Artificial Intelligence in Radiology (AIRad)

The hypothesis of this project is that the use of AI in conventional radiography ssytems can improve the acquisition workflow and provide significant improvements in image quality to allow a reduction of ionizing radiation dose and greater diagnostic accuracy, from image enhancement avoiding the use of the anti-diffusion grid and eliminating tissue overlap either in the line of dual energy or tomosynthesis systems. The overall objective of the project is the development of a new generation of radiological devices that apply artificial intelligence in the different stages of a radiological study, from the positioning of the patient to the interpretation of the image by the radiologist.

The most outstanding elements of innovation in this project derive from the incorporation of advanced AI tools in the field of conventional radiology, characterized until now by the that fact that the major manufacturers are improving the hardware of the equipment, but little has changed with respect to workflows and how to achieve better machine-operator cooperation. The combination of resources coming from a research group of international level with experience the processing of X-ray image (UC3M), one of the major national hospitals focused on research and innovation, and capable of establishing user requirements, testing and validating with agility the developments achieved (HGGM) and one of the few national companies in the sector (SEDECAL), leaders in conventional radiology systems, ensures the combination of suitable resources to achieve highly innovative developments that can reach an international commercial exploitation in a very short period of time. The final expected level of technological development is TRL5, reaching within the framework of the project the verification of a demonstrator by radiology specialists in a relevant environment, though not yet in a real clinical environment.