Ultrasonic thyroid nodule detection method based on U-Net network

Deep Learning-Based Segmentation and Risk Stratification for Gastrointestinal Stromal Tumors in Transabdominal Ultrasound Imaging

PURPOSE

To develop a deep neural network system for the automatic segmentation and risk stratification prediction of gastrointestinal stromal tumors (GISTs).

METHODS

A total of 980 ultrasound (US) images from 245 GIST patients were retrospectively collected. These images were randomly divided (6:2:2) into a training set, a validation set, and an internal test set. Additionally, 188 US images from 47 prospective GIST patients were collected to evaluate the segmentation and diagnostic performance of the model. Five deep learning-based segmentation networks, namely, UNet, FCN, DeepLabV3+, Swin Transformer, and SegNeXt, were employed, along with the ResNet 18 classification network, to select the most suitable network combination. The performance of the segmentation models was evaluated using metrics such as the intersection over union (IoU), Dice similarity coefficient (DSC), recall, and precision. The classification performance was assessed based on accuracy and the area under the receiver operating characteristic curve (AUROC).

RESULTS

Among the compared models, SegNeXt-ResNet18 exhibited the best segmentation and classification performance. On the internal test set, the proposed model achieved IoU, DSC, precision, and recall values of 82.1, 90.2, 91.7, and 88.8%, respectively. The accuracy and AUC for GIST risk prediction were 87.4 and 92.0%, respectively. On the external test set, the segmentation models exhibited IoU, DSC, precision, and recall values of 81.0, 89.5, 92.8, and 86.4%, respectively. The accuracy and AUC for GIST risk prediction were 86.7 and 92.5%, respectively.

CONCLUSION

This two-stage SegNeXt-ResNet18 model achieves automatic segmentation and risk stratification prediction for GISTs and demonstrates excellent segmentation and classification performance.

Hybrid deep learning assisted multi classification: Grading of malignant thyroid nodules

Thyroid nodules are commonly diagnosed with ultrasonography, which includes internal characteristics, varying looks, and hazy boundaries, making it challenging for a clinician to differentiate between malignant and benign forms based only on visual identification. The advancement of AI, particularly DL, provides significant breakthroughs in the domain of medical image identification. Yet, there are certain obstacles to achieving accuracy as well as efficacy in thyroid nodule detection. The thyroid nodules in this study are detected and classified using an inventive hybrid deep learning-assisted multi-classification method. The median blur method is applied in this work to eliminate the salt and pepper noise from the image. Then MPIU-Net-based segmentation is utilized to segment the image. The LGBPNP-based features are retrieved from the segmented image to obtain a single histogram sequence of the LGBP pattern in addition to other features like extraction of multi-texton and LTP-based features. After the feature extraction, the data augmentation process is applied and then the features are fed to the hybrid classification-based nodule classification model that comprises Deep Maxout and CNN, this hybrid classification trains the features and predicts the thyroid nodule. Additionally, the TIRADS score classification is used for the projected malignant thyroid nodule coupled with statistical features collected from the segmented. The DBNAAF with transfer learning model is employed to classify the grading of malignant thyroid nodules, where the weights of the model are learned with transfer learning. The MCC of the Hybrid Model is 0.9445, whereas the DCNN is 0.6858, YOLOV3-DMRF is 0.7229, CNN is 0.7780, DBN is 0.7601, Bi-GRU is 0.7038, Deep Maxout is 0.7528, and RNN is 0.8522, respectively.

Semantic segmentation for tooth cracks using improved DeepLabv3+ model

Objective

Accurate and prompt detection of cracked teeth plays a critical role for human oral health. The aim of this paper is to evaluate the performance of a tooth crack segmentation model (namely, FDB-DeepLabv3+) on optical microscopic images.

Method

The FDB-DeepLabv3+ model proposed here improves feature learning by replacing the backbone with ResNet50. Feature pyramid network (FPN) is introduced to fuse muti-level features. Densely linked atrous spatial pyramid pooling (Dense ASPP) is applied to achieve denser pixel sampling and wider receptive field. Bottleneck attention module (BAM) is embedded to enhance local feature extraction.

