Early assessment of acute kidney injury using targeted field of view diffusion-weighted imaging: An in vivo study

Hanjing Kong,Chengyan Wang,Fei Gao,Xiaodong Zhang,Min Yang,Li Yang,Xiaoying Wang,Jue Zhang
DOI: https://doi.org/10.1016/j.mri.2018.10.012
IF: 3.13
2019-01-01
Magnetic Resonance Imaging
Abstract:Acute kidney injury (AKI) is a common complication in various clinical settings. In recent years, AKI diagnostics have been investigated intensively showing the emerging need for early characterization of this disease. To verify whether targeted field-of-view diffusion-weighted imaging (tFOV-DWI) is feasible to significantly improve the performance of traditional full field-of-view diffusion-weighted imaging (fFOV-DWI) in the early assessment of AKI. 14 rabbits with unilateral AKI were induced by injection of microspheres under the guidance of digital subtraction angiography (DSA). All rabbits underwent tFOV-DWI and fFOV-DWI immediately after the surgery. Artifacts, distortion and lesion identification were graded by two experienced radiologists, and the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were measured. Apparent diffusion coefficient (ADC) maps were then derived. Blood samples were collected pre- and post-surgery and serum creatinine weres measured. Renal specimen and biopsy were performed as the reference standard. Student t-test was used to ascertain statistical significance between the above parameters for tFOV-DWI and fFOV-DWI. The interobserver agreement and ADC measurements agreement were assessed. A higher percentage of renal lesions (17 out of 19) were detected in tFOV-DWI compared with fFOV-DWI (14 out of 19). Significant differences were observed in ADC value for both techniques between the lesion regions and normal tissues (p < 0.001). Histological findings were inversely correlated with ADC values of tFOV-DWI (r = −0.97, P < 0.001 for cortex; r = −0.98, P < 0.001 for medulla) and fFOV-DWI sequences (r = −0.95, P < 0.001 for cortex; r = −0.98, P < 0.001 for medulla). Those tFOV-DW images rated by the radiologists exhibit superior performance in terms of all assessed measures (P < 0.05), and interobserver agreement was excellent (ICC, 0.78 to 0.92). Besides, the ADC values derived from tFOV-DWI had a satisfactory agreement with those estimated by fFOV-DWI. The animal study demonstrates that the tFOV-DWI strategy provided visually better image quality and lesion depiction than conventional fFOV-DWI for early assessment of AKI.
What problem does this paper attempt to address?