Varying diffusion curvature, intravoxel incoherent motion, and conventional monoexponential diffusion models for evaluating Ki-67 expression in rectal carcinoma
Related Experiment Video
Updated: Nov 6, 2026

Easy Measurement of Diffusion Coefficients of EGFP-tagged Plasma Membrane Proteins Using k-Space Image Correlation Spectroscopy
Published on: May 10, 2014
Yingying Cui1, Huijia Yin2, Ying Wang1
1Department of Radiology, Henan Provincial People's Hospital and Zhengzhou University People's Hospital, Zhengzhou, China.
Objectives:
This study aimed to systematically evaluate and compare the diagnostic performance of three diffusion models - varying diffusion curvature (VDC), intravoxel incoherent motion (IVIM), and conventional monoexponential diffusion-weighted imaging (DWI) - in predicting Ki-67 expression status in rectal adenocarcinoma.
Methods:
A total of 84 patients with rectal carcinoma were enrolled, comprising 48 cases with high proliferative activity (Ki-67 > 50%) and 36 cases with low proliferative activity (Ki-67 ≤ 50%). All patients underwent pelvic MRI examinations. The following imaging parameters were calculated and compared: VDC-derived parameters (D0, D1), IVIM-derived parameters (ADCslow, ADCfast, f), and the DWI-derived parameter (ADC). Logistic regression analysis was performed to identify independent predictive factors, and diagnostic performance was assessed using the area under the receiver operating characteristic curve (AUC), the DeLong test, and calibration curves.
Results:
Compared with the low-proliferation group, the high-proliferation group of rectal carcinoma exhibited significantly higher D1 and f values, along with markedly lower D0, ADC, and ADCslow values (all P < 0.05). Logistic regression analysis identified D1 and ADCslow as independent predictors of Ki-67 status in rectal carcinoma. The combined diagnostic model incorporating D1 and ADCslow yielded an AUC of 0.924, which was significantly superior to the individual parameters, including ADC (AUC = 0.780), ADCslow (AUC = 0.832), D0 (AUC = 0.696), D1 (AUC = 0.857), and f (AUC = 0.802) (all P < 0.05). The calibration curve demonstrated good stability and reliability of the combined diagnostic model.
Conclusion:
Both VDC, IVIM, and DWI demonstrate efficacy in evaluating the Ki-67 expression status in rectal carcinoma. The combined parameters of D1 and ADCslow may serve as a novel and reliable imaging biomarker for assessing the proliferative activity of rectal carcinoma.