Prostate MR image quality of apparent diffusion coefficient maps versus fractional intracellular volume maps from

Saurabh Singh1, Francesco Giganti2, Louise Dickinson3

  • 1Centre for Medical Imaging, Division of Medicine, University College London, London, UK; Department of Radiology, University College London Hospitals NHS Foundation Trust, London, UK.

PubMed
Abstract

Insights

Image quality of fractional intracellular volume (FIC) maps from VERDICT MRI and apparent diffusion coefficient (ADC) maps were comparable. Inter-reader agreement varied by scoring method, with higher agreement for ADC maps using the Likert score.

Area of Science:

  • Radiology
  • Medical Imaging
  • Oncology

Background:

  • Diffusion-weighted MRI (DW-MRI) provides apparent diffusion coefficient (ADC) maps for tumor assessment.
  • Vascular, Extracellular, Restricted Diffusion for Cytometry in Tumours (VERDICT) MRI offers fractional intracellular volume (FIC) maps.
  • Assessing image quality and inter-reader agreement is crucial for clinical adoption.

Purpose of the Study:

  • To evaluate and compare the image quality of ADC maps and FIC maps.
  • To assess the inter-reader agreement for both types of diffusion imaging maps.
  • To analyze data from patients in the INNOVATE trial.

Main Methods:

  • Two readers assessed image quality of ADC and FIC maps from 57 INNOVATE trial patients.
  • Prostate Imaging Quality (PI-QUAL) and Likert image quality (Likert-IQ) scores were used.
  • Statistical analysis included Wilcoxon Signed Ranks test and Cohen's kappa for inter-reader agreement.

Main Results:

  • No significant difference in PI-QUAL scores between FIC and ADC maps.
  • Reader 1 found higher Likert-IQ for FIC maps (p=0.021), but Reader 2 did not (p=0.663).
  • Inter-reader agreement was fair for PI-QUAL (FIC: 0.27, ADC: 0.39) and Likert-IQ (FIC: 0.38, ADC: 0.62).

Conclusions:

  • Image quality of FIC and ADC maps is comparable.
  • Inter-reader agreement was similar for PI-QUAL, but substantially higher for ADC maps using Likert-IQ.