Investigation of the effect of "Nicotiana rustica/Maras Otu" use on gray matter using image processing techniques from brain MRI images

In this study, it was investigated on the brain gray matter whether the effect of Nicotiana rustica, which is widely used in Kahramanmaras and its environs and user age can be lower than secondary school, that harms are not clearly revealed, which can be bought quite cheaply and contains more nicotine than cigarettes. Segmentation of parts of the brain such as gray matter, white matter, and cerebrospinal fluid (CSF) with SPM and calculation of their volume were performed with the Voxel-Based Morphometry (VBM) interface. Segmentation of images in the study was carried out using the Gaussian Mixture Model. With these processes, gray matter volumes of users and non-users of Nicotiana rustica were calculated and classified using classical machine learning and deep learning techniques. A data set was created for those user of Nicotiana rustica and non users, with the volumes of gray and white matter, CSF volume and their ratios to each other obtained with VBM. This data set has been classified using classical machine learning methods (Naive Bayes, J48, kNN, Support Vector Machines, Random Forest), as well as deep learning methods (1D-CNN, CNN + LSTM). Methods classification success for classical methods, respectively: 70%, 85%, 90%, 75%, 95%; for deep learning methods: 85% and 90%, respectively. According to our knowledge, there is no study in the literature that makes classification using various information of people who use Nicotiana rustica. This is the original aspect of the study and contributes to the literature.