المستخلص: |
تهدف هذه الدراسة إلى بناء أنظمة تمييز سرطان الثدى والتي تستخدم من قبل مختصى علم الخلية للفصل بين الحالات الحميدة والخبيثة. ويتم ذلك بتحليل الصور الطبية المستخدمة على نطاق عالمى واسع. أن الطريقة المستخدمة في البحث تستند إلى عدد من العوامل الهندسية الإقليدية مثل المحيط والمساحة وعامل الملئ لصور الخلايا المستقطعة من الصور البصرية للثدى، ومن ثم التصنيف اعتمادا على الشبكات العصبية ذات الانتشار العكسي
The study is based on research into the improvement of a breast cancer screening system that can be used by cytologists to differentiate between benign and malignant types using images that are typical of those currently interpreted by cytologists world-wide. The approach is considered based on features vector which is composed of Euclidian geometric parameters such as the object perimeter, area and infill coefficient in segmented cells of optical images of breast. The aim of study to create a system for classification of breast cancer, which is used by professional cytology of separation between benign and malignant cases. Medical images were analyzing used a global scale and widespread. The method used in the study based on a number of factors such as Euclidean engineering parameters, diameter, space and filling factor for cells withheld images from the visual images of the breast, and then depending on the rating of backpropagation neural networks
|