A Hybrid CNN-GLCM Classi�er for Detection and Grade Classi�cation of Brain Tumor

using Deep and segmentation of the brain tumor in The Deep Net architecture extracts the features internally from enhanced image and classifies them into normal and abnormal tumor images. The segmentation of tumor region is performed by the global thresholding with area morphological function method. This proposed method of fully automated classification and segmentation of brain tumor the spatial invariance and inheritance. on its feature the proposed CNN Deep net classifier, classifies the detected tumor image either to (low grade)benign or (high grade)malignant. proposed Deep net classification methodology approach with grading system both quantitatively and qualitatively. The quantitative measures sensitivity, specificity, accuracy, Dice similarity coefficient, precision, F-score of the proposed classification and segmentation methodology a segmentation accuracy of with to


Introduction
A tumor is a volume of irregular and abnormal cells affecting the function of nearby healthy cells in a human body. Meningioma is the most commonly occurring tumor in adults with high risk factors.
Meningiomas are seen in the dura mater, which are the outer tissues of the brain and called meninges. III is dangerous, and it occurs in fewer numbers. But the mitotic rate of Grade III is higher than 18 and patients are at a high risk of death. In general, the diagnosis of brain tumor begins with magnetic resonance imaging (MRI) as it provides detailed information on both hard and soft tissues with fat and fluid substances of the brain through electromagnetic fields when compared to other imaging modalities like CT, X ray, PET etc. A specific pattern or a regular range of radiation intensity is obtained when human brain is exposed to radiation by different modalities. The NCIS and WHO has reported that every year around 13,000 people are affected by the tumor. Every year the death rate is progressively increasing because of the late diagnosis and semi-automated grading system. The detection and classification of tumors by manual method is the greatest challenge which pathologist's face. According to the reports published in Cancer.Net and WHO 2019, automated classifier architecture for the detection and classification of tumor is an emerging research area in the field of medicine. This provokes number of researchers to develop a cost effective as well as more precise automated algorithm for detection, classification and diagnosis of tumor.
After the emergence of machine learning techniques in medical imaging analysis the detection and classification approaches has become easier and effective in terms of accuracy. An image analysis scheme of CNN Deep Net is proposed for the detection and classification of meningioma tumors with high accuracy and classification rate. This article uses the standard medical image processing sub modules such as preprocessing ,feature extraction ,classification and segmentation for detection and localization of tumor regions in brain MR images.In this regard ,the article is organized as follows: In section 2 ,various existing brain tumor classification and segmentation approaches using deep learning ,machine learning are discussed. In section 3, novel brain tumor detection, classification and segmentation methodology is proposed and section 4 discusses the results of the proposed tumor method. Finally, section 5 concludes the work with future scope.  2017) proposes a network to identify affected tumor regions in brain using deep learning approaches. The method applies DWT for segmentation of detected image and uses PCA feature extraction method. This non-linear approach produced a significant simulation result of 96.7 %. Rajagopal (2019) proposes a ML approach to find the brain tumor regions by energy technique and segmented the affected regions .In addition, classification is also performed using random forest classification approach and achieves 98% accuracy. Abdelmajid Classification. With validation dataset and data augmentation, the proposed approach achieves 96.49% accuracy.

Contribution
Machine learning approaches for the brain tumor classification requires large number of extracted features for efficient classification. Thus different ML approaches uses a separate feature extraction method and classifies the tumor from normal brain MRI image. These feature extraction and classification approaches are not fully automated.

Materials
The performance of the proposed CNN Deep Net is observed on the MR brain images of open access BRATS datasets. This dataset includes ground truth images that are obtained from expert radiologist. This paper access 600 brain images from the dataset and it has both normal brain images (340 brain images) and abnormal brain images (260 brain images). This dataset is grouped into training and testing set. The training dataset contains 90 normal images and 75 abnormal images .The testing dataset contains 250 normal images and 185 abnormal MR images. The images used in the proposed work are T1 and T2 weighted sequence with contrast agent in axial position with 1000 slices per pixel resolution. As T1 weighted sequence shows better delineating anatomy and morphology, where T2 weighted images shows pathology to highlight the lesions well, in brain images.

