Iron quantification in Parkinson's disease using an age-based threshold on susceptibility maps: The advantage of local versus entire structure iron content measurements.

BACKGROUND: Elevated brain iron has been observed in Idiopathic Parkinson's disease (IPD) within the deep gray matter. Using quantitative susceptibility mapping (QSM) and a thresholded high-iron region, we quantified iron content in the midbrain of patients with Parkinson's disease as a function of age. METHODS: We used MRI to scan 24 IPD patients at 3-Tesla. Susceptibility-weighted images were collected with the following parameters, TE: 6 and 20 ms, TR: 30 ms, FA: 15°, and resolution: 0.5 × 0.5 × 2.0 mm3. QSM images were reconstructed from the source phase images. Whole-region and thresholded high-iron (RII) region boundaries for the Substantia Nigra (SN) and Red Nucleus (RN) were traced. Iron content was measured via mean susceptibilities and volumes, which were compared between the groups, as well as between right and left side of the structures within groups. RESULTS: Twenty patients with mild to moderate IPD were used in this study. For the SN, mean RII and whole-region iron and volumes were higher in the IPD group compared to HC, as well as mean RII for the RN, while no differences were seen between the groups when considering whole-region mean susceptibility bilaterally for the RN. CONCLUSION: Using a two-region of interest analysis on QSM, we showed that abnormal iron occurs in IPD patients in the SN and with greater volumes compared to HC. This method may have application as a biomarker for disease diagnosis and early intervention.

Computationally rapid method of estimating signal-to-noise ratio for phased array image reconstructions.

Measuring signal-to-noise ratio (SNR) for parallel MRI reconstructions is difficult due to spatially dependent noise amplification. Existing approaches for measuring parallel MRI SNR are limited because they are not applicable to all reconstructions, require significant computation time, or rely on repeated image acquisitions. A new SNR estimation approach is proposed, a hybrid of the repeated image acquisitions method detailed in the National Electrical Manufacturers Association (NEMA) standard and the Monte Carlo based pseudo-multiple replica method, in which the difference between images reconstructed from the unaltered acquired data and that same data reconstructed after the addition of calibrated pseudo-noise is used to estimate the noise in the parallel MRI image reconstruction. This new noise estimation method can be used to rapidly compute the pixel-wise SNR of the image generated from any parallel MRI reconstruction of a single acquisition. SNR maps calculated with the new method are validated against existing SNR calculation techniques.