Classification of radiographic lung pattern based on texture analysis and machine learning

Abstract

This study evaluated the feasibility of using texture analysis and machine learning to distinguish radiographic lung patterns. A total of 1200 regions of interest (ROIs) including four specific lung patterns (normal, alveolar, bronchial, and unstructured interstitial) were obtained from 512 thoracic radiographs of 252 dogs and 65 cats. Forty-four texture parameters based on eight methods of texture analysis (first-order statistics, spatial gray-level-dependence matrices, gray-level-difference statistics, gray-level run length image statistics, neighborhood gray-tone difference matrices, fractal dimension texture analysis, Fourier power spectrum, and Law's texture energy measures) were used to extract textural features from the ROIs. The texture parameters of each lung pattern were compared and used for training and testing of artificial neural networks. Classification performance was evaluated by calculating accuracy and the area under the receiver operating characteristic curve (AUC). Forty texture parameters showed significant differences between the lung patterns. The accuracy of lung pattern classification was 99.1% in the training dataset and 91.9% in the testing dataset. The AUCs were above 0.98 in the training set and above 0.92 in the testing dataset. Texture analysis and machine learning algorithms may potentially facilitate the evaluation of medical images.

Comparison of a Deep Learning Algorithm vs. Humans for Vertebral Heart Scale Measurements in Cats and Dogs Shows a High Degree of Agreement Among Readers

“…Recent progress in the field of computer vision gave the medical community access to powerful computer-aided diagnosis (CAD) systems and recently emerged in the field of veterinary medicine (13)(14)(15). Simply put, CAD systems can be categorized into classification (the image can have different status and the system has to choose between them, for example, prediction of osteoarthrosis severity on hips radiographs and regression algorithms; the system has to locate specific elements such as anatomical parts or lesions).…”

Evaluating artificial intelligence algorithms for use in veterinary radiology

“…Multiple individuals, who are usually recognized experts in veterinary radiology, often perform image labeling. Ground truth is also specifically defined (i.e., radiologist report, multiple radiologist consensus, national screening database, echocardiography) 3,5–7 . Research papers should include comparison (“benchmarking”) of AI algorithms to experienced radiologists 39 and those that do 5 can help practitioners in weighing their use for clinical practice.…”

“…The scope of most current AI software is limited to estimating the probability or risk that a specific radiographic pattern (lesion, variation, abnormality) is present, and algorithms do not provide complete radiographic interpretation. [2][3][4][5][6][7][8] Complete radiographic interpretation is a complex task that integrates the assessment of radiographic quality, the perception of abnormalities ("pattern detection"), and the weighting of pertinent patient information and clinical context into a list of reasonable differential diagnoses. 9 Thus, it remains paramount for the veterinarian, be they a general practitioner, non-imaging specialist, or radiologist, to consider any AI algorithm as a potentially useful assistant, a "decision support system" (DSS), not a replacement for clinical judgment.…”

“…A number of recent academic studies have described the development of AI algorithms and automated measurement tools for veterinary radiology 1–7 . These studies show that AI‐assisted detection of certain radiographic abnormalities is feasible, at least in the research setting.…”

Texture-based Feature Extraction for COVID-19 Pneumonia Classification using Chest Radiography

“…The quantitative analysis of morphological, intensity, and texture features are helpful on diagnosis and prognosis. Texture analysis can be further categorized into structural, model-based, transformational, and statistics-based [5].…”