Results

Through testing on a self-made hidden cracked tooth dataset, the proposed method outperforms four classical networks (FCN, U-Net, SegNet, DeepLabv3+) on segmentation results in terms of mean pixel accuracy (MPA) and mean intersection over union (MIoU). The network achieves an increase of 11.41% in MPA and 12.14% in MIoU compared to DeepLabv3+. Ablation experiments shows that all the modifications are beneficial.

Conclusion

An improved network is designed for segmenting tooth surface cracks with good overall performance and robustness, which may hold significant potential in computer-aided diagnosis of cracked teeth.

Multi-class classification of thyroid nodules from automatic segmented ultrasound images: Hybrid ResNet based UNet convolutional neural network approach

BACKGROUND AND OBJECTIVES

Early detection and diagnosis of thyroid nodule types are important because they can be treated more effectively in their early stages. The types of thyroid nodules are generally stated as atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS), benign follicular, and papillary follicular. The risk of malignancy for AUS/FLUS is typically stated to be between 5 and 15 %, while some studies indicate a risk as high as 25 %. Without complete histology, it is difficult to classify nodules and these diagnostic operations are pricey and risky. To minimize laborious workload and misdiagnosis, recently various AI-based decision support systems have been developed.

METHODS

In this study, a novel AI-based decision support system has been developed for the automated segmentation and classification of the types of thyroid nodules. This system is based on a hybrid deep-learning procedure that makes both an automatic thyroid nodule segmentation and classification tasks, respectively. In this framework, the segmentation is executed with some U-Net architectures such as ResUNet and ResUNet++ integrating with the feature extraction and upsampling with dropout operations to prevent overfitting. The nodule classification task is achieved by various deep nets architecture such as VGG-16, DenseNet121, ResNet-50, and Inception ResNet-v2 considering some accurate classification criteria such as Intersection over Union (IOU), Dice coefficient, accuracy, precision, and recall.

RESULTS

In analysis, a total of 880 patients with ages ranging from 10 to 90 years were included by taking the ultrasound images and demographics. The experimental evaluations showed that ResUNet++ demonstrated excellent segmentation outcomes, attaining remarkable evaluation scores including a dice coefficient of 92.4 % and a mean IOU of 89.7 %. ResNet-50 and Inception ResNet-v2 trained over the images segmented with UNets have shown better performance in terms of achieving high evaluation scores for the classification accuracy such as 96.6 % and 95.0 %, respectively. In addition, ResNet-50 and Inception ResNet-v2 classified AUS/FLUS from the images segmented with UNets with AUC=97.0 % and 96.0 %, respectively.

CONCLUSIONS

The proposed AI-based decision support system improves the automatic segmentation performance of AUS/FLUS and it has shown better performance than available approaches in the literature with respect to ACC, Jaccard and DICE losses. This system has great potential for clinical use by both radiologists and surgeons as well.

A devised thyroid segmentation with multi-stage modification based on Super-pixel U-Net under insufficient data

OBJECTIVES

The application of deep learning to medical image segmentation has received considerable attention. Nevertheless, when segmenting thyroid ultrasound images, it is difficult to achieve good segmentation results using deep learning methods because of the large number of nonthyroidal regions and insufficient training data.

METHODS

In this study, a Super-pixel U-Net, designed by adding a supplementary path to U-Net, was devised to boost the segmentation results of thyroids. The improved network can introduce more information into the network, boosting auxiliary segmentation results. A multi-stage modification is introduced in this method, which includes boundary segmentation, boundary repair, and auxiliary segmentation. To reduce the negative effects of non-thyroid regions in the segmentation, U-Net was utilized to obtain rough boundary outputs. Subsequently, another U-Net is trained to improve and repair the coverage of the boundary outputs. Super-pixel U-Net was applied in the third stage to assist in the segmentation of the thyroid more precisely. Finally, multidimensional indicators were used to compare the segmentation results of the proposed method with those of other comparison experiments.

DISCUSSION

The proposed method achieved an F1 Score of 0.9161 and an IoU of 0.9279. Furthermore, the proposed method also exhibits better performance in terms of shape similarity, with an average convexity of 0.9395. an average ratio of 0.9109, an average compactness of 0.8976, an average eccentricity of 0.9448, and an average rectangularity of 0.9289. The average area estimation indicator was 0.8857.