Methods
The meningioma tumor prediction and classification in MR image is carried out using proposed CNN Deep net. All dataset images of 512*512 pixel resolution are resized into 256*256 pixels and achieves same dimensionality with scaling range of [1 1.2] . This preprocessed brain MRI image is classified either into normal or abnormal using the proposed CNN Deep net classifier .The tumor regions are segmented using global thresholding approach in fusion with connected component method. A combined GLCM and CNN classifier is proposed in this paper for diagnosis of segmented regions.

Preprocessing
The preprocessing technique includes resizing of brain image sources to uniform dimensionality images with flipping up /down directions and rotating images at ± angles and skewing etc through data augmenting. Thus data 6 | P a g e augmentation helps the proposed CNN Deep net architecture to attain a high accuracy and precision in classification.
The data augmentation of the proposed CNN Deep net is listed in the Table 1. Fig.3 shows the augmented results of input brain MR images.
. proposed methodology for the classification of 2D images. The structure of CNN is stated using convolution, max pooling, and fully connected layers. The entire internal architecture of CNN classifier is depicted in Fig. 4.

Convolutional Filters
The convolutional layer focuses with both higher and lower level features namely edge detection, smoothening and sharpening etc. The convolution of an input image with a feature detector called kernel filter is carried out across the image through the sliding window method. As the depth of the filter is equal to the input image, the proposed method uses 3x 3 kernel filter .An element wise multiplication is done and added up to give feature maps ,which are the output of convolutional layers. The value of feature map is the value of matched input image and filter value .Hence the filter map is a 2D matrix. They are also known as activation maps or convolved features.
The convolution function is depicted in the following equation.
The symbol * indicates convolution process and the equation represents percentage of filter (J) area overlapping with input image (I) at a time τ .
Applying the tighter bounds to the integral the equation can be rewritten as The equation corresponds to single entry in 1D to compute complete convolution tensors .The multidimensional kernel tensors are required and it is given by, The convolution slides the kernel over the input image .It is mathematically given by the above equation and the output of equation will be scalar value. This process is repeated for every xy pixels and results are stored in a convolutional matrix called feature maps. The main purpose of this convolutional layer is to reduce the size of input image for the upcoming fast process. The performance of CNN architecture can be improved with the use of consecutive convolutional layer with different filter .To obtain more number of learned features for the image, more convolutional layers along with filters are used and thereby spatial preservation is achieved effectively. Thus to be simple and sharp, in convolutional layer the feature map extracts the features that are important and essential for classification. The spatial relationship between pixel, called spatial invariance, hierarchical feature learning and scalability are preserved in convolutional layer. The convolved feature or activation maps are incorporated with ReLu function to increase non linearity of the network. Recognition of objects or images using ReLu is the transition of pixels with adjacent pixels and so images are processed even when it is rotated, tilted or shifted.

Polling layers
The pooling of features can be done through the pooling layers which is actually a down sampling process. Here high data pixels obtained from convolved feature are minimized to a level that is suitable for further processing in  The main advantage of spatial distortion and size reduction of 75% is achieved through max pooling layer. Thus in this paper a Max pooling filter with 2x2 filter mask is used and shows better performance in classification of images.
This max pooling filter is applied to the result of the convolutional layer .It is observed that the depth of feature map after applying pooling also remains same with non overlapping features. Hence the proposed method eliminates the chance of over fitting occurrence, and reduces the number of parameters to 25 % when compared to the original number of parameters obtained from the input MRI brain image.

