Deep learning on ultrasound images of thyroid nodules

Sharifi, Yasaman; Bakhshali, Mohamad Amin; Dehghani, Toktam; DanaiAshgzari, Morteza; Sargolzaei, M.; Eslami, Saeid

doi:10.1016/j.bbe.2021.02.008

Cited by 36 publications

(26 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…There are several lesion types, the automated classification of which is commonly studied in ultrasound imaging. Some of these are: benign versus malignant breast lesion differentiation [19][20][21][22][23][24]; thyroid cancer detection [25][26][27][28] or liver disease classification [29,30]. Lesion diagnosis techniques belong either to the traditional Computer-Aided Diagnostic (CAD) class of methods, or to the relatively recent class of deep learning (DL) methods.…”

Section: Overview Of Ultrasound-based Lesion Diagnostic Methodsmentioning

confidence: 99%

“…This requirement is usually met in the case of "grand challenges" using popular lesion types, where anyone can test their algorithm on publicly available datasets. Recent examples of the use of DL in ultrasound-based lesion diagnosis are to be found in the classification of thyroid nodules [27,28], breast lesion differentiation [24] and lung ultrasound for the detection of COVID-19-associated lesions [36].…”

Section: Overview Of Ultrasound-based Lesion Diagnostic Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Automated Skin Lesion Classification on Ultrasound Images

et al. 2021

View full text Add to dashboard Cite

The growing incidence of skin cancer makes computer-aided diagnosis tools for this group of diseases increasingly important. The use of ultrasound has the potential to complement information from optical dermoscopy. The current work presents a fully automatic classification framework utilizing fully-automated (FA) segmentation and compares it with classification using two semi-automated (SA) segmentation methods. Ultrasound recordings were taken from a total of 310 lesions (70 melanoma, 130 basal cell carcinoma and 110 benign nevi). A support vector machine (SVM) model was trained on 62 features, with ten-fold cross-validation. Six classification tasks were considered, namely all the possible permutations of one class versus one or two remaining classes. The receiver operating characteristic (ROC) area under the curve (AUC) as well as the accuracy (ACC) were measured. The best classification was obtained for the classification of nevi from cancerous lesions (melanoma, basal cell carcinoma), with AUCs of over 90% and ACCs of over 85% obtained with all segmentation methods. Previous works have either not implemented FA ultrasound-based skin cancer classification (making diagnosis more lengthy and operator-dependent), or are unclear in their classification results. Furthermore, the current work is the first to assess the effect of implementing FA instead of SA classification, with FA classification never degrading performance (in terms of AUC or ACC) by more than 5%.

show abstract

Section: Overview Of Ultrasound-based Lesion Diagnostic Methodsmentioning

confidence: 99%

Section: Overview Of Ultrasound-based Lesion Diagnostic Methodsmentioning

confidence: 99%

Automated Skin Lesion Classification on Ultrasound Images

et al. 2021

View full text Add to dashboard Cite

show abstract

“…104 They describe the following four applications methods: classification, detection, segmentation and feature selection, in the order of most used in the reviewed literature. 104 These application methods can use various networks such as "VGGNet," "GoogleNET" and "ResNet," amongst others. Training and testing of these methods is performed with a variety of different, often small, datasets of which most are not open-access, which makes comparing methods difficult.…”

Section: Future Trends-computer Aided Diagnosis (Cad) and Deep Learni...mentioning

confidence: 99%

Ultrasound imaging in thyroid nodule diagnosis, therapy, and follow‐up: Current status and future trends

Boers

Braak

Rikken

et al. 2023

J of Clinical Ultrasound

View full text Add to dashboard Cite

Ultrasound, the primary imaging modality in thyroid nodule management, suffers from drawbacks including: high inter-and intra-observer variability, limited field-of-view and limited functional imaging. Developments in ultrasound technologies are taking place to overcome these limitations, including three-dimensional-Doppler, -elastography, -nodule characteristics-extraction, and novel machine-learning algorithms. For thyroid ablative treatments and biopsies, perioperative use of three-dimensional ultrasound opens a new field of research. This review provides an overview of the current and future applications of ultrasound, and discusses the potential of new developments and trends that may improve the diagnosis, therapy, and follow-up of thyroid nodules.

show abstract

“…However, the deep learning only needs to pass the data directly to the neural networks. Thus, one of the most growing trends of ML is deep learning (Sharifi et al, 2021 ). Song et al ( 2018 ) developed a cascade convolution neural network framework, which confirmed the feasibility of CNN used for thyroid nodules detection and recognition.…”

Section: Related Workmentioning

confidence: 99%

A Multi-Scale Densely Connected Convolutional Neural Network for Automated Thyroid Nodule Classification

et al. 2022

View full text Add to dashboard Cite

Automated thyroid nodule classification in ultrasound images is an important way to detect thyroid nodules and to make a more accurate diagnosis. In this paper, we propose a novel deep convolutional neural network (CNN) model, called n-ClsNet, for thyroid nodule classification. Our model consists of a multi-scale classification layer, multiple skip blocks, and a hybrid atrous convolution (HAC) block. The multi-scale classification layer first obtains multi-scale feature maps in order to make full use of image features. After that, each skip-block propagates information at different scales to learn multi-scale features for image classification. Finally, the HAC block is used to replace the downpooling layer so that the spatial information can be fully learned. We have evaluated our n-ClsNet model on the TNUI-2021 dataset. The proposed n-ClsNet achieves an average accuracy (ACC) score of 93.8% in the thyroid nodule classification task, which outperforms several representative state-of-the-art classification methods.

show abstract

Deep learning on ultrasound images of thyroid nodules

Cited by 36 publications

References 70 publications

Automated Skin Lesion Classification on Ultrasound Images

Automated Skin Lesion Classification on Ultrasound Images

Ultrasound imaging in thyroid nodule diagnosis, therapy, and follow‐up: Current status and future trends

A Multi-Scale Densely Connected Convolutional Neural Network for Automated Thyroid Nodule Classification

Contact Info

Product

Resources

About