Related Experiment Videos
Partial volume effects in MRI studies of multiple sclerosis
M J Firbank1, A Coulthard, R M Harrison
1University Department of Radiology, Royal Victoria Infirmary, Newcastle Upon Tyne, UK. m.j.firbank@ncl.ac.uk
Magnetic Resonance Imaging
|May 7, 1999
Summary
Accurate magnetic resonance imaging (MRI) volume measurements of multiple sclerosis (MS) lesions require thin slices. Thicker slices significantly increase measurement errors, emphasizing the need for precise patient positioning and thin, 1.5-mm slices for reliable MS lesion quantification.
Area of Science:
- Medical Imaging
- Neurology
- Radiology
Background:
- Accurate quantification of multiple sclerosis (MS) lesions using magnetic resonance imaging (MRI) is crucial for disease monitoring.
- Image slice thickness and object diameter can significantly impact the accuracy of lesion volume measurements.
Purpose of the Study:
- To estimate the accuracy of MRI-based volume measurements for MS lesions, particularly when lesion diameter is comparable to image slice thickness.
- To investigate the influence of image dimensions, geometry, and outlining techniques on volume measurement accuracy.
Main Methods:
- Utilized a phantom with objects of known volume and acquired MRI images at various slice thicknesses.
- Assessed a theoretical model using phantom measurements to analyze the effects of slice thickness and object position on volume accuracy.
- Evaluated different signal intensity threshold values and edge detection techniques for lesion outlining.
Main Results:
- Measured lesion volume was highly dependent on slice thickness, with thinner slices yielding more accurate results.
- Increasing slice thickness relative to object diameter led to volume measurement errors up to 100%.
- A 50% signal intensity threshold provided the most accurate volume estimates; lower thresholds and edge detection increased overestimation.
Conclusions:
- Slice thickness should be no more than one-fifth of the anticipated lesion diameter for accurate MRI volume measurement (e.g., 1.5 mm for 7 mm MS lesions).
- Patient repositioning errors can introduce significant variability in serial studies.
- Recommends using 3D MRI sequences with 1 mm cubic voxels and emphasizes consistent quality assurance for accurate MS lesion volume quantification.