CONCLUSION

The proposed method exhibited superior performance, proving the improvements of the multi-stage modification and Super-pixel U-Net.

The Evolution and Application of Artificial Intelligence in Rhinology: A State of the Art Review

OBJECTIVE

To provide a comprehensive overview on the applications of artificial intelligence (AI) in rhinology, highlight its limitations, and propose strategies for its integration into surgical practice.

DATA SOURCES

Medline, Embase, CENTRAL, Ei Compendex, IEEE, and Web of Science.

REVIEW METHODS

English studies from inception until January 2022 and those focusing on any application of AI in rhinology were included. Study selection was independently performed by 2 authors; discrepancies were resolved by the senior author. Studies were categorized by rhinology theme, and data collection comprised type of AI utilized, sample size, and outcomes, including accuracy and precision among others.

CONCLUSIONS

An overall 5435 articles were identified. Following abstract and title screening, 130 articles underwent full-text review, and 59 articles were selected for analysis. Eleven studies were from the gray literature. Articles were stratified into image processing, segmentation, and diagnostics (n = 27); rhinosinusitis classification (n = 14); treatment and disease outcome prediction (n = 8); optimizing surgical navigation and phase assessment (n = 3); robotic surgery (n = 2); olfactory dysfunction (n = 2); and diagnosis of allergic rhinitis (n = 3). Most AI studies were published from 2016 onward (n = 45).

IMPLICATIONS FOR PRACTICE

This state of the art review aimed to highlight the increasing applications of AI in rhinology. Next steps will entail multidisciplinary collaboration to ensure data integrity, ongoing validation of AI algorithms, and integration into clinical practice. Future research should be tailored at the interplay of AI with robotics and surgical education.

A Multiscale Attentional Unet Model for Automatic Segmentation in Medical Ultrasound Images

Ultrasonography has become an essential part of clinical diagnosis owing to its noninvasive, and real-time nature. To assist diagnosis, automatically segmenting a region of interest (ROI) in ultrasound images is becoming a vital part of computer-aided diagnosis (CAD). However, segmenting ROIs on medical images with relatively low contrast is a challenging task. To better achieve medical ROI segmentation, we propose an efficient module denoted as multiscale attentional convolution (MSAC), utilizing cascaded convolutions and a self-attention approach to concatenate features from various receptive field scales. Then, MSAC-Unet is constructed based on Unet, employing MSAC instead of the standard convolution in each encoder and decoder for segmentation. In this study, two representative types of ultrasound images, one of the thyroid nodules and the other of the brachial plexus nerves, were used to assess the effectiveness of the proposed approach. The best segmentation results from MSAC-Unet were achieved on two thyroid nodule datasets (TND-PUH3 and DDTI) and a brachial plexus nerve dataset (NSD) with Dice coefficients of 0.822, 0.792, and 0.746, respectively. The analysis of segmentation results shows that our MSAC-Unet greatly improves the segmentation accuracy with more reliable ROI edges and boundaries, decreasing the number of erroneously segmented ROIs in ultrasound images.

A systematic review on artificial intelligence techniques for detecting thyroid diseases

The use of artificial intelligence approaches in health-care systems has grown rapidly over the last few years. In this context, early detection of diseases is the most common area of application. In this scenario, thyroid diseases are an example of illnesses that can be effectively faced if discovered quite early. Detecting thyroid diseases is crucial in order to treat patients effectively and promptly, by saving lives and reducing healthcare costs. This work aims at systematically reviewing and analyzing the literature on various artificial intelligence-related techniques applied to the detection and identification of various diseases related to the thyroid gland. The contributions we reviewed are classified according to different viewpoints and taxonomies in order to highlight pros and cons of the most recent research in the field. After a careful selection process, we selected and reviewed 72 papers, analyzing them according to three main research questions, i.e., which diseases of the thyroid gland are detected by different artificial intelligence techniques, which datasets are used to perform the aforementioned detection, and what types of data are used to perform the detection. The review demonstrates that the majority of the considered papers deal with supervised methods to detect hypo- and hyperthyroidism. The average accuracy of detection is high (96.84%), but the usage of private and outdated datasets with a majority of clinical data is very common. Finally, we discuss the outcomes of the systematic review, pointing out advantages, disadvantages, and future developments in the application of artificial intelligence for thyroid diseases detection.

