Separability of Acute Cerebral Infarction Lesions in CT Based Radiomics: Toward Artificial Intelligence-Assisted Diagnosis

Figure 5

Feature map of (a) lesion region (left) and its symmetric region (right) showed separable by calculating short-run low gray-level emphasis on the square transformed images, (b) normal region (left) and its symmetric region (right) showed inseparable by calculating run entropy on the wavelet transformed images, and (c) lesion region (left) and same position of normal region (right) showed separable by calculating 10th percentile on the wavelet transformed images.

(a)

(b)

(c)