Just as the ability to capture the “decisive moment” counts in fine art photography, until now, it has been the radiology operator's experience that has made the difference in visualising scar areas in the heart during cardiac MRI examinations.
However, an Italian team of scientists has developed an innovative software called THAITI, which uses artificial intelligence to make this diagnostic practice more accurate and efficient. The software has been developed and patented by an interdisciplinary team consisting of Daniela Besozzi and Daniele M. Papetti from the University of Milano-Bicocca, Marco Salvatore Nobile from Ca' Foscari University of Venice, and Camilla Torlasco from Auxologico Italian Institute (IRCCS) in Milan.
The global conference on cardiovascular magnetic resonance (CMR 2024, in London) confirmed its value. It awarded the Italian invention first prize in the Shark Tank Competition, judging THAITI as “innovative, clinically impactful, with a translational value and commercially viable.” The model is the subject of an Italian and international patent application and aims to find investment through the collaboration of the technology transfer offices of the University of Milan-Bicocca, Ca' Foscari University, and Auxologico.
The software is designed to calculate the best "inversion time", a parameter needed to obtain images that can identify the presence of scar tissue in the heart after administering contrast medium. The inversion time is unique for each patient and changes frequently during the examination based on the amount of contrast medium in the heart. Usually, the operator chooses and adjusts the inversion time based on experience, the appearance of the previous image and the patient's characteristics. THAITI, on the other hand, uses an artificial intelligence model to determine the optimal, personalised and dynamic inversion time based on the patient's physiological and anthropometric information and the technical information about the examination. This allows for acquiring high-quality images of the cardiac tissue throughout the entire MRI examination.
"Our invention optimises the quality of the images acquired, making it a valuable tool to support diagnosis," explains Camilla Torlasco, a cardiologist and Coordinator of the Cardiac Magnetic Resonance Service at Auxologico. " THAITI makes the workflow more fluid, reducing operator fatigue and enhancing the patient experience."
“THAITI is a tool made to apply a precision medicine approach,” says Daniele M. Papetti, research fellow at the University of Milano-Bicocca, “because the turnaround time is calculated by exploiting specific characteristics of each patient. Furthermore, THAITI ensures the possibility of consistently acquiring images of standardised quality, which facilitates the reproducibility of examinations and their results”.
The prototype of THAITI has reached an advanced stage of development. "The fundamental functionalities of THAITI have been fine-tuned," says Marco S. Nobile, Professor of Computer Science at Ca' Foscari University. The team is still working on the user interface and on the scalability of the system, which must be able to respond in real-time to a large number of requests coming from potentially anywhere in the world.
The software requires no investment in instrumentation; just a user licence. As Daniela Besozzi, Professor of Informatics at the University of Milan-Bicocca, points out," This makes it easy for low—or lower-middle-income countries to adopt the technology, where the spread of cardiac MRI is limited due to the difficulty of providing adequate training for radiology operators. It is also beneficial for low-volume or recently opened cardiac MRI centres.
Cardiac MRI is a crucial test for determining the presence and characteristics of possible scar tissue in patients with ischaemic heart disease and cardiomyopathies. It is an essential tool for diagnosing and managing patients due to its diagnostic and prognostic significance. It also helps accurately study congenital heart disease, diseases of the pericardium and aorta, and, to some extent, the cardiac valvular system. THAITI has been trained on various cardiovascular diseases, making it highly generalisable.
