Spectral computed tomography parameters could be surrogate imaging markers to detect early perfusion changes in diabetic kidneys

Wei Zheng; Ronghua Mu; Xiaoyan Qin; Xin Li; Fuzhen Liu; Zeyu Zhuang; Peng Yang; Yahui Liang; Xiqi Zhu

doi:10.21037/qims-22-1400

Spectral computed tomography parameters could be surrogate imaging markers to detect early perfusion changes in diabetic kidneys

Quant Imaging Med Surg. 2023 Sep 1;13(9):6116-6128. doi: 10.21037/qims-22-1400. Epub 2023 Jul 25.

Authors

Wei Zheng^#^{1

2}, Ronghua Mu^#¹, Xiaoyan Qin^#¹, Xin Li¹, Fuzhen Liu¹, Zeyu Zhuang^{1

2}, Peng Yang¹, Yahui Liang³, Xiqi Zhu¹

Affiliations

¹ Department of Radiology, Nanxishan Hospital of Guangxi Zhuang Autonomous Region, Guilin, China.
² Department of Radiology, Graduate School of Guilin Medical University, Guilin, China.
³ Department of Radiology, Guilin Medical University, Guilin, China.

^# Contributed equally.

Abstract

Background: Kidney microvasculopathy is the baseline pathophysiological feature of diabetic kidney disease (DKD). We aimed to evaluate the spectral computed tomography (CT) parameters for detecting renal perfusion changes among diabetic patients.

Methods: From August 2020 to June 2022, 34 patients (age, 57.7±10.7 years; male, 20) clinically diagnosed with type 2 diabetes mellitus (DM) and 19 DM-free individuals (age, 48.1±16.9 years; male, 12) were selected for analysis. The series participants formed the DM group and control group, respectively. Spectral parameters, including effective atomic number (Z_eff), iodine density (ID), normalized iodine density (NID) and the slope of the energy spectrum curves (λ), between the 2 groups were analyzed using independent samples t-test. Receiver operator characteristic (ROC) curves were used to evaluate the diagnostic performance of spectral parameters for detecting renal perfusion changes.

Results: The results indicate that in both cortical and medullary phases, the values of Z_eff, ID, NID, and λ_40-70 for the renal cortex of the DM group were significantly higher than those in the control group (P<0.05). In the cortex phase, the diagnostic efficacy of cortical spectral CT parameters discriminating DM patients from controls was as follows: the area under ROC curve (AUC) of ID value was 0.816 [95% confidence interval (CI): 0.679-0.921] at the optimal cutoff value 4.14, the AUC of Z_eff value was 0.800 (95% CI: 0.668-0.901) at the optimal cutoff value 9.26, the AUC of λ_40-70 value was 0.822 (95% CI: 0.675-0.918) at the optimal cutoff value 8.26, and the AUC of NID value was 0.851 (95% CI: 0.684-0.926) at the optimal cutoff value 0.37. In medullary phase: the AUC of ID value was 0.769 (95% CI: 0.617-0.846) at the optimal cutoff value 5.08, the AUC of Z_eff value was 0.763 (95% CI: 0.614-0.837) at the optimal cutoff value 9.58, the AUC of λ_40-70 value was 0.766 (95% CI: 0.617-0.839) at the optimal cutoff value 10.07, and the AUC of NID value was 0.79 (95% CI: 0.623-0.855) at the optimal cutoff value 1.37.

Conclusions: Spectral CT could serve as an alternative protocol for the early identification of kidney injury in diabetic patients.

Keywords: X-ray; computed tomography (CT); diabetic kidney disease; perfusion disorder; spectral parameters.