Lung Nodule Classification With Multilevel Patch-Based Context Analysis

Abstract

In this paper, we propose a novel classification method for the four types of lung nodules, i.e., well-circumscribed, vascularized, juxta-pleural, and pleural-tail, in low dose computed tomography scans. The proposed method is based on contextual analysis by combining the lung nodule and surrounding anatomical structures, and has three main stages: an adaptive patch-based division is used to construct concentric multilevel partition; then, a new feature set is designed to incorporate intensity, texture, and gradient information for image patch feature description, and then a contextual latent semantic analysis-based classifier is designed to calculate the probabilistic estimations for the relevant images. Our proposed method was evaluated on a publicly available dataset and clearly demonstrated promising classification performance.

Medical image classification with convolutional neural network

“…Feature extraction techniques commonly used in medical imaging include intensity histograms, filter-based features [18], [22], and the recently very popular scale-invariant feature transform (SIFT) [17], [22] and local binary patterns (LBP) [16], [18], [22]. The feature vectors extracted are normally used to train a classification model, e.g.…”

Assessing diagnostic complexity: An image feature-based strategy to reduce annotation costs

“…The union of the most important features from the two data distributions (balanced and unbalanced) resulted in a set of 21 features that included 8 uncorrelated features. Then three classification models were created on all features (42), most important features (21), and most important uncorrelated features (8). Table 6 shows the results for the combination of trees parameters that resulted in the highest average accuracy for balanced and unbalanced datasets, and for each feature set.…”

“…While all the above CAD studies relate to the prediction of malignancy, there are other studies that look into predicting specific characteristics of lung nodules that are important in the diagnosis process. For example, [42] proposed a patch-based context analysis to differentiate between well-circumscribed, vascularized, juxto-pleural, and pleural tail types of nodules. Further, [43] classified lung nodules into round, lobulated, densely spiculated, ragged, and halo based on their margin characteristic.…”

Text- and Content-Based Medical Image Retrieval in the VISCERAL Retrieval Benchmark

“…The text-based approach is performed given the manual clinical / pathological descriptions, which require that the experts manually index the images with alphanumerical keywords if no text is already available with the images. The content-based retrieval is based on the image visual content information, which automatically extracts the rich visual properties / features to characterize the images [10][11][12]. While the text-based retrieval is the more common method, the content-based approach is attracting more interest due to the fact that medical image data have expanded rapidly in the past two decades [13,[15][16][17].…”