Segmentation of thyroid glands and nodules in ultrasound images using the improved U-Net architecture

BACKGROUND

Identifying thyroid nodules' boundaries is crucial for making an accurate clinical assessment. However, manual segmentation is time-consuming. This paper utilized U-Net and its improved methods to automatically segment thyroid nodules and glands.

METHODS

The 5822 ultrasound images used in the experiment came from two centers, 4658 images were used as the training dataset, and 1164 images were used as the independent mixed test dataset finally. Based on U-Net, deformable-pyramid split-attention residual U-Net (DSRU-Net) by introducing ResNeSt block, atrous spatial pyramid pooling, and deformable convolution v3 was proposed. This method combined context information and extracts features of interest better, and had advantages in segmenting nodules and glands of different shapes and sizes.

RESULTS

DSRU-Net obtained 85.8% mean Intersection over Union, 92.5% mean dice coefficient and 94.1% nodule dice coefficient, which were increased by 1.8%, 1.3% and 1.9% compared with U-Net.

CONCLUSIONS

Our method is more capable of identifying and segmenting glands and nodules than the original method, as shown by the results of correlational studies.

A Novel Deep-Learning-Based CADx Architecture for Classification of Thyroid Nodules Using Ultrasound Images

Nodules of thyroid cancer occur in the cells of the thyroid as benign or malign types. Thyroid sonographic images are mostly used for diagnosis of thyroid cancer. The aim of this study is to introduce a computer-aided diagnosis system that can classify the thyroid nodules with high accuracy using the data gathered from ultrasound images. Acquisition and labeling of sub-images were performed by a specialist physician. Then the number of these sub-images were increased using data augmentation methods. Deep features were obtained from the images using a pre-trained deep neural network. The dimensions of the features were reduced and features were improved. The improved features were combined with morphological and texture features. This feature group was rated by a value called similarity coefficient value which was obtained from a similarity coefficient generator module. The nodules were classified as benign or malignant using a multi-layer deep neural network with a pre-weighting layer designed with a novel approach. In this study, a novel multi-layer computer-aided diagnosis system was proposed for thyroid cancer detection. In the first layer of the system, a novel feature extraction method based on the class similarity of images was developed. In the second layer, a novel pre-weighting layer was proposed by modifying the genetic algorithm. The proposed system showed superior performance in different metrics compared to the literature.

Automatic segmentation of thyroid with the assistance of the devised boundary improvement based on multicomponent small dataset

Deep learning has been widely considered in medical image segmentation. However, the difficulty of acquiring medical images and labels can affect the accuracy of the segmentation results for deep learning methods. In this paper, an automatic segmentation method is proposed by devising a multicomponent neighborhood extreme learning machine to improve the boundary attention region of the preliminary segmentation results. The neighborhood features are acquired by training U-Nets with the multicomponent small dataset, which consists of original thyroid ultrasound images, Sobel edge images and superpixel images. Afterward, the neighborhood features are selected by min-redundancy and max-relevance filter in the designed extreme learning machine, and the selected features are used to train the extreme learning machine to obtain supplementary segmentation results. Finally, the accuracy of the segmentation results is improved by adjusting the boundary attention region of the preliminary segmentation results with the supplementary segmentation results. This method combines the advantages of deep learning and traditional machine learning, boosting the accuracy of thyroid segmentation accuracy with a small dataset in a multigroup test.

Application of dual-source computed tomography in the diagnosis of thyroid cancer and evaluation of biological behaviors

OBJECTIVE

Thyroid cancer (TC) is an extremely prevailing malignant endocrine tumor. Therefore, effective diagnostic tools are necessary. This study explored the application value of dual-source computed tomography (DSCT) in TC diagnosis and biological behavior assessment.

