We are seeking an ambitious PhD candidate to work on this three-year project, which should result in a PhD thesis. The goal of this PhD project is to quantify the added value of specific AI tools in radiology. You will implement and evaluate two commercial AI tools in clinical practice across 3 hospitals in the Netherlands. In the Radboudumc, together with Isala (Zwolle) and the Meander MC (Amersfoort), AI tools for the evaluation chest radiographs and CT brain will be implemented. The project focuses both on the evaluation of (cost)efficiency and its diagnostic accuracy.
Background In recent years, the number of radiological exams has grown significantly, but this growth has not kept pace with the number of available radiologists. This has led to an increasing workload and a rise in burnout complaints among radiologists. The use of artificial intelligence (AI) offers a promising solution to alleviate this pressure. AI can not only improve diagnostic accuracy but also enhance workflow efficiency and the comfort of radiologists. This aligns with the government's goal to implement AI in healthcare, ensuring that care remains affordable and staffing shortages are addressed.
AI has the potential to significantly enhance radiology by improving both diagnostic accuracy and workflow efficiency. By automating routine tasks, such as the detection of normal images, AI allows radiologists to focus their expertise on more complex and critical cases. This not only reduces the cognitive load on radiologists but also helps streamline the overall diagnostic process, potentially leading to faster and more accurate results. Furthermore, AI can assist in identifying subtle abnormalities that may be missed by the human eye, thereby improving diagnostic outcomes. Overall, the integration of AI into radiology holds great promise for increasing efficiency, reducing burnout, and improving patient care, all while helping to address the growing demand for radiological services.
The Radboudumc, and the supervising team, has a great experience in developing and evaluating AI tools in radiology. In previous projects of the diagnostic image analysis group (
DIAG) of the radiology department, other AI tools have been tested in a pre-clinical or clinical setting. Radiology
HealthAIRegister was developed to create an overview of existing CE marked AI products in radiology. In
Project AIR, multiple AI tools for the same tasks were compared. Now prospective clinical evaluation studies should determine the added value of AI in clinical practice, for both radiologists and patients.