SU‐E‐I‐70: Comparison of Brain Volume Quantification with NeuroQuant & FreeSurfer

• Published in: Medical physics (Lancaster) Vol. 42; no. 6Part6; p. 3258
• Main Authors: Zhou, Y, Zambelli, J, Heredia, G
• Format: Journal Article
• Language: English
• Published: American Association of Physicists in Medicine 01.06.2015
• Subjects: Brain; Computer simulation; Germanium; Image scanners; Magnetic resonance imaging; Medical image artifacts
• ISSN: 0094-2405; 2473-4209
• DOI: 10.1118/1.4924067

Purpose: This study is to evaluate the precision of volumetric evaluation using two types of image processing platforms with the MRI images obtained on a single subject from different types of scanners. Methods: The MRI scans have been performed on a single healthy volunteer during a span of 3 months using the parameters specified by the NeuroQuant manufacturer. The imaging parameters can be found at http://www.cortechslabs.com/resources/#EDUCATIONTRAINING‐0. Nine scanners with field strengths of 1.5T and 3T were used. The types of scanners are GE HDx, MR450w, MR750, MR750w, and Siemens Aera and Avanto. Some scans were repeated on the same scanner to assess the short term repeatability. The images were examined to be free of artifacts (motion or otherwise) before they were processed using both the NeuroQuant and FreeSurfer (www.freesurfer.net) platforms. Both platforms generate the volumetric data of various brain structures. During the analysis the results from both the left and right hemispheres were combined to avoid the complication from left‐right segmentation inconsistency. Over all 26 independent measurements were analyzed for comparison. Results: We have observed that both processing methods can achieve similar precision. The results for the Hippocampus, Amygdala, and Putamen show excellent agreement, while the Result of Thalamus shows some deviation. Conclusion: The robustness of both methods is excellent with available datasets, but FreeSurfer requires significant more computing power. The good agreement between two methods should increase the diagnostic confidence for future clinical use.