Classification of CT brain images based on deep learning networks

Abstract

While Computerised Tomography (CT) may have been the first imaging tool to study human brain, it has not yet been implemented into clinical decision making process for diagnosis of Alzheimers disease (AD). On the other hand, with the nature of being prevalent, inexpensive and non-invasive, CT does present diagnostic features of AD to a great extent. This study explores the significance and impact on the application of the burgeoning deep learning techniques to the task of classification of CT brain images, in particular utilising convolutional neural network (CNN), aiming at providing supplementary information for the early diagnosis of Alzheimers disease. Towards this end, three categories of CT images (N=285) are clustered into three groups, which are AD, Lesion (e.g. tumour) and Normal ageing. In addition, considering the characteristics of this collection with larger thickness along the direction of depth (z) (∼3-5mm), an advanced CNN architecture is established integrating both 2D and 3D CNN networks. The fusion of the two CNN networks is subsequently coordinated based on the average of Softmax scores obtained from both networks consolidating 2D images along spatial axial directions and 3D segmented blocks respectively. As a result, the classification accuracy rates rendered by this elaborated CNN architecture are 85.2%, 80% and 95.3% for classes of AD, Lesion and Normal respectively with an average of 87.6%. Additionally, this improved CNN network appears to outperform the others when in comparison with 2D version only of CNN network as well as a number of state of the art hand-crafted approaches. As a result, these approaches deliver accuracy rates in percentage of 86.3, 85.6±1.10, 86.3±1.04, 85.2±1.60, 83.1±0.35 for 2D CNN, 2D SIFT, 2D 1 KAZE, 3D SIFT and 3D KAZE respectively. The two major contributions of the paper constitute a new 3-D approach while applying deep learning technique to extract signature information rooted in both 2D slices and 3D blocks of CT images and an elaborated hand-crated approach of 3D KAZE.

Applications of Deep Learning and Reinforcement Learning to Biological Data

“…Due to its facilitating structure, CNN has been the most popular DL architecture for image analysis. CNN was applied to classify breast masses from mammograms (MMM) [151][152][153][154][155], diagnose AD using different neuroimages (e.g., brain MRI [126], brain CT scans [135], and (f)MRIs [128]), and rheumatoid arthritis from hand radiographs [150]. CNN was also used extensively: on CT scans to detect-anatomical structure [136], sclerotic metastases of spine along with colonic polyps and lymph nodes (LN) [137], thoracoabdominal LN and interstitial lung disease (ILD) [139], pulmonary nodules [138,140,141]; on (f)MRI and diffusion tensor images to extract deep features for brain tumor patients' survival time prediction [129]; on MRI to detect neuroendocrine carcinoma [127]; on UlS images to diagnose Breast Lesions [138] and ILD [147]; on CFI to detect hemorrhages [148]; on endoscopy images to diagnose digestive organ related diseases [149]; on PET images to identify oesophagal carcinoma and predict responses of neoadjuvant chemotherapy [143].…”

A Generic Framework for Task Selection Driven by Synthetic Emotions

“…where the E α are defined in (2). The global objective, to align character C A and experience E A , is achieved by minimizing the Kullback-Leibler divergence between C A and E A with respect to the q α .…”

Comparative study of performing features applied in CNN architectures

“…There are many techniques available for diagnosis of brain tumor from the brain tissues, detection of brain tumor such as conventional radiology, ultrasonography, magnetic resonance imaging, computerized tomography and etc., but the process of diagnosing a number of CT-scan images manually becomes tiresome and also susceptible to error. Therefore, computer aided systems are used to assist the physicians as a second option to reduce the mistakes and errors, this raise the need of the automated com-puterized system, such as CNN [3]- [6]. In literature, there are many applications of CNN in the medical field and has played an important role for automated detection of cancerous cells from mammographic images, Computed topography (CT) and Magnetic resonance (MR) imaging are the most widely used radiographic techniques in diagnosis, clinical studies and treatment planning.…”