China
Africa
Explore chapters and articles related to this topic
Neuroimaging in Nuclear Medicine
Published in Michael Ljungberg, Handbook of Nuclear Medicine and Molecular Imaging for Physicists, 2022
Anne Larsson Strömvall, Susanna Jakobson Mo
A common type of benign brain tumour is meningioma, arising from cells in one of the membranes (meninges) around the brain. There are several types of malignant primary brain tumours. These may arise from any cell type in the brain. For example, a type of malignant tumour occurring in children and called medulloblastoma, arises from nerve cells. However, the largest group of brain tumours in adults are called gliomas, arising from different kind of supporting cells in the brain, the glia cells. There are different types of gliomas. For example, a subtype of glioma called astrocytoma and glioblastoma multiforme arise from astroglia cells, and oligodendrogliomas arise from glia cells called oligodentrocytes. Gliomas may also be classified according to malignancy, as in the WHO-classification [1], which by characterization of specific features in the tumour cells separates them into four grades of malignancy; a WHO grade I tumour is benign and a grade IV tumour is highly malignant.
Image Edge Detection Using Fractional Conformable Derivatives in Liouville-Caputo Sense for Medical Image Processing
Published in Devendra Kumar, Jagdev Singh, Fractional Calculus in Medical and Health Science, 2020
J. E. Lavín-Delgado, J. E. Solís-Pérez, J. F. Gómez-Aguilar, R. F. Escobar-Jiménez
A meningioma is a tumour that arises from a layer of tissue (the meninges) that covers the brain and spine [54]. Although most meningiomas are encapsulated benign tumours with limited numbers of genetic aberrations, their intracranial location often leads to serious and potentially lethal consequences [55]. By using an MRI image, it is possible to determine the location of meningiomas; however, the detection and accurate diagnosis of meningiomas can be drastically improved with the use of the proposed operator due to its high capacity to detect edges and textures. The T1-weighed MRIs shown in Figures 1.10through 1.13 were taken from [56-59], respectively. In these figures, the boundaries of the meningiomas are correctly defined even in noisy images. By using the fractional conformable Gaussian edge, it is possible to make a better identification of the meningiomas achieving a more accurate diagnosis. On the other hand, with the boundaries of the meningiomas being well-defined, it is possible to obtain a better estimate of their size. In conclusion, the fractional conformable operator is able to identify much better the edges in MRI images for meningioma identification than classic operators such as Sobel, Prewitt, Roberts, LoG, and Canny.
The Role of Multiparametric MR Imaging—Advanced MR Techniques in the Assessment of Cerebral Tumors
Published in Ioannis Tsougos, Advanced MR Neuroimaging, 2018
A meningioma is a tumor that arises from the meninges, that is the membranous layers surrounding the central nervous system. Although not technically a brain tumor (it is situated on the brain), it is included in this category because it may compress or squeeze the adjacent brain. Therefore, meningiomas are the most common extra-axial cerebral tumors, and due to their characteristic location a relatively straightforward diagnosis is achieved. However, although the majority of meningiomas are benign, they can have malignant transformations. According to the WHO classification system, there are three types of meningiomas. The majority of meningiomas (up to 90%) are benign or Grade I and usually full recovery is achieved with surgical resection. Grade II (atypical) and Grade III (malignant) meningiomas are less common but more aggressive than Grade I, thus they are more likely to recur even after complete resection. The differences between benign and atypical/malignant meningiomas relate to the number of mitoses, cellularity, and nucleus-to-cytoplasm ratio as well as their histologic patterns (Perry et al., 2007). Conventional MR imaging provides useful information regarding their localization and morphology, however there can be cases where meningiomas may have atypical imaging findings, like heterogeneous contrast enhancement and necrotic areas mimicking high-grade tumors. Hence, a correct diagnosis and accurate histologic grading is of great importance for beneficial treatment planning. An atypical/malignant meningioma with a large heterogeneous enhancement and intense mass effect is illustrated in Figure 9.8.
Multi-Channel CNN based image classification using SKIP connection and MSVM
Published in International Journal of Computers and Applications, 2022
We used the freely available dataset of brain tumor type images [35] for the analysis. The dataset consists of 3064 T1-Contrast Enhanced MRI images of 233 patients, including all the three orientations coronal, sagittal and axial. The slices consist of patients having glioma, meningioma and pituitary tumor. Glioma begins from glial cells found in the brain’s supportive tissue. Meningioma is a benign tumor found in the outer coverings of the brain. Finally, pituitary tumors are located at the pituitary gland. The tumor’s location helps to classify them, accordingly, as shown in Figure 4. 80% is used for training and 20% for testing from the total number of images. Of the 80% for training, 20% is again split up as a validation set. The input images are grayscale images of the size 512 × 512 pixels. Our input image dimension is designed to be 227 × 227 × 3. So, we need to resize the image and repeat it three times to get the appropriate input size for the architecture. It is seen that all the leading architectures have an input size dimension in the range of 224,227, etc. So, we adopted a similar range for better comparison with these existing systems.
Conditional random field-recurrent neural network segmentation with optimized deep learning for brain tumour classification using magnetic resonance imaging
Published in The Imaging Science Journal, 2023
Geetha M, Prasanna Lakshmi K, Sajeev Ram Arumugam, Sandhya N
Brain tumours emerge from the cells around the meninges (the brain membranes), nerves, or glands in the brain. They can unambiguously affect and damage the brain cells by causing increased pressure in the skull’s interior [10]. Brain tumour treatment can be performed if the tumour is identified at an initial stage. Brain tumours can be classified as malignant and benign, wherein the malignant tumour is devoid of any constant and organized uniformity. Generally, benign tumours have non-existent and rigid uniformity [11]. Further, brain tumours can be categorized as Pituitary, Glioma, and Meningioma. Meningiomas are commonly non-cancerous tumours that most commonly form in the thin walls surrounding the brain. Pituitary tumours are non-cancerous and originate in the pituitary glands [12]. Conversely, Gliomas are the main class of brain tumours that emerge from the glial cells and penetrate the neighbouring brain tissues, and they can be categorized into High-Grade Glioma (HGG) and Low-Grade Glioma (LGG). LGG has a slower growth rate, and with the help of appropriate treatment, the survival rate of the affected individuals can be prolonged. The successful treatment of brain tumours depends on the evaluation of the progression rate of the disease, considering the accuracy of the neuroimaging modalities [13]. Clinically, brain tumour is diagnosed based on the analysis of the imaging data of tumour images. Examining the tumour images is the most significant step in identifying the individual’s condition. But, the accuracy of the analysis depends on various factors, such as visual fatigue, variations in the experiences, and knowledge of the medical professionals [14–16]. Thus, the most significant process in treating brain tumours is to decide the detection process [17]. MRI can offer numerous data [18, 19] concerning the position, dimensions, and shape of the body’s organs and tissues without harmful ionizing radiation. Further, the efficiency of the diagnosis is greatly enhanced, as the images are highly precise and clear and can be effectively used for localizing the lesions [20].