Document Type : Original Article(s)
Authors
1
PhD Student, Department of Medical Physics, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran
2
Resident, Department of Neurosurgery, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran.
3
BS Student, Department of Radiation technology, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran
4
Radiology Expert, Ghaem Educational, Research and Treatment Center, Mashhad, Iran.
5
MSc student, Department of Medical Physics, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran
6
Associate Professor, Department of Medical Physics, Faculty of Medicine, Mashhad University of Medical Sciences AND Medical Physics Research Center, Basic Sciences Research Institute, Mashhad University of Medical Sciences, Mashhad, Iran
10.48305/jims.v43.i816.0544
Abstract
Background: This study evaluated the performance of mono-exponential, bi-exponential, and stretched exponential diffusion models in diffusion-weighted imaging (DWI) for grading gliomas.
Methods: Thirty patients with confirmed gliomas underwent DWI using 10 b-values. We extracted apparent diffusion coefficient (ADC) from the mono-exponential model, D, f, and D* from the bi-exponential model, and DDC and α from the stretched exponential model. These parameters were compared between high- and low-grade tumors and normal tissue. Data normality was assessed, followed by statistical analysis (T-tests or Mann-Whitney U tests, P < 0.05). We also evaluated the sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) for significant parameters.
Findings: The results of this study show that all parameters of the different models have the ability to create a significant difference in cancerous tissue, but none of them have this ability in normal tissue. Despite the small difference between the parameters of the different models, the parameter D has the highest area under the ROC curve, with approximately 78 percent. Also, these results show that if the numbers are divided by the normal region, the ability to differentiate between grades is lost.
Conclusion: Our results show that DWI technique with each model can provide valuable information about tissue without contrast agent, and all parameters of models can differentiate between high- and low-grade tumors. The mono-exponential method is the most practical option if rapid imaging and calculations are essential. However, the bi-exponential method is better suited when accurate absolute values are required.
Highlights
Hadi Akbari-Zadeh: Google Scholar, PubMed
Alireza Montazerabadi: Google Scholar, PubMed
Keywords
Main Subjects
)