Abstract #4333

Automatic Detection of Lipid Peaks in MR Spectroscopic Image Data Using Artificial Neural Networks

Balasrinivasa R. Sajja¹, Himanshu Bhat², Michael D. Boska¹, Ponnada A. Narayana³

¹Department of Radiology, Univ. of Nebraska Medical Center, Omaha, NE, USA; ²Department of Biomedical Engineering, Northwestern University, Chicago, IL, USA; ³Department of Diagnostic and Interventional Imaging, Univ. of Texas Medical School, Houston, TX, USA

Presence of lipid in brain may represent active de- or re-myelination in multiple sclerosis. Manual examination of each voxel for lipid peak in MRSI data is tedious and this becomes acute in multicenter clinical trials where large data need to be analyzed. In this study, we present a rapid and robust method based on artificial neural networks for automatic identification of lipid peaks in MRSI data.