No abstract

Innovation in latest technologies have provided means to gather data using various ways. Almost all the domains generate, store, and analyze the data for the improvement of the services they provide. The healthcare industry too generates a significant data which is used for improving public healthcare. While dealing with health data, it is necessary to follow appropriate data ethics as health data is considered the most sensitive and it needs to be properly collected, stored, processed and shared with different domains. This research paper discusses about the various data ethics to be followed to handle individual's health data, suggests a framework to deal with this data and a use case is suggested to understand the data ethics tenets.

No abstract

Early detection and prevention is the only way to treat lung cancer to avoid the loss of life. Where Computed Tomography (CT) screening is viewed as perhaps the best technique for discovering the early indications of lung malignant growth. The primary goal of this study is nodule detection and classification of collected CT scans images as benign or malignant. Sometimes some human errors can occur in the checking of a long series of CT slices of a single patient manually. This automated system (CAD-Computer Aided System) can help to radiologist or doctors to know the current stages and condition of the disease to diagnose correctly and quickly on a single click which will useful for radiologists and doctors to avoid the serious disease stage. The key four processes of our proposed system are input CT images, pre-processing, features extraction, and classification. In the proposed approach firstly we read all the CT image database (70 thoracic lung CT scans) having Dicom format then applied some pre-processing techniques of Matlab to enhance the image quality and obtained texture features. Using texture features, we extracted several features. At the end, we classified the dataset as benign or malignant using the K-means clustering method, and we achieved an accuracy of 92.8 percent.

No abstract