The experiments with this artificial intelligence software, developed by the University of Turin, are at an advanced stage at the Pinerolo Hospital and will be validated by the end of the summer. A technology that can therefore also be used for other pathologies
The algorithm is voracious, hungry for knowledge, eating thