Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Radiomics analysis workflow.

First, the clinical CT images of malignant and benign pulmonary nodules were collected. Second, image segmentation was used to delineate the pulmonary nodules. Next, the image features were extracted by the automated high-throughput feature analysis algorithm. Finally, the statistical analysis was applied and the sequential forward search was used for feature selection for the classification of lung nodules.

More »

Fig 1 Expand

Table 1.

Characteristics of population.

More »

Table 1 Expand

Fig 2.

Examples of lung lesion segmentation.

Original CT image (a) and target segmentation (b) of a benign lung lesion (tuberculosis) in patient’s left upper lobe. Another original CT image (c) and target segmentation (d) of a malignant lung tumor (adenocarcinoma) in patient’s left upper lobe.

More »

Fig 2 Expand

Table 2.

Radiomics feature characteristics.

More »

Table 2 Expand

Table 3.

Radiomics feature list that had significant difference (p<0.05) between malignant and benign groups.

More »

Table 3 Expand

Fig 3.

Heat map of the selected 4-features radiomics signature.

Radiomics features expression with Z-score. Hierarchical clustering of lung lesions is on the x axis (n = 75, B = Benign, M = Malignant). The 4-feature radiomics signature expression is on the y axis.

More »

Fig 3 Expand

Fig 4.

Prediction performance of the three different feature sets.

A leave-one-out cross-validation was performed and the accuracies in the malignant and benign nodules were plotted. The randomly selected 4 features group was examined in a 1000-time permutation test.

More »

Fig 4 Expand

Table 4.

Selected feature.

More »

Table 4 Expand