International Journal of Image and Graphics

Skeleton-based action recognition has become popular with the recent developments in sensor technology and fast pose estimation algorithms. The existing research works have attempted to address the action recognition problem by considering either spatial or temporal dynamics of the actions. But, both the features (spatial and temporal) would contribute to solve the problem. In this paper, we address the action recognition problem using 3D skeleton data by introducing eight Joint Distance Maps, referred to as Spatio Temporal Joint Distance Maps (ST-JDMs), to capture spatio temporal variations from skeleton data for action recognition. Among these, four maps are defined in spatial domain and remaining four are in temporal domain. After construction of ST-JDMs from an action sequence, they are encoded into color images. This representation enables us to fine-tune the Convolutional Neural Network (CNN) for action classification. The empirical results on the two datasets, UTD MHAD and NTU RGB+D, show that ST-JDMs outperforms the other state-of-the-art skeleton-based approaches by achieving recognition accuracies 91.63% and 80.16%, respectively.

Skin diseases have become prevalent in the present times. It has been observed in a study that every year the percentage of global population suffering from skin diseases is 1.79%. These diseases have a potential to become extremely dangerous if they are not treated in the nascent stages. It is extremely important that skin diseases are detected and diagnosed at the starting stages so that serious risks to life are avoided. Often, exhaustive tests are required so as to arrive on a conclusion regarding skin condition, the patient may be affected with. Thus, an expert system is required that has the ability to identify diseases and propose the required diagnosis. Presently, only a few solutions are available for diagnosis of skin diseases using computerized system but this is an era which is under extensive research and can be developed further. As the existing system has certain loopholes, this system attempts to override the present problems by applying a different approach. As a result of comparison of results from numerous research papers, an expert system has been developed by choosing residual neural networks (ResNet) and this system can be used to aid skin specialists in identifying and diagnosing various major diseases of skin like (Eczema, Psoriasis & Lichen Planus, Benign Tumors, Fungal Infections and Viral Infections) in more effective and efficient manner. The causes for identified skin disease can be outlined through this system and treatment can be provided. We have used Python language for implementing the proposed expert system that uses a 50-layer ResNets for training a dataset that has been taken from DERMNET. We achieved an accuracy of 95% using ResNet for training of the model and prediction of results at an epoch value of 10.

In recent years, many improvements have been made in image processing techniques which aid pathologists to identify cancer cells. Nowadays, convolutional neural networks (CNNs), also known as deep learning algorithms have become popular for the applications of image processing and examination in histopathology image (tissue and cell images). This study aims to present the detection of histopathology images associated to detection of invasive ductal carcinoma (IDC) and non-IDC in breast. However, detection of IDC is a challenging task in histopathology image which needs deep examination as cancer comprises of minor entities with a diversity of forms which can be easily mixed up with different objects or facts contained in image. Hence, the proposed study suggests three types of CNN architectures which is called 8-layer CNNs, 9-layer CNNs and 19-layer CNNs, respectively, in the detecting IDC using histopathology images. The purpose of the study is to identify IDC from histopathology images by taking proper layer in deep layer CNNs with the maximum accuracy, highest sensitivity, precision and least classification error. The result shows better performance for deep layer-convolutional neural networks architecture by using 19-layer CNNs.

In this modern world of biomedical medicine, the classification of breast density has been considered a very important part of the process of breast diagnosis. Furthering the same research, this research aims to determine the patient’s breast density by mammogram image with the help of modern techniques such as computerized devices and machine learning algorithms, which will greatly help the radiologist. To carry out this process, this research paper introduces a Convolutional Neural Network (CNN) model of deep learning that will work as a basis for waveform conversion and fine-tune. This deep learning model will prove effective in automatically classifying a patient’s breast density. With the help of this method, the last two layers which are fully connected are removed and joined with two newly formed layers. This would have helped in addressing a pre-trained AlexNet model that further improved the classification process. In this model, the original or preprocessed images are used at level 1 of the input (which is in sharp contrast to the usual methods based on the CNN model), which also makes the model compatible with the use of redundant wavelet coefficients. Because in the field of radiologists it is very important to underline the difference between scattered density and heterogeneous density, so the main objective of this research is focused on this end. As the proposed method has an accuracy of 82.2%, it shows a better performance. This research paper further compares the effectiveness and performance of the proposed method to traditional fine-tuning CNN models, with satisfactory results. The comparative results of the proposed method suggest that the proposed method is in the field of radiologists representing a helpful tool. This method may be intended to act as a second eye for doctors in the medical field with the intention of classifying the categories of breast density in the patient during breast cancer screening.

In this paper, we discuss the classification of the brain tumor in Magnetic Resonance Imaging (MRI) images using the U-Net model, then evaluate parameters that indicate the performance of the model. We also discuss the extraction of the tumor region from brain image and description of the tumor regarding its position and size. Here, we consider the case of Gliomas, one of the types of brain tumors, which occur in common and can be fatal depending on their position and growth. U-Net is a model of Convolutional Neural Network (CNN) which has U-shaped architecture. MRI employs a non-invasive technique and can very well provide soft-tissue contrast and hence, for the detection and description of the brain tumor, this imaging method can be beneficial. Manual delineation of tumors from brain MRI is laborious, time-consuming and can vary from expert to expert. Our work forms a computer aided technique which is relatively faster and reproducible, and the accuracy is very much on par with ground truth. The results of the work can be used for treatment planning and further processing related to storage or transmission of images.