Texture is one of the important characteristics used in identifying objects or regions of interest in an image, whether the image be a photomicrograph, an aerial photograph, or a satellite image. This paper describes some easily computable textural features based on gray-tone spatial dependancies, and illustrates their application in category-identification tasks of three different kinds of image data: photomicrographs of five kinds of sandstones, 1:20 000 panchromatic aerial photographs of eight land-use categories, and Earth Resources Technology Satellite (ERTS) multispecial imagery containing seven land-use categories. We use two kinds of decision rules: one for which the decision regions are convex polyhedra (a piecewise linear decision rule), and one for which the decision regions are rectangular parallelpipeds (a min-max decision rule). In each experiment the data set was divided into two parts, a training set and a test set. Test set identification accuracy is 89 percent for the photomicrographs, 82 percent for the aerial photographic imagery, and 83 percent for the satellite imagery. These results indicate that the easily computable textural features probably have a general applicability for a wide variety of image-classification applications.

Textural Features for Image Classification

Computer-Aided Diagnosis in Medical Imaging: Historical Review, Current Status and Future Potential

Background Radiomic features may quantify characteristics present in medical imaging. However, the lack of standardized definitions and validated reference values have hampered clinical use. Purpose To standardize a set of 174 radiomic features. Materials and Methods Radiomic features were assessed in three phases. In phase I, 487 features were derived from the basic set of 174 features. Twenty-five research teams with unique radiomics software implementations computed feature values directly from a digital phantom, without any additional image processing. In phase II, 15 teams computed values for 1347 derived features using a CT image of a patient with lung cancer and predefined image processing configurations. In both phases, consensus among the teams on the validity of tentative reference values was measured through the frequency of the modal value and classified as follows: less than three matches, weak; three to five matches, moderate; six to nine matches, strong; 10 or more matches, very strong. In the final phase (phase III), a public data set of multimodality images (CT, fluorine 18 fluorodeoxyglucose PET, and T1-weighted MRI) from 51 patients with soft-tissue sarcoma was used to prospectively assess reproducibility of standardized features. Results Consensus on reference values was initially weak for 232 of 302 features (76.8%) at phase I and 703 of 1075 features (65.4%) at phase II. At the final iteration, weak consensus remained for only two of 487 features (0.4%) at phase I and 19 of 1347 features (1.4%) at phase II. Strong or better consensus was achieved for 463 of 487 features (95.1%) at phase I and 1220 of 1347 features (90.6%) at phase II. Overall, 169 of 174 features were standardized in the first two phases. In the final validation phase (phase III), most of the 169 standardized features could be excellently reproduced (166 with CT; 164 with PET; and 164 with MRI). Conclusion A set of 169 radiomics features was standardized, which enabled verification and calibration of different radiomics software. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Kuhl and Truhn in this issue.

https://pubs.rsna.org/doi/pdf/10.1148/radiol.2020191145

The Image Biomarker Standardization Initiative: Standardized Quantitative Radiomics for High-Throughput Image-based Phenotyping

The process of making judgments and decisions requires a method for combining data. To compare the accuracy of clinical and mechanical (formal, statistical) data-combination techniques, we performed a meta-analysis on studies of human health and behavior. On average, mechanical-prediction techniques were about 10% more accurate than clinical predictions. Depending on the specific analysis, mechanical prediction substantially outperformed clinical prediction in 33%-47% of studies examined. Although clinical predictions were often as accurate as mechanical predictions, in only a few studies (6%-16%) were they substantially more accurate. Superiority for mechanical-prediction techniques was consistent, regardless of the judgment task, type of judges, judges' amounts of experience, or the types of data being combined. Clinical predictions performed relatively less well when predictors included clinical interview data. These data indicate that mechanical predictions of human behaviors are equal or superior to clinical prediction methods for a wide range of circumstances.

/pdf/clinical-versus-mechanical-prediction-a-meta-analysis-4wov3bdnn5.pdf

Clinical versus mechanical prediction: a meta-analysis.

Segmentation of anatomical structures in chest radiographs using supervised methods: a comparative study on a public database.

Feature extraction is one of the more difficult steps in image pattern recognition. Some sources of difficulty are the presence of irrelevant information and the relativity of a feature set to a particular application. Several preprocessing techniques for enhancing selected features and removing irrelevant data are described and compared. The techniques include gray level distribution linearization, digital spatial filtering, contrast enhancement, and image subtraction. Also, several feature extraction techniques are illustrated. The techniques are divided into spatial and Fourier domain operations. The spatial domain operations of directional signatures and contour tracing are first described. Then, the Fourier domain techniques of frequency signatures and template matching are illustrated. Finally, a practical image pattern recognition problem is solved using some of the described techniques.

A Survey of Preprocessing and Feature Extraction Techniques for Radiographic Images

One goal of digital processing of radiographic images is to provide the radiologist with quantitative measurements of human anatomy as well as an indication as to whether or not this anatomy is within normal limits. A computer algorithm is described, designed to automatically detect, extract quantitative measurements from, and diagnose the cardiac projection present in full-size anteriorview chest radiographs. A normal-abnormal diagnosis is demonstrated utilizing abnormal data from five classes of heart disease. In addition, normal-abnormal as well as normal-differential diagnoses are demonstrated for the rheumatic heart disease class. A feature extraction algorithm is developed using several ad hoc techniques, some of which were adapted from other feature extraction uses. The extracted features are classified into diagnostic classes using linear and quadratic discriminant functions. A concurrent study of physician diagnostic accuracy is also undertaken using the averaged diagnostic rates of ten radiologists on a representative subset of the radiographs used in the computer study.

Automated Radiographic Diagnosis via Feature Extraction and Classification of Cardiac Size and Shape Descriptors

The authors briefly describe their experiences in the computer screening of 14 × 17-inch chest radiographs and the identification of cardiac abnormalities. The computer identified the cardiac silhouette, measured its important parameters, and applied discriminant function analysis techniques to decide whether or not the heart was within normal limits; if not, the computer reported the likely type of abnormality within the spectrum of rheumatic valvular disease. With the use of the test library of normals and cases of rheumatic heart disease, the computer had a higher overall accuracy than the radiologists viewing the same library of proved cases.

Direct computer diagnosis of rheumatic heart disease.

Computer diagnosis of heart disease.

This paper deals with digital, position variant and invariant linear and nonlinear techniques for image enhancement (Rozenfeld, 1969). The radiographic system, the image processing and display system, and the human visual system are discussed. The enhancement techniques tailor image content of radiographs to the type of information most useful to the human visual system.

R.P. Kruger

Papers

A Survey of Preprocessing and Feature Extraction Techniques for Radiographic Images

Automated Radiographic Diagnosis via Feature Extraction and Classification of Cardiac Size and Shape Descriptors

Direct computer diagnosis of rheumatic heart disease.

Computer diagnosis of heart disease.

Digital techniques for image enhancement of radiographs.