by Giulia Mazza
Recognizing and classifying lung nodules using artificial intelligence (AI). It happens with the research project Ailanton, born from the collaboration between theImaging Center of the Campus Bio-Medico University Hospital Foundation e Sferanet (Olidata group), and financed by Lazio Region. An all-Italian project and the first of its kind, both for the integrated structured reporting system and the AI system used. With an accuracy in reporting tumor lesions of up to 98%.
The primary objective is to develop an artificial intelligence solution applied to the e-health sector, interconnected within the Radiology workflow so that the search, identification, and classification of lung nodules on chest Computed Tomography (CT) scans are performed automatically, and the results are presented to radiologists without any disruption to their normal operations. For each patient, in fact, the radiologist must examine thousands of images, searching for possible lesions. With Ailanton, however, the algorithm, after processing millions of images, is able to directly write patient reports, flagging the presence of possible tumors along with other information such as the size and volume of the lesion, and the likelihood of interference. At that point, the physician can decide whether to modify the data or validate it by sending it directly to the medical record. teacher. Bruno Beomonte Zobel, director of the Diagnostic Imaging Unit and coordinator of the Imaging Center of the Campus Bio-Medico University Hospital, where the system is already in use on an experimental basis, explains that this method "It allows doctors to save 20% of their time. Time that we can instead dedicate to patients, talking to them.".
This system also represents an important starting point for prevention. International scientific evidence shows that screening for the early diagnosis of lung cancer using chest CT scans can reduce mortality by 20 to 26%.
