Simple baseline correction for 1H MRSI data of the prostate

Area of Science:

• Medical Imaging
• Spectroscopy
• Oncology

Background:

• Proton magnetic resonance spectroscopic imaging (sMRI) detects prostate tumors.
• Pattern recognition in sMRI data can differentiate tumor from normal tissue.
• Accurate baseline is crucial for sMRI analysis, but lipid artifacts can interfere.

Purpose of the Study:

• To develop a simple and effective baseline correction method for prostate sMRI.
• To address the challenge of lipid resonances in prostate sMRI data.
• To improve the reliability of sMRI for prostate cancer diagnosis and treatment monitoring.

Main Methods:

• A novel baseline correction technique using subtraction of a single simulated-Lorentzian resonance was proposed.
• Prostate sMRI data acquisition was optimized for a flat baseline with long echo time and water/fat suppression.
• The proposed method was tested on a dataset of prostate sMRI spectra.

Main Results:

• The proposed method successfully restored flat baselines to the test spectra.
• Lipid artifacts, particularly at the prostate margins, were effectively removed.
• The performance of the new method favorably compared to a state-space modeling approach.

Conclusions:

• A simple simulated-Lorentzian resonance subtraction method is effective for baseline correction in prostate sMRI.
• This technique enhances the diagnostic accuracy of sMRI for prostate cancer.
• The method offers a valuable tool for improving prostate cancer detection and treatment monitoring.