Classification of  Intracranial Hemorrhage  CT images for Stroke Analysis with Transformed and Image-based GLCM  Features

Abstract

Intracranial hemorrhage (ICH) is one of the severe types of brain stroke. Artery burst results in bleeding inside the brain and its surrounding tissue. Depending upon brain bleed location, hemorrhage gets classified. This paper presents a hybrid feature selection approach to form joint feature vector sets using transformed image features and image-based gray level co-occurrence matrix (GLCM) texture features. Feature extraction is performed by applying discrete wavelet transform, discrete cosine transform, and GLCM features. Feature vector selection model is built with the combination of Discrete wavelet transform, Discrete cosine transform, and GLCM features. A joint feature vector is formed with transformed image features and GLCM features. These joint feature vectors form an acute input for the further classification process. The machine learning algorithm i.e. Random Tree, Random Forest, and REPTree are used for the classification of Intracranial hemorrhage CT images. The classification results obtained are further analyzed for accuracy considering transformed features and GLCM features. It is observed that Random forest classifier results in the highest classification accuracy of 87.97% for discrete wavelet transform and GLCM (DWT+GLCM) feature set

Innovations in Stroke Identification: A Machine Learning-Based Diagnostic Model Using Neuroimages

“…In [18], a hybrid feature selection approach uses transformed CT image features and grayscale co-occurrence matrix texture features. Features are extracted using discrete wavelet transforms, discrete cosine curves, and GLCMs.…”

An automated system for classifying types of cerebral hemorrhage based on image processing techniques

“…Moreover, the U-Net model achieved a 60% score, effectively discerning discrepancies between ischemic and hemorrhagic strokes [8]. Gudadhe and Thakare [9] developed a hybrid feature selection approach to extract a common feature vector from CT images by applying discrete wavelet transforms (DWT), discrete cosine transforms (DCT), and gray-level co-occurrence matrixes (GLCM) in a CT image. The CT images of intracranial hemorrhage are then classified by running random tree, random forest, and REP-Tree machine learning algorithms.…”

“…The CT images of intracranial hemorrhage are then classified by running random tree, random forest, and REP-Tree machine learning algorithms. The classification results showed that the highest accuracy was for the random forest classifier, which reached 87.97% for the DWT and GLCM feature sets [9]. The novelty of this work lies in using Orange3 as a new data mining method to extract a huge number of efficient features from CT images of patients with cerebral hemorrhage and employing these features in the cerebral hemorrhage classification process by applying four classification algorithms: logistic regression (LR), convolutional neural networks (CNN), support vector machine (SVM), and k-nearest neighbor algorithm (KNN).…”

An efficient stacked bidirectionalGRU‐LSTMnetwork for intracranial hemorrhage detection

“…25 Santwana and Anuradha proposed a hybrid feature group technique using discrete wavelet transform and image-based gray-level cooccurrence matrix (GLCM) that achieves 87.89% classification accuracy for a random forest classifier. 26 Pérez Del Barrio et al proposed a novel hybrid architecture based on image input from a 3D CNN and a feed-forward network for clinical data input. The model final component combines the outputs of the feed-forward and convolutional blocks and feeds this hybrid input to the final dense and activation layers to detect the presence of ICH.…”

“…On the collected dataset, the author performed validation by comparing the results of his proposed model with existing architectures, such as MobileNetV2 and VGG16 25 . Santwana and Anuradha proposed a hybrid feature group technique using discrete wavelet transform and image‐based gray‐level co‐occurrence matrix (GLCM) that achieves 87.89% classification accuracy for a random forest classifier 26 . Pérez Del Barrio et al proposed a novel hybrid architecture based on image input from a 3D CNN and a feed‐forward network for clinical data input.…”