Related Experiment Video
Updated: Jul 7, 2026
A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia
Published on: September 16, 2017
Minimum detectable difference of MR diffusion maps in acute ischemic stroke
Ashley D Harris1, Mirthula Govindaraj, Richard Frayne
1Seaman Family MR Research Centre, Foothills Medical Centre, Calgary Health Region, Calgary, Alberta, Canada.
Purpose:
To determine the minimum detectable difference (MDD) and investigate variability of region-of-interest (ROI) analysis of apparent diffusion coefficient (ADC) and fractional anisotropy (FA) in acute ischemic stroke.
Materials And Methods:
Ten patients with acute stroke (<24 hours) and moderate-to-large infarcts were imaged using a fast diffusion tensor technique. Four observers repeated three trials, during which each of two ROI types (free-hand polygon and ellipse) were drawn in white and gray matter (WM and GM) on FA and ADC maps. Analysis-of-variance techniques examined tissue and ROI type effects as well as inter- and intraobserver variability. F-tests examined the variability differences between ROI types.
Results:
The MDD for ADC was 0.160 x 10(-3) mm(2) s(-1) in WM and 0.212 x 10(-3) mm(2) s(-1) in GM. The FA MDD was 0.19 in WM and 0.10 in GM. Tissue but not ROI type affected the mean values for both ADC and FA maps. Intraobserver reliability was substantial, while interobserver reliability was poor-to-moderate. No variability differences were found by ROI types.
Conclusion:
The MDD for WM and GM in normal and ischemic tissue were calculated. Inter- and intraobserver variability and tissue type affect ROI analysis of ADC and FA maps.
Insights
This study determined the minimum detectable difference (MDD) for apparent diffusion coefficient (ADC) and fractional anisotropy (FA) in acute ischemic stroke. Results show that tissue type impacts analysis, while inter-observer reliability is a key concern for region-of-interest (ROI) analysis.
Area of Science:
- Neuroimaging
- Diffusion Tensor Imaging
- Stroke Research
Background:
- Region-of-interest (ROI) analysis is crucial for quantifying diffusion metrics in acute ischemic stroke.
- Understanding the variability and minimum detectable difference (MDD) of these metrics is essential for reliable interpretation.
- Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) are key diffusion tensor imaging (DTI) parameters.
Purpose of the Study:
- To determine the minimum detectable difference (MDD) for ADC and FA in acute ischemic stroke.
- To investigate the variability associated with ROI analysis of ADC and FA.
- To compare the effects of different ROI types and tissue types (white matter vs. gray matter) on these metrics.
Main Methods:
- Ten patients with acute ischemic stroke (<24 hours) underwent fast DTI.
- Four observers repeatedly drew two types of ROIs (free-hand polygon, ellipse) on ADC and FA maps in white and gray matter.
- Analysis of variance (ANOVA) was used to assess tissue and ROI type effects, and inter/intra-observer variability.
Main Results:
- MDD for ADC was 0.160 x 10(-3) mm(2)/s in white matter (WM) and 0.212 x 10(-3) mm(2)/s in gray matter (GM).
- MDD for FA was 0.19 in WM and 0.10 in GM.
- Tissue type significantly affected mean ADC and FA values, but ROI type did not. Intra-observer reliability was substantial; inter-observer reliability was poor-to-moderate.
Conclusions:
- MDD values for WM and GM in normal and ischemic tissues were established.
- Inter- and intra-observer variability, along with tissue type, significantly influence the reliability of ROI analysis for ADC and FA in acute stroke.
- These findings highlight the importance of standardized ROI placement and awareness of observer variability in DTI analysis of stroke.
