Disease object detection in brain
יהודה פרסי , דוד מזרחי
:שמות הסטודנטים המציגים
ד"ר כהן שראל
שיתופי פעולה במחקר
Our overarching goal is to enhance the work of a dedicated team of researchers by integrating the latest advancements in deep learning into their existing research. Specifically, we seek to enhance the team's current work on detection, aiming to enable them to locate various types of cancerous tumors in MRI brain scans with greater accuracy.
In this project, we surpass the prevailing benchmark results. We achieved this by generating more realistic brain images using advanced Generative Adversarial Network (GAN) architectures. These machine learning systems are renowned for their ability to produce high-quality, intricate synthetic images, an aspect that we believe will greatly enhance the visual data quality of the research team's work.
Moreover, employed advanced data augmentation techniques on the object detection data of the research team. This approach not only expands their dataset but also improves the model's understanding of various aspects of the data, thus improving its performance and robustness. By adopting these innovative strategies, we aim to significantly enrich the work of the research team, setting new standards in medical image analysis and diagnosis.