Diagnosis of Thyroid Nodules Based on Image Enhancement and Deep Neural Networks

Abstract

The diagnosis of thyroid nodules at an early stage is a challenging task. Manual diagnosis of thyroid nodules is labor-intensive and time-consuming. Meanwhile, due to the difference of instruments and technical personnel, the original thyroid nodule ultrasound images collected are very different. In order to make better use of ultrasound image information of thyroid nodules, some image processing methods are indispensable. In this paper, we developed a method for automatic thyroid nodule classification based on image enhancement and deep neural networks. The selected image enhancement method is histogram equalization, and the neural networks have four-layer network nodes in our experiments. The dataset in this paper consists of thyroid nodule images of 508 patients. The data are divided into 80% training and 20% validation sets. A comparison result demonstrates that our method can achieve a better performance than other normal machine learning methods. The experimental results show that our method has achieved 0.901961 accuracy, 0.894737 precision, 1 recall, and 0.944444 F1-score. At the same time, we also considered the influence of network structure, activation function of network nodes, number of training iterations, and other factors on the classification results. The experimental results show that the optimal network structure is 2500-40-2-1, the optimal activation function is logistic function, and the best number of training iterations is 500.

B-mode Ultrasound Characteristics of Thyroid Nodules With High-Benign Probability and Nodules With Risk of Malignancy

“…In addition, variability between observers may be present affecting the impression of the outcome diagnostic report [ 17 , 18 ]. It has been proposed that quantitative grey-scale analysis could improve the diagnosis rate and may be a useful tool to discriminate benign nodules from those with risk of malignancy, thus helping to reduce unnecessary FNA/CNB procedures [ 7 ]. Future studies investigating the use of quantitative grey-scale analysis alongside visual assessment and their correlation with FNA/CNB as gold standards are required.…”

“…Incidental thyroid nodules can be also detected with the use of diagnostic imaging of the neck for purposes unrelated to the thyroid. Ultrasound is the first-line imaging examination for the identification of thyroid nodules [ 7 ] and has improved the malignancy risk assessment of thyroid nodules cancer through sonographic findings, including assessment of the nodule echogenicity, internal composition, calcification and border regularity [ 5 , 8 ]. In addition, patient characteristics including age and gender have been reported to be associated with increased risk factor of thyroid cancer in which thyroid cancer predominately affects women, but may have higher mortality in men and worse prognosis in older age [ 2 ].…”

An Improved k‐Nearest Neighbor Algorithm for Recognition and Classification of Thyroid Nodules

“…In recent years, many methods for diagnosing and recognizing thyroid nodules based on machine learning algorithms have been proposed. [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] Among them, the k-nearest neighbor (KNN) algorithm, as a classic classification algorithm, has been used in many literatures for the diagnosis of thyroid nodules. Savelonas, Maroulis and Sangriotis proposed a novel computer-based approach for malignancy risk assessment of thyroid nodules in ultrasound images.…”

“…In recent years, many methods for diagnosing and recognizing thyroid nodules based on machine learning algorithms have been proposed 5–20 . Among them, the k ‐nearest neighbor (KNN) algorithm, as a classic classification algorithm, has been used in many literatures for the diagnosis of thyroid nodules.…”