http://hdl.handle.net/123456789/7923 2026-04-04T12:13:12Z http://hdl.handle.net/123456789/7963 Study of Early Detection of Lungs Cancer Using Support Vector Machine and Artificial Neural Network Faisal Imran; Syed Hassan Tanvir; Abubakar Yamin Lung and bronchial Cancer is one of the world’s leading cause of death. It is world’s second deadliest disease which is now getting very common among men and women. Due to rapid deformation of climate, excess use of tobacco and working in hazardous waste side i.e. nuclear waste, explosive demolition centers left traces of harmful gases in air, which later can cause lung cancer and other fatal diseases. Cure of this deadliest disease is only possible by regularly examining of individuals, who are working and are exposed to health hazardous environment. X-rays images and Computed Tomography (CT) scans are the sources to detect nodules in lungs, which are the primary source of lungs cancers. Detection, identification and classification of these nodules are always a challenging job for doctors and researchers in medical imaging field. Researchers have developed many methods for nodules detection. Some of these methods include machine learning and artificial neural network. In this paper we have discusses few of these methods both from machine learning and artificial neural network for the early detection of nodules from CT scan. 2016-01-01T00:00:00Z http://hdl.handle.net/123456789/7960 A Review Analysis on Ophthalmology Caused by Hypertensionusing Structural Features of Eye Joddat Fatima; M. Usman Akram; Adeel M. Syed; M. Usman Akbar The Hypertension is considered as one of the most common diseases that lead to an ophthalmological illness that is Hypertensive Retinopathy. It affects the arterial structure which damages eye vision; later at severity it causes complete vision loss. A diligent study has been carried out to get in depth and detailed knowledge about some of the previous clinical and machine learning methodologies. We have analyzed the functional and structural features along with clinical findings using fundus images. A review analysis deduced that Hypertensive Retinopathy has different severity estimations. The highest level is the swelling in Optic Disk and the initial level has arterial narrowing. A review also infers about the techniques to diagnose at early stage require precise data about its arterial features. Furthermore, by utilizing hybrid set of features in training and classification of arteries and veins in fundus images, the turnout results can yield to accurate and early exposure of grading level in Hypertensive Retinopathy. 2016-01-01T00:00:00Z http://hdl.handle.net/123456789/7962 Automated Techniques for Detection and Classification of Diabetic Macular Edema: A Review Adeel M. Syed; Muhammad Faizan; M. Usman Akbar; Joddat Fatima Diabetic Macular Edema is the main cause of vision loss in diabetic patients caused by the accumulation of fluid in the macular region of retina. Detecting it at an early stage is a herculean task and requires great expertise and consumption of time. Many automated techniques developed to do so were analyzed in this study. Moreover, a fine survey of preprocessing techniques, feature selection, feature extraction, Machine Learning (ML) techniques and the data sets used for its training and testing was conducted. Many automated techniques in that matter have been able to achieve high accuracy in detection as well as classification of DME. Optical Coherence Tomography (OCT) stands out to be more effective and better result yielding than others for the detection and classification of DME 2016-01-01T00:00:00Z http://hdl.handle.net/123456789/7961 Digital Fundus Image Stitching for Generation of Retinal Mosaicto Help Ophthalmologists M. Usman Akbar; M. Usman Akram; Adeel M. Syed; Joddat Fatima These days medical Imaging is a very important field and is helping doctors in various ways to diagnose diseases and treat them. Fundus image is used by doctors has a limited field of view and to overcome this problem a new technique is presented in this paper to stitch the images together to generate a mosaic image. Proposed methodology uses Weber local descriptor for feature extraction and then a new technique is presented for seamless stitching. A locally gathered database is used and results are compared with other state of the art techniques. Significance of proposed system can be seen from the results generated. 2016-01-01T00:00:00Z