METHODS

This study retrospectively selected 68 TC patients and another 74 benign patients with thyroid adenoma, nodular goiter, or adenomatous hyperplasia. All patients were confirmed by pathological examination and underwent DSCT examination. The iodine concentration (IC) obtained from plain computed tomography (CT) scanning and normalized iodine concentration (NIC) in the arterial phase and venous phase were recorded. The positive expression rates of estrogen receptor alpha (ERα), estrogen receptors beta (ERβ), and Ki67 in pathological tissues were determined by immunohistochemistry, and their correlation with IC in plain CT was assessed by Pearson correlation analysis, respectively. The diagnostic values of IC in plain CT and venous phase NIC in TC patients were evaluated using the receiver operating characteristic curve.

RESULTS

Malignant patients had lower IC in plain DSCT scanning, venous phase NIC, and ERβ, and higher ERα and Ki67 than benign patients. IC level in plain DSCT scanning was inversely-correlated with ERα and Ki-67 positive expression rates, but positively-related to ERβ to different degrees. For the diagnosis of TC patients, the AUC of IC level in plain DSCT was 0.771, with a cut-off value of 1.250 (97.06% sensitivity and 41.89% specificity), and the AUC of venous phase NIC was 0.738, with a cut-off value of 0.825 (100% sensitivity and 43.24% specificity).

CONCLUSION

The IC level obtained from DSCT scanning could assist in the differential diagnosis of malignant and benign thyroid nodules and evaluation of biological behaviors.

Differentiation and prediction of pneumoconiosis stage by computed tomography texture analysis based on U-Net neural network

BACKGROUND AND OBJECTIVE

The progressive worsening of pneumoconiosis will ensue a hazardous physical condition in patients. This study details the differential diagnosis of the pneumoconiosis stage, by employing computed tomography (CT) texture analysis, based on U-Net neural network.

METHODS

The pneumoconiosis location from 92 patients at various stages was extracted by U-Net neural network. Mazda software was employed to analyze the texture features. Three dimensionality reduction methods set the best texture parameters. We applied four methods of the B11 module to analyze the selected texture parameters and calculate the misclassified rate (MCR). Finally, the receiver operating characteristic curve (ROC) of the texture parameters was analyzed, and the texture parameters with diagnostic efficiency were evaluated by calculating the area under curve (AUC).

RESULTS

The original film was processed by Gaussian and Laplace filters for a better display of the segmented area of pneumoconiosis in all stages. The MCR value obtained by the NDA analysis method under the MI dimension reduction method was the lowest, at 10.87%. In the filtered texture feature parameters, the best AUC was 0.821.

CONCLUSIONS

CT texture analysis based on the U-Net neural network can be used to identify the staging of pneumoconiosis.

A multi-channel deep convolutional neural network for multi-classifying thyroid diseases

BACKGROUND AND OBJECTIVE

Thyroid disease instances have been continuously increasing since the 1990s, and thyroid cancer has become the most rapidly rising disease among all the malignancies in recent years. Most existing studies focused on applying deep convolutional neural networks for detecting thyroid cancer. Despite their satisfactory performance on binary classification tasks, limited studies have explored multi-class classification of thyroid disease types; much less is known of the diagnosis of co-existence situation for different types of thyroid diseases.

METHOD

This study proposed a novel multi-channel convolutional neural network (CNN) architecture to address the multi-class classification task of thyroid disease. The multi-channel CNN merits from computed tomography characteristics to drive a comprehensive diagnostic decision for the overall thyroid gland, emphasizing the disease co-existence circumstance. Moreover, this study also examined alternative strategies to enhance the diagnostic accuracy of CNN models through concatenation of different scales of feature maps.

RESULTS

Benchmarking experiments demonstrate the improved performance of the proposed multi-channel CNN architecture compared with the standard single-channel CNN architecture. More specifically, the multi-channel CNN achieved an accuracy of 0.909±0.048, precision of 0.944±0.062, recall of 0.896±0.047, specificity of 0.994±0.001, and F1 of 0.917±0.057, in contrast to the single-channel CNN, which obtained 0.902±0.004, 0.892±0.005, 0.909±0.002, 0.993±0.001, 0.898±0.003, respectively. In addition, the proposed model was evaluated in different gender groups; it reached a diagnostic accuracy of 0.908 for the female group and 0.901 for the male group.

CONCLUSION

Collectively, the results highlight that the proposed multi-channel CNN has excellent generalization and has the potential to be deployed to provide computational decision support in clinical settings.

