Major depressive disorder (MDD) is a debilitating illness and characterized with abnormalities of mood, cognitive, and even motor abnormalities. This study was designed to investigate the longitudinal effect of sertraline on the findings on MR imaging. We hypothesized that a predictive MR imaging model of efficacy of sertraline treatment could be established to help treatment planning for patients with MDD.Patients aged from 20 to 65 years old with satisfactory DSM-IV criteria for MDD and age- and gender-matched healthy subjects were recruited. Patients took sertraline 25mg/day orally for 6 weeks. All patients underwent MR imaging in a 3-T MR scanner (Discovery MR 750, GE Healthcare) within one week before the start of sertraline treatment and within one week after completion of the sertraline treatment. T1-weighted volumetric images were processed using SPM8. For follow-up data preprocessing, longitudinal pipeline are utilized to rise up the accuracy and the ability in detecting subtle changes between baseline and follow-up image in each subject. Firstly, images (baseline and follow-up) were first realigned together to remove motion between time points for each subject. After the initial alignment, mean realigned image is calculated and used as a reference image in creating group specific DARTEL template in MNI coordinate. All realigned images are then corrected for signal inhomogeneous bias with regard to the mean image as references. Normalization parameters are then applied to gray-matter (GM) segments of bias-corrected images. All normalized, segmented, and modulated MNI standard space images were then smoothed with an 8-mm Gaussian kernel before voxelwise group comparisons. Segmented tissue volumes were estimated in cubic millimeters by counting the voxels representing the native space of the GM, white matter, and cerebrospinal fluid. The coordinates of each significant cluster were transformed from MNI coordinates into Talairach coordinates by using the GingerALE toolbox. The anatomical structures of the coordinates representing significant clusters were identified using the Talairach and Tournoux atlas. The mean volume of each significant cluster was extracted for each participant and further correlated to cognitive assessments in the whole study group. A total of 62 subjects were recruited. Among those, 26 patients (8 man and 18 women) were diagnosed with major depression (mean range, 39.9 years; SD, 10.8) while 36 (12 man and 24 women) were healthy controls (mean range, 39.7 years; SD, 10.5). Table 1 shows volume difference of the GM volume between patients and healthy controls. The GM volumes of the bilateral inferior temporal gyri and left inferior frontal opercularis in the patients with major depression were significantly smaller than those in healthy controls (Figure 1). After six week of sertraline treatment, patients showed atrophic change of brain volume in the left medial frontal gyrus, left parietal lobule, right postcentral gyrus, right middle temporal gyrus, and right middle occipital gyrus (Table 2 and Figure 2).The lingual gyrus is primarily involved in the process of vision and thought to play a role in the analysis of the logical conditions and encoding visual memories. It locates between the calcarine sulcus next to the fusiform gyrus with a direct connection to the limbic system.¹ The lingual gyrus has been shown to being larger in antidepressant responsive patients as compared with non-responsive patients and healthy controls.² The activity of the lingual gyrus is mediated by a top-down attention control system, in which the frontoparietal network is involved.³ Our results showed increased volume of the lingual and occipital gyri in patients with MDD as compared with those in healthy controls and the occipital gyrus showed decreased size after 6-week treatment of sertraline. In a MR imaging study of Desseilles et al., an abnormal filtering of irrelevant information in visual cortices and altered functional connectivity between frontoparietal network and visual cortex were found, which might suggest an abnormal visual processing interfered with recruitment of attention/executive neural network in patients with MDD.⁴ Hence, the top-down influences from the frontoparietal network might be abnormal in patients with MDD with abnormal neural processing the visual cortices. The association between the lingual gyrus and neuropsychological features of MDD in our study supported that patients with MDD might have a functional abnormality of visual areas and antidepressant treatment might shift the abnormal activity in the antidepressant-susceptible brain region to a normal level.Our results showed increase volume of the lingual and occipital gyri in patients with MDD as compared with those in healthy controls and the occipital gyrus showed decreased size after 6-week treatment of sertraline.