Expert systems help identify Benign and Precancerous colorec
Now open: Certificate Course in Management of Covid-19 by Govt. Of Gujarat and PlexusMDKnow more...Now open: Certificate Course in Management of Covid-19 by Govt. Of Gujarat and PlexusMDKnow more...
During the past two decades, CT colonography emerged as a noninvasive screening method for colorectal cancer. A recent study suggests that the machine-learning-based image analysis enabled noninvasive detection and differentiation of benign and premalignant colorectal polyps with CT colonography.

The purpose of this study published by the journal Radiology was to perform machine learning–based differentiation of benign and premalignant colorectal polyps detected with CT colonography in an average-risk asymptomatic colorectal cancer screening sample with external validation using radiomics.

In this secondary analysis of a prospective trial, colorectal polyps of all size categories and morphologies were manually segmented on CT colonography images and were classified as benign (hyperplastic polyp or regular mucosa) or premalignant (adenoma) according to the histopathologic reference standard. Quantitative image features characterizing shape, gray level histogram statistics, and image texture was extracted from segmentations after applying 22 image filters, resulting in 1906 feature-filter combinations. Based on these features, a random forest classification algorithm was trained to predict the individual polyp character. Diagnostic performance was validated in an external test set.

The random forest model was fitted using a training set consisting of 107 colorectal polyps in 63 patients comprising 169 segmentations on CT colonography images.

--The external test set included 77 polyps in 59 patients comprising 118 segmentations.

--Random forest analysis yielded an area under the receiver operating characteristic curve of 0.91, a sensitivity of 82%, and a specificity of 85% in the external test set.

--In two subgroup analyses of the external test set, the area under the receiver operating characteristic curve was 0.87 in the size category of 6–9 mm and 0.90 in the size category of 10 mm or larger.

--The most important image feature for decision-making was quantifying first-order gray level histogram statistics.

In this proof-of-concept study, machine learning–based image analysis enabled noninvasive differentiation of benign and premalignant colorectal polyps with CT colonography.

Source: https://doi.org/10.1148/radiol.2021202363
Like
Comment
Share