A lightweight network for automatic thyroid nodules location and recognition with high speed and accuracy in ultrasound images

BACKGROUND

The intelligent diagnosis of thyroid nodules in ultrasound image is an important research issue. Automatically locating the region of interest (ROI) of thyroid nodules and providing pre-diagnosis results can help doctors to diagnose faster and more accurate.

OBJECTIVES

This study aims to propose a model, which can detect multiple nodules stably and accurately in order to avoid missed detection and misjudgment. In addition, the detection speed of the model needs to be fast for real-time diagnosis in ultrasound images.

METHODS

Based on the object detection technology, we propose an accurate, robust and high-speed network with multiscale fusion strategy called Efficient-YOLO, which can realize the localization and recognition of nodules at the same time. Finally, multiple metrics are used to measure the diagnostic ability of the model.

RESULTS

Experimental results conducted on 3,562 ultrasound images show that our new model greatly increases the accuracy and speed of the detection compared with the baseline model. The best mAP is 92.64%, and the fastest detection speed is 45.1 frames per second.

CONCLUSIONS

This study proposed an effective method to diagnosis thyroid nodules automatically, which can meet the real-time requirements, indicating that its effectiveness and feasibility for future clinical application.

Deep convolutional neural networks in thyroid disease detection: A multi-classification comparison by ultrasonography and computed tomography

BACKGROUND AND OBJECTIVE

As one of the largest endocrine organs in the human body, the thyroid gland regulates daily metabolism. Early detection of thyroid disease leads to reduced mortality rates. The diagnosis of thyroid disease is usually made by radiologists and pathologists, which heavily relies on their experience and expertise. To mitigate human false-positive diagnostic rates, this paper proves that deep learning-driven techniques yield promising performance for automatic detection of thyroid diseases which offers clinicians assistance regarding diagnostic decision-making.

METHOD

This research study is the first of its kind, which adopts two pre-operative medical image modalities for multi-classifying thyroid disease types (i.e., normal, thyroiditis, cystic, multi-nodular goiter, adenoma, and cancer). Using the current state-of-the-art performing deep convolutional neural network (CNN) architecture, this study builds a thyroid disease diagnostic model for distinguishing among the disease types.

RESULTS

The model obtains unprecedented performance for both medical image sets, and it reaches an accuracy of 0.972 and 0.942 for ultrasound images and computed tomography (CT) scans correspondingly.

CONCLUSION

The experimental results illustrate that the selected CNN can be adapted to both image modalities, indicating the feasibility of the deep learning model and emphasizing its further applications in clinics.

TNSNet: Thyroid nodule segmentation in ultrasound imaging using soft shape supervision

BACKGROUND AND OBJECTIVES

Thyroid nodules are a common disorder of the endocrine system. Segmentation of thyroid nodules on ultrasound images is an important step in the evaluation and diagnosis of nodules and an initial step in computer-aided diagnostic systems. The accuracy and consistency of segmentation remain a challenge due to the low contrast, speckle noise, and low resolution of ultrasound images. Therefore, the study of deep learning-based algorithms for thyroid nodule segmentation is important. This study utilizes soft shape supervision to improve the performance of detection and segmentation of boundaries of nodules. Soft shape supervision can emphasize the boundary features and assist the network in segmenting nodules accurately.

METHODS

We propose a dual-path convolution neural network, including region and shape paths, which use DeepLabV3+ as the backbone. Soft shape supervision blocks are inserted between the two paths to implement cross-path attention mechanisms. The blocks enhance the representation of shape features and add them to the region path as auxiliary information. Thus, the network can accurately detect and segment thyroid nodules.

RESULTS

We collect 3786 ultrasound images of thyroid nodules to train and test our network. Compared with the ground truth, the test results achieve an accuracy of 95.81% and a DSC of 85.33. The visualization results also suggest that the network has learned clear and accurate boundaries of the nodules. The evaluation metrics and visualization results demonstrate the superior segmentation performance of the network to other classical deep learning-based networks.

CONCLUSIONS

The proposed dual-path network can accurately realize automatic segmentation of thyroid nodules on ultrasound images. It can also be used as an initial step in computer-aided diagnosis. It shows superior performance to other classical methods and demonstrates the potential for accurate segmentation of nodules in clinical applications.