Structural abnormalities in the substantia nigra have been reported in patients with Parkinson’s disease (PD) [1-2]. With the ability to probe tissue microstructures and microenvironment, high b-value diffusion imaging can be very useful to study structural changes in Parkinsonian patients. Unfortunately, conventional diffusion image techniques rely on single-shot EPI which has poor spatial resolution and excessive geometric distortion. These can be particularly problematic for imaging substantia nigra in the brainstem where the magnetic susceptibility-induced geometric distortion is excessive and high-resolution is needed to visualize the small structure. These technical challenges can be addressed using reduced Field of View (rFOV) techniques [3-5]. The goal of this study is to investigate structural abnormalities in the substantia nigra on PD patients using a combination of an rFOV technique developed recently [5] and a novel diffusion model – the continuous-time random-walk (CTRW) model.Subjects: With approval of the Institutional Review Board, 30 clinically confirmed PD patients were included in the study (20 males, 10 females, age ranges from 60 to 80) and 10 subjects (6 males, 4 females, age ranges from 29 to 76) were recruited as healthy controls. Three patients (2 males, 1 female) were excluded from data analysis because of excessive head motion, resulting in a subject group comprising 27 patients (9 females) and 10 healthy controls (no exclusion). Image Acquisition: All subjects underwent diffusion MR scan on a 3T GE Signa scanner with an 32-channel head coil (Nova Medical, Inc.,Wilmington, MA). High b-value diffusion images were acquired using an rFOV sequence [5], with six b-values (50, 200, 500, 1000, 2000, 3000 $$$sec/mm^{2}$$$). The other parameters were: TR/TE=30800/86ms, slice number=26, slice thickness=3mm, Δ=47 ms, δ=32.2ms, FOV=10cm×6cm, and matrix size=80×48. Trace-weighted images were used to reduce the effect of diffusion anisotropy. Image and Statistical Analysis: According to the CTRW model, the signal intensity in diffusion-weighted imaging decays according to a Mittag-Leffler function (MLF): $$M(b)=M_{0}E_{\alpha}(-(bD_{m})^{\beta})$$ where Dm is the anomalous diffusion coefficient, α and β correspond to temporal and spatial diffusion heterogeneity, respectively [6-7]. Using this equation, the CTRW model was fit to the multi-b-value diffusion images, producing parametric maps, Dm, α and β, on a pixel-by-pixel basis. The region of interest (ROIs) were drawn on both sides of the brainstem on the b=0 images on each subject. The mean and standard deviation of each parameter in the ROIs were calculated and compared between the PD patients and the healthy controls. A 2-tailed Student’s t-test was then performed for all CTRW parameters. For comparison, the conventional ADC values were also computed from the same ROIs.Figure 1 shows the maps of Dm, α, and β, as well as the diffusion weighted image at b value of 1000 $$$sec/mm^{2}$$$, from a representative patient. Figure 2 shows the boxplots of the mean values of the CTRW parameters and ADC value for PD patients and healthy controls. Figure 2 indicates that α can differentiate the healthy and PD patient groups with a p-value < 0.05, which outperformed ADC (p=0.19). Figure 3 displays a boxplot of the mean values of α for PD patients and healthy controls by analyzing the male and female groups separately. A larger difference between female patients and healthy controls was observed compared to the male sub-group. Table 1 summarizes the descriptive statistics, mean and standard deviation values, and the p-value of the CTRW parameters and ADC.Our results showed that the rFOV pulse sequence can be successfully combined with the CTRW diffusion model for studying structural abnormalities in the brainstem of PD patients. Comparing to the conventional ADC value in a mono-exponential model, the α parameter in the CTRW model offers an advantage in identifying abnormalities in the substantia nigra in PD patients relative to the healthy controls. A major limitation of this study is the small number of healthy subjects. With further validation, the α parameter in the CTRW model may become a useful marker for PD-related brain structural alterations.