همایش ملی بیوانفورماتیک ایران

صفحه اصلی / 4th international edition and 13th Iranian Conference on Bioinformatics

HETLN: A Hybrid Ensemble Model for Precise Localization of Breast Cancer Tumors in Radiotherapy Treatment

نویسندگان :

Hassan Salarabadi¹ Dariush Salimi² Negin Farshchian³ Mehdi Zoberi⁴

1- دانشگاه زنجان 2- دانشگاه زنجان 3- دانشگاه علوم پزشکی کرمانشاه 4- دانشگاه علوم پزشکی کرمانشاه

کلمات کلیدی :

Breast cancer،radiotherapy،semantic segmentation،transfer learning،ensemble learning

چکیده :

Abstract Breast cancer is characterized by the uncontrolled proliferation of cells within the breast tissue, leading to the formation of tumors. Following surgical removal of tumors, radiotherapy techniques are commonly employed to prevent the growth of any remaining cancer cells in the surrounding area. However, accurately identifying the precise location of cancer cells for targeted X-ray radiation poses a significant challenge in radiotherapy treatment. In recent times, semantic segmentation has emerged as a suitable deep-learning method for object detection, particularly in the context of medical images. The objective of this study is to develop a hybrid ensemble model, referred to as HETLN (Hybrid Ensemble Transfer Learning Network), which combines the UNet architecture with transfer learning networks as a backbone. The purpose of this model is to identify the critical regions required for radiotherapy treatment accurately. The proposed model provides 98% accuracy and 91% Mean-IoU score respectively. The conducted qualitative and quantitative performance analysis experiments show that the proposed model performs better in specifying the areas for targeted X-ray radiation.