Alteration in shape, size and packing density of cellular and sub cellular microstructure affect the diffusion of water, so the diffusion MRI (dMRI) signal ; however traditional dMRI parameters are not sensitive enough to probe specific microstructural components. Modeling of dMRI data has successfully detected changes in cellular and sub-cellular microstructure, for example neurite density (ν), axonal water fraction, intra and extra axonal diffusivity, to name a few (1, 2, 3). Neurite density parameters are sensitive and potentially useful in disease, diagnosis and monitoring treatment outcome (1, 2). However, histological validation of these findings is limited, but highly valuable, and may provide more pertinent explanations of parameter alteration as well as to establish the usefulness of model parameters (4). Although histology is considered a gold standard, correlating MRI to histology is often unfeasible because it requires extensive time and labor, and is thus a low throughput technique. Here we employed morphological operation tools included with the image processing toolbox of Matlab to compute histology based neurite, nuclear and glial volume fractions from the amygdala region of control and chronic mild stress (CMS) rat brains. The novel image processing and analysis framework can process and analyze the images robustly with high throughput. Twenty four adult wistar rats (Taconic, Denmark) were used in the present study. The CMS paradigm was implemented (5) and distinguished as anhedonic and resilient on the basis of sucrose consumption test, described elsewhere (5). Brains were perfusion fixed and post-fixed in buffered formalin. After ex-vivo dMRI (TR/TE =6500 /26 ms, Δ/δ 15/5 ms, 12 directions, 14 b-values (0-8000 s/mm2) and 250 µm isotropic resolution), the same hemispheres were sectioned into 60 µm thick sections on a vibratome. One set of tissue sections were double stained with lipophilic dye (DiI), and nuclear stain (Hoechst). Another set of tissue sections were immunostained with ALDH1L1, an astrocyte marker. Image stacks in 3D from the amygdala region were acquired from DiI and Hoechst stained sections on a confocal microscope (Figure 1a, 1d). Multiple images were captured on a light microscope from the amygdala region of immunostained (ALDH1L1) sections (Figure 1g). Images in 3D underwent depth dependent intensity correction and were subsequently imported as ‘tiff’ images in Matlab, followed by contrast enhancement (Figure 2b, e, h). Contrast enhanced images were thresholded to binary images and morphologically opened with rectangular structuring elements. Binary images were compared with contrast enhanced image to visually confirm successful delineation of targeted microstructure (Figure 2c, f, i). Pixels in the binary images were counted and normalized to approximate neurite, nuclear and astrocyte volume fractions. MRI and histological data were separately fit in Matlab to a linear mixed effect models with animals as random effects and group as a fixed effect. Figures 2 and 3 show estimated means with confidence intervals (95%) providing an estimate of fixed effect size and variability. F tests were used to assess the level of significance. Biophysical modeling of dMRI data based neurite density model showed significant increase in neurite density only in amygdala in anhedonic (p = 0.008) and resilient groups (p = 0.011) (Figure 2). Histology based neurite density was also significantly increased in anhedonic and resilient groups (p = 0.018 and p = 0.048) respectively, in comparison to control, while nuclear and astrocyte densities showed modest non-significant decrease in both stressed group (Figure 3).The present study demonstrates histological validation of dMRI based neurite density increase in the amygdala of CMS rat brains. Histological validation of MR based findings establishes the microstructural underpinnings of the model parameters and provides a clear understanding of microstructural alterations driving observed parameter. Despite the fundamental difference between the MRI and microscopy modalities, previous quantitative analysis has shown a significant correlation between the two (2, 4). Here, we present a novel image processing and analysis framework of histological data to validate dMRI derived neurite density. Our analysis comparing amygdala from stressed rat to normal controls provides results in agreement with a previous neuronal tracing study which found dendritic hypertrophy in amygdala after chronic stress (6). The modest decrease in nuclear and astrocyte volume fraction reported here is also supported by previous stereology based findings in major depressive disorder (7). Biophysical modelling combined with dMRI revealed increases in neurite density in the amygdala of stressed rats. We developed novel robust and high throughput image processing and analysis tools validating these findings with histology and immunohistochemistry. These techniques can be applied to validate specificity of MRI model parameters in an unbiased and efficient manner.