Whole-tumour diffusion kurtosis MR imaging histogram analysis of rectal adenocarcinoma: Correlation with clinical pathologic prognostic factors

Abstract

• K correlated positively with some important prognostic factors of rectal cancer. • K showed higher AUC and specificity for differentiation of nodal involvement. • DKI metrics with whole-tumour volume histogram analysis depicted tumour heterogeneity.

Diffusion kurtosis imaging‐derived histogram metrics for prediction of KRAS mutation in rectal adenocarcinoma: Preliminary findings

“…The 10 th , 25 th , 50 th , mean, 75 th , and 90 th percentile pixel values were generated, and the kurtosis (a measure of the peakedness of the histogram) and skewness (a measure of asymmetry of the histogram about its mean) were calculated as well. The interobserver agreement was proved to be good to excellent for the whole‐tumor volume histogram metrics of DW‐MRI‐derived parameters (ADC, D, K), as reported in our previous study with the same observers . Furthermore, whole‐tumor slice (WTS)‐outline ROI resulted in the best intra‐ and interobserver ICC, compared with single and three slices ROI analysis .…”

“…Similar results were found in our study, where K derived from the DKI model significantly showed higher AUC and sensitivity for predicting KRAS mutations than conventional ADC analysis. The increased K value might be associated with a more complex microenvironment, with the presence of increased cellular density, larger percentage of gland formation, and marked variation of nuclear pleomorphism that reflect more peak distribution of tissue diffusivities in the setting of non‐Gaussian diffusion behavior . Additionally, mutations of the RAS oncogene were responsible for the constitutive activation of the RAS–RAF–MAPK pathway in an EGFR binding‐independent manner, which could result in tumor cell diffuse proliferation and decreased apoptosis .…”

“…The oblique axial and oblique coronal planes were perpendicular and parallel to the long axis of the tumor by using the sagittal plane (repetition time / echo time [TR/TE] = 3000/80 msec; flip angle = 90°; field of view [FOV] = 18 × 18 cm; NSA = 2; slice thickness = 3 mm; slice gap = 0.3 mm). The DKI (b = 0, 700, 1400, and 2100 s/mm 2 ) sequence in the transverse plane was undertaken with the parallel acquisition technique (sensitivity encoding [SENSE]) using a single‐shot echo‐planar imaging (SS‐EPI) sequence with the following parameters: TR/TE = 4000/80 msec; FOV = 25 × 25 cm; slice thickness = 3 mm, no slice gap; slice number = 32, flip angle = 90°; parallel imaging acceleration factor = 2; matrix = 256 × 256, as described in our previous study . The motion‐sensitive diffusion gradients were set along three orthogonal directions.…”

“…Diffusion kurtosis imaging (DKI), an advanced non‐Gaussian DWI model, can provide more precise information in terms of the heterogeneity and irregularity of tissue components . Superior performance of DKI over conventional DWI has been found in tumor characterization and the prediction and assessment of therapeutic response of CRT in patients with rectal cancer …”

Diffusion Kurtosis at 3.0T as an in vivo Imaging Marker for Breast Cancer Characterization: Correlation With Prognostic Factors

“…Kurtosis reflects the deviation of diffusion properties from Gaussian behavior, whereas diffusivity is the diffusion coefficient corrected for non‐Gaussian bias. DKI has shown promising results for characterizing gliomas, prostate cancer, rectal adenocarcinoma, and breast lesions . However, the diagnostic performance of DKI in comparison with DWI, DCE‐MRI, and 1 H‐MRS for breast cancer characterization has not been reported.…”