AR application for sentinel lymph node detection in breast cancer

Abstract

With the growing relevance of breast cancer in worldwide health reports, its treatment techniques are constantly assessed for improvement. Most of breast cancer patient face surgery at some point of the disease and one of the most typical interventions they may undergo is sentinel node lymph biopsy, a radio-guided surgery in which the patient’s sentinel lymph nodes are extracted to be analyzed in order to assess further treatment. In order to perform an accurate and minimally invasive resection of lymph nodes, it is necessary to have a clear localization of the element and for this reason, new visualization systems like the one proposed in this project are emerging. The aim of this project is to create an AR application to improve nuclear image data visualization in sentinel node biopsies, this way the burden on nuclear physicians will lessen. Accordingly, two different AR 3D models based on nuclear images were built to assess its performance for advanced image visualization in sentinel nodes biopsies. While the first approach was performed on a breast phantom, the second one was based on a real patient nuclear images, both provided by the Nuclear Medicine Department of Hospital Clinic. The final AR application was developed in Unity and, while the results obtained were limited, they were satisfactory and encouraging for further development in the emerging field of AR in healthcare.