Fully connected layer
The obtained convolved features are pooled and it is required to be flatten sequentially into a single column. This

Proposed CNN Deep Net
An efficient and supervised deep learning architecture is proposed for the classification of brain tumors. The

Segmentation
The classified meningioma brain tumor image is segmented to obtain dilated and eroded images by global opening and closing functions respectively. These two images are subtracted to obtain threshold image. In order to improve the segmentation process, an area morphological function is utilized in threshold image to eliminate the missed pixels.Here, an important process of image analysis called labeling is carried through connected component method    11) where, P(i, j) is the GLCM matrix constructed at the orientation of 90 degree, the rows and columns of the GLCM matrix are represented as i and j. The number of rows and columns in GLCM matrix are represented by P and Q, respectively. Fig. 10(a) and Fig. 10(b) shows the proposed GLCM-CNN classifier results for low grade brain MRI images and high grade brain images.

Results and Discussion
In this article, the proposed CNN Deep net detection and classification methodology is applied on the set of brain MR images in the BRATS, an open access dataset. From the dataset, 600 brain images of 512*512 pixels are used in this article which includes 340 normal brain images and 260 tumor affected brain images. MATLAB R2020 b is used as simulation tool for detection, classification and segmentation of tumor region with Intel core i5 processor with 2.30 GHz and 16GB RAM memory. Table 3 shows the performance metrics of proposed CNN Deep net classifier on BRATS brain MR image datasets.
The validation accuracy, validation loss and training loss with different epochs are obtained from simulation results and the average accuracy of proposed Deep net is tabulated as shown.   regularization,normalization and 50% dropout with 0.01 learning rate are performed as hypertuning. The evaluation metrics for the performance analysis of the proposed method is stated through a confusion matrix which has the value of TP and TN, providing correctly detected tumor and correctly detected non tumor pixels as well as FP and FN to provide non-correctly detected tumor and non-correctly detected non tumor pixels . Table  4 is illustrated from confusion matrix and listed as sensitivity (Sen),Specificity(Sp),Accuracy(Acc),Precision(Pr),F-score, Dice Similarity Index (DSI).  Table 5, the meningioma brain tumor detection using CNN Deep Net architecture provides higher classification rate while compared with the brain tumor detection system using conventional CNN architecture.  Table 6 is the comparisons of proposed simulation results with other conventional methods on same dataset images.
It is inferred from these comparisons, the proposed system using CNN Deep net classification approach yields high simulation values when compared with other conventional methods on same dataset brain MRI images.  Table 7 with circumference (C), perimeter (P), area(A) and eccentricity(Ecc) .   Table 9.

Discussion
It is observed that the metrics sensitivity, specificity, accuracy, precision and F-score of the proposed CNN Deep Net works better when compared with the conventional machine learning methods.

Conclusion
In this paper, detection of meningioma brain tumor and classification by CNN Deep net is proposed. The CNN Deep net architecture is designed with five convolutional layer with ReLU activations, Max pooling layers and multi neuron feed forward neural network to obtain deep features from input brain MRI images. The proposed methodology uses multilayer perceptron architecture for detection and classification of brain tumor with grade classification .Global threshold segmentation approach is used for segmenting the detected tumor affected region where both dilation and erosion are used for locating the tumor regions. Further, a novel diagnosis system using GLCM CNN classifier is proposed and achieves a high classification rate and accuracy. Open access dataset BRATS is used for the analysis and evaluation of the proposed method. Thus the proposed CNN Deep net architecture has achieved higher classification rate of 99.5% with better specificity ( 98.6%) and sensitivity (97.2%).The Deep net achieves a segmentation accuracy of 99.4% where the diagnosis of segmented tumor regions achieves greater diagnosis rate of 98.3 % .

Conflict of interest
The authors declare that they have no conflict of interest.    The ow of proposed method of CNN Deep net methodology for classi cation of tumor and non tumor brain MRI images   Performance analysis of proposed CNN Deep net for classi cation of brain tumors images with different epoch

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. BIBChecklist1.docx