Spectral peak quantification of metabolites of interest requires a long postprocessing time, which becomes longer when a large array of data as in CSI acquisition are to be elaborated. Furthermore, peak overlapping and lower signal-to-noise ratio in short echo time (20ms) spectra make peak quantification very difficult. We propose therefore an automatic method for metabolites quantification, using Wavelet Packets to decompose the time domain FID signal in subsignals. To each subsignal a Linear Prediction Singular Value Decomposition (LPSVD) method has been applied to compute the peak parameters such as amplitude, phase, frequency and damping factor. The estimated amplitude is used to calculate metabolite ratios and to extract metabolic maps. The proposed method has been validated on simulated data and on phantom and then applied to healthy volunteers and on some patients. The proposed automatic method allows peak quantification as well as metabolic maps representation in a reliable way, eliminating operator dependence and time consuming corrections.
|Number of pages||6|
|Journal||Rivista di Neuroradiologia|
|Publication status||Published - 2000|
ASJC Scopus subject areas
- Clinical Neurology
- Radiology Nuclear Medicine and imaging
- Radiological and Ultrasound Technology