Visual grading regression: analysing data from visual grading experiments with regression models

Abstract

For visual grading experiments, which are an easy and increasingly popular way of studying image quality, hitherto used data analysis methods are often inadequate. Visual grading analysis makes assumptions that are not statistically appropriate for ordinal data, and visual grading characteristic curves are difficult to apply in more complex experimental designs. The approach proposed in this paper, visual grading regression (VGR), consists of an established statistical technique, ordinal logistic regression, applied to data from single-image and image-pair experiments with visual grading scores selected on an ordinal scale. The approach is applicable for situations in which, for example, the effects of the choice of imaging equipment and post-processing method are to be studied simultaneously, while controlling for potentially confounding variables such as patient and observer identity. The analysis can be performed with standard statistical software packages using straightforward coding of the data. We conclude that the proposed statistical technique is useful in a wide range of visual grading studies.

Application of adaptive non-linear 2D and 3D postprocessing filters for reduced dose abdominal CT

“…Subjective image quality was analyzed using visual grading regression (ordinal logistic regression), which is suitable for the ordinal scoring system (13). Visual grading regression allows for simultaneous evaluation of effects and interactions of different independent variables, such as imaging equipment, types of post-processing techniques, variations in patient size and contrast enhancement phases as well as differences between patients and between observers.…”

Clinical Audit of Image Quality in Radiology Using Visual Grading Characteristics Analysis

“…The results of the visual grading analysis for each criterion for lumbar spine AP are presented in Table 3. O12-5 Image quality audit using VGC analysis 8 The parameters presented are the area under the VGC-curve, AVGC, the standard deviation sVGC, and the inter-observer reliability (AC1). Table 4 and 5 show the corresponding results for chest AP and for abdominal CT, respectively.…”

“…One important advantage of visual grading is that observers can assess almost any radiograph, provided that the image criteria are properly selected, and have validity and relevance to the observer. Visual grading characteristics (7) and visual grading regression (8) are two practical implementations of visual grading techniques in clinical settings. They both rely on the hypothesis that the possibility to detect pathology correlates with the clarity of the reproduction of important anatomy.…”

Clinical Applications of Synthetic MRI of the Brain

“…This kind of data are ordinal-type data and need to be handled accordingly statistically. [84] In this study synthetic images had lower scores on image quality compared to conventional images. However, they were comparable with conventional images when it came to a diagnostically acceptable level, with 82-89% of synthetic images having sufficient diagnostic quality compared with 90-95% of conventional images.…”

“…Scores for the image quality criteria were analysed with visual grading regression. [84] The agreement between observers concerning diagnosis was described with the unweighted kappa coefficient. The frequency of correct diagnosis was compared between conventional and synthetic acquisition with binary logistic regression.…”