Computer-aided diagnosis for classifying benign versus malignant thyroid nodules based on ultrasound images: A comparison with radiologist-based assessments

Abstract

The use of thyroid cad to differentiate malignant from benign lesions shows accuracy similar to that obtained via visual inspection by radiologists. Thyroid cad might be considered a viable way to generate a second opinion for radiologists in clinical practice.

Investigation of thyroid nodules: A practical algorithm and review of guidelines

“…The advantageous perspective with CAD involves a systematic weighting of the various ultrasound parameters, as opposed to biases from interobserver variability in knowledge and experience . However, current evidence shows no significant difference between CAD systems and visual inspection by radiologists for receiver operating characteristic curves quantifying the degree of successful diagnoses …”

“…Similar to irregular margins, changes in lesion perfusion are common in both thyroiditis and malignancy, illustrating an absence of color flow and intrinsic hypervascularity, respectively . Hypervascularity and prominent central blood flow has been demonstrated in 69%‐74% of malignant thyroid nodules . For this reason, recognition of the difference between peripheral and intranodular flow is essential to rule out thyroiditis nodules with noninternal vascularity .…”

Deep convolutional neural network for the diagnosis of thyroid nodules on ultrasound

“…When 3 classifiers including the K‐Nearest Neighbor, Probabilistic Neural Network, and Decision Tree were used, sensitivity, specificity, PPV, and accuracy showed very excellent values, more than 98% . However, diagnostic performances for evaluating thyroid malignancy were better with examinations by radiologists than classifier models including the radial basis function–neural network, Naive Bayes classifier, and support vector machine (SVM) . Recently, CNN, CNN combined with other classifiers, and other classifiers alone have been investigated for diagnosing thyroid nodules .…”

“…However, when the diagnostic performances of quantitative analysis and visual assessment by radiologists were compared in predicting thyroid malignancy, quantitative analysis showed relatively inferior diagnostic performance compared to traditional grayscale US . A number of machine learning algorithms using US images have been investigated for the differential diagnosis of thyroid nodules and have shown conflicting results when diagnostic performances were compared between these algorithms and visual assessment by radiologists . The convolutional neural network (CNN) is a typical type of deep learning technique with fully trainable models and is accepted as a state‐of‐the‐art method .…”

Computer‐aided system for diagnosing thyroid nodules on ultrasound: A comparison with radiologist‐based clinical assessments

“…Computer‐aided diagnosis (CAD) was recently developed to diagnosis thyroid nodules according to sonographic features. Previous study results showed that the CAD system had an acceptable diagnostic accuracy for thyroid nodules . We compared the diagnostic performance of the thyroid ultrasound CAD system using artificial intelligence with that of an experienced radiologist using the K‐TIRADS pattern, the ACR TIRADS pattern, and the ATA guidelines.…”