Physicians are under increasing pressure to assess complex patient data quickly and accurately. Moreover, image interpretation especially for the early diagnosis of cancer is error-prone and labour intensive. To tackle these challenges, Radboud University Medical Center (RUMC) is joining force with ScreenPoint Medical and Thirona to develop generic AI software that can be used autonomously to improve and accelerate the early diagnosis of breast and lung cancer through image interpretation without compromising transparency and control (the physician remains the decision-maker).
Thirona’s primary objective is to automate bronchoscope route planning by generating detailed anatomical data and calculating potential navigation paths for the physician. During the biopsy, these routes are displayed in real-time and projected onto intra-operative images, providing the physician with critical guidance to successfully reach the target nodule.
Firstly, Thirona will develop and improve the segmentation algorithms for preoperative and intraoperative image. Secondly, the registration algorithms will be developed and improved to create a spatial link between two images taken by different modalities and points of time. Finally, by combining these algorithms, the most optimal navigation route to the airway of peripheral lesions will be automatically generated. Following this process, a 3D view of the routes will be generated for simulating the bronchoscope’s position, ultimately assisting physicians in both biopsy preparation and live navigation
