, Zhifang; Xue, Jun

doi:10.1145/3429889.3429903

Cited by 2 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results validate the efficacy of these design principles of the improved U‐Net model, resulting in a performance improvement of 4.9% against the original U‐Net architecture 33 . Liu et al 39 . combine the advantages of the residual network into the U‐Net model to achieve end‐to‐end segmentation output and high‐performance segmentation with dice coefficient of 89.50%.…”

Section: Related Workmentioning

confidence: 55%

“…The model also introduces an attention gate mechanism multiplied by the weighted feature maps obtained from the shallow and deep layers.The results validate the efficacy of these design principles of the improved U-Net model, resulting in a performance improvement of 4.9% against the original U-Net architecture. 33 Liu et al 39 combine the advantages of the residual network into the U-Net model to achieve end-to-end segmentation output and high-performance segmentation with dice coefficient of 89.50%. Considering the complex tissue structure around the thyroid, Wu et al 40 propose a segmentation model using the Atrous Spatial Pyramid Pooling fusion feature, which increases the receptive field of each layer of dilated convolution.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A two‐stage network with prior knowledge guidance for medullary thyroid carcinoma recognition in ultrasound images

et al. 2022

View full text Add to dashboard Cite

Accurate recognition of medullary thyroid carcinoma (MTC) is of great importance in medical diagnosis, as MTC is rare but second-most malignant thyroid cancers with a high case-fatality ratio. 1 But there is a lower recognition rate on distinguishing MTC from other thyroid nodules in ultrasound images, even by experienced experts. This paper introduces the computer-aided method to tackle the challenge of recognizing MTC from ultrasound images, including limited MTC samples, and ambiguities among MTC, benign nodules, and papillary thyroid carcinoma (PTC). Methods: The recognition of MTC based on large MTC samples of ultrasound images has never been explored, as only one existing work presented a relevant dataset with a limited 21 MTC samples. This study proposes a novel method for primarily differentiating MTC samples from benign nodules and PTC that is the most common thyroid cancer. Our method is a two-stage schema with two important components including a cascaded coarse-to-fine segmentation network and a knowledge-based classification network. The cascaded coarseto-fine segmentation network incorporates two U-Net++ networks for improving the segmentation results of thyroid nodules. Meanwhile, our knowledge-based classification network extracts and fuses semantic features of solid tissues and calcification for better recognizing the segmented nodules from the ultrasound images. In our experiments, dice similarity coefficient (DSC), intersection over union (IoU),precision,recall,and Hausdorff distance (HD) are adopted for evaluating the segmentation results of thyroid nodules, and accuracy, precision, recall, and F1-score are used for classification evaluation. Results: We present a well-annotated dataset including samples of 248 MTC, 240 benign nodules, and 239 PTC. For thyroid nodule segmentation, our designed cascaded segmentation network attains values of 0.776 DSC, 0.689 IoU, 0.778 precision, and 0.821 recall, respectively. By incorporating prior knowledge, our method achieves a mean accuracy of 82.1% in classifying thyroid nodules of MTC, PTC, and benign ones. Especially, our method gains the higher performance in recognizing MTC with an accuracy of 86.8%,compared to nearly 70% diagnosis accuracy of experienced doctors. The experimental results on our Fujian Provincial Hospital dataset further validate the efficiency of our proposed method. Conclusions: Our proposed two-stage method incorporates pipelines of thyroid nodules segmentation and classification of MTC, individually. Quantitative and qualitative results indicate that our proposed model achieves accurate segmentation of thyroid nodules. The results also validate that our learning-based

show abstract

Section: Related Workmentioning

confidence: 55%

Section: Related Workmentioning

confidence: 99%

A two‐stage network with prior knowledge guidance for medullary thyroid carcinoma recognition in ultrasound images

et al. 2022

View full text Add to dashboard Cite

show abstract

“…A model for thyroid nodule image segmentation in medical imaging was described by Liu et al [3] and was built on an optimised U-Net Convolutional Neural Network. In order to structure the segmentation process, the Test Time Augmentation (TTA) approach was applied to the U-Net.…”

Section: Literature Surveymentioning

confidence: 99%

“…In the previous five years, 53,000 new cases of thyroid cancer were found and diagnosed only in America. Nearly 2000 people died as a result of thyroid cancer [3]. It is more challenging to distinguish the nodules from the normal tissues when they are clearly apparent.…”

Section: Literature Surveymentioning

confidence: 99%

Swin Transformer and Attention Guided Thyroid Nodule Segmentation on Ultrasound Images

Beyyala,

Priya,

Choudari

et al. 2024

ISI

View full text Add to dashboard Cite

The early detection of thyroid cancers depends on the ability to segment thyroid nodules. Thyroid ultrasound imaging is critical for diagnosing thyroid nodules. It is very challenging to correctly segment and extract the thyroid nodule from ultrasound pictures. To solve this issue, a deep learning-based segmentation module called Thyroid Region Prior Guided is recommended for the first segmentation process. This module uses three different forms of encoder, decoder, and thyroid region previous guidance to prepare for the nodule picture. This method uses ultrasound to precisely segment thyroid nodules utilizing the Swin Transformer and Attention Guided network. The DDTI dataset is used to assess the effectiveness and performance of the proposed model. The performance of the proposed model is evaluated using a number of metrics, such as accuracy of 83.43, F1-score of 68.35, IoU of 59.67, and Dice of 71.87, which ensures the best accuracy improvement over the models currently in use.

show abstract

An End to End Thyroid Nodule Segmentation Model based on Optimized U-Net Convolutional Neural Network

Cited by 2 publications

References 4 publications

A two‐stage network with prior knowledge guidance for medullary thyroid carcinoma recognition in ultrasound images

A two‐stage network with prior knowledge guidance for medullary thyroid carcinoma recognition in ultrasound images

Swin Transformer and Attention Guided Thyroid Nodule Segmentation on Ultrasound Images

Contact Info

Product

Resources

About