Reproducibility of brain ADC histograms

Area of Science:

• Radiology
• Neuroimaging
• Medical Physics

Background:

• Apparent diffusion coefficient (ADC) histogram analysis is a quantitative method for evaluating diffusion-weighted imaging (DWI) data.
• Understanding the impact of acquisition parameters on ADC histograms is crucial for consistent interpretation.
• Scan-rescan reproducibility is essential for the clinical utility of any imaging biomarker.

Purpose of the Study:

• To evaluate how different DWI acquisition techniques affect whole-brain ADC histogram parameters.
• To assess the scan-rescan reproducibility of ADC histogram analysis.
• To investigate the utility of FLAIR DWI in mitigating confounding factors.

Main Methods:

• Healthy subjects underwent DWI with varying maximum b-values (0-800, 0-1000, 0-1500 s/mm²) and FLAIR DWI (0-1000 s/mm²).
• Sequences were repeated with and without repositioning to assess reproducibility.
• Paired-sample t tests were used to compare histogram parameters (peak location, peak height, mean ADC) and region of interest (ROI) mean ADC.

Main Results:

• Increasing maximum b-values shifted ADC histograms to lower values with increased peak height (p<0.01).
• FLAIR DWI resulted in lower ADC values and a significantly higher, narrower peak (p<0.01) compared to standard DWI.
• No significant differences were found in ROI mean ADC or scan-rescan reproducibility across techniques.

Conclusions:

• DWI acquisition techniques significantly influence ADC histogram parameters.
• ADC histogram analysis demonstrates robustness and reproducibility with a given pulse sequence.
• FLAIR DWI effectively reduces the partial-voluming effect of cerebrospinal fluid, improving histogram analysis reliability.