Meditation can be conceptualized as a family of complex emotional and attentional regulatory practices, that can be classified into two main styles: focused attention (FA) and open monitoring (OM)^1,2. While FA meditation entails the capacities of monitoring the focus of attention and detecting distraction^1,2, OM encompasses an attentive set that is characterized by an open presence and a non judgemental awareness of sensory, cognitive and affective fields of experience in the present moment^1,2. Here we want to describe how experience modulates ROI-based fMRI functional connectivity patterns in two different meditators groups: experts and novices. We used a pattern classification approach^3,4 to discriminate these groups and find which connections and nodes are important to classify subject experience.Participants included 12 Theravada Buddhist monks (males, mean age 37.9 years, SD 9.4 years) and 8 novice meditators (males, mean age 32 years, SD 3 years).The novice participants were given oral and written instructions on how to perform Samatha (FA) and Vipassana (OM) meditation styles⁵. The FA–OM experimental paradigm consisted of 6 min FA and 6 min OM meditation blocks, each preceded and followed by a 3 min non-meditative resting state block, for three times; conditions were switched by an auditory signal. Data acquistion was performed on a Siemens Magnetom Vision Scanner at 1.5 T, BOLD signal images were obtained using T2*-weighted echo planar (EPI) (TR = 4.087 s, voxel size = 4x4x4 mm3) for a total of 860 images each containing 64x64x28 voxels. Preprocessing consisted in a motion and time correction, temporal filtering (high-pass filter of two cycles per time course) and linear and non-linear trend removal; the first five scans of each run were discarded due to T1 saturation effect. Connectivity preprocessing consisted in white matter, CSF regression, and temporal filtering (band-pass filter 0.08 – 0.009 Hz), then timecourses were extracted, for each meditation session, from ROIs defined by AAL90 atlas⁶ and the correlation matrix between ROIs computed. The value of the connections were used as feature for the classifier while subjects were the samples, for each subject we computed 3 matrices per meditation style. Subject number within groups (i.e. Monks and Novices) were different, resulting in an unbalanced dataset; we therefore decided to analyze the dataset using two different approaches⁷: down-sampling, and classifier weighting. We divided the dataset in 4 folds for each group and we down-sampled the larger class within each fold; this procedure was repeated 100 times, each dataset was analyzed using a linear SVM, with an ANOVA Univariate Feature Selection that picked 1% of most significant features in a k-fold cross validation procedure with k = 2; permutation tests (n=2000) were used to assess statistical significance.Decoding results were above chance for both Vipassana (average accuracy across balanced datasets: 60.70%) and Samatha (63.60%). We tested with a one sample t-test against chance level if the accuracy distribution of each of 100 balanced datasets was significative (OM p<0.001; FA p<0.001), moreover we tested the statistical significance of the average accuracies for each style using permutation test⁸ finding significant values (OM p<0.05; FA p<0.05). We also found above chance classification for accuracy obtained with classifier weighting (OM 61.67% p<0.05 through permutation test; FA 66.67% p<0.01). Connections that drive the classifier were found summing the number of times a connection is chosen for each balanced dataset, then we selected only the connections chosen with a value higher than mu+2std and we use the classifier weights as magnitude of connection importance. In Figure 1 and 2 are plotted the feature choice frequency and the feature weights, our results suggests that Amygdala, Putamen and Insula nodes play an important role to discriminate experts and novice medidators during FA style. Several studies ^9,10,11 have shown that these regions are involved during attention control. Superior Frontal Gyrus is an important node to discriminate these categories during OM meditation, this node has been demonstrated to be highly involved in self-awareness¹².We showed how pattern classification and functional connectivity could be used to detect fine-grained differences in fMRI connectome of different meditators categories. Regions having a role in decoding are those implicated in self-awareness and attention control.