[Retracted] Application of Dynamic Enhanced Magnetic Resonance Imaging in the Diagnosis of Hematological Malignancies
Table 2
DE-MRI image features.
Common features
Description
Evaluation
Morphological characteristics
Roundness: the roundness of the lesion Area: the area of the lesion Edge sharpness: average signal gradient at the edge of the lesion Border irregularity: the shape deviation of the edge of the lesion
Important guiding role in the judgment of benign and malignant tumors, and the calculation is relatively simple. Classification effect is not as good as other features.
Texture feature
Contrast: it reflects the sharpness of the image Autocorrelation: it reflects the texture consistency of the image ASM energy: it reflects the uniformity of image grayscale Inverse moment of difference: it reflects the homogeneity of image texture Entropy: it reflects the nonuniformity of image texture
Reflect the texture characteristics of the image. Classifiers trained as features are better than morphological features.
Hemodynamic characteristics
Maximum absorption: the peak of the kinetic curve Peak time: the time required to reach the peak Absorption rate: the absorption rate of the contrast agent Attenuation rate: the rate of attenuation of the contrast agent Enhancement ratio: the ratio of signal strength before enhancement to average signal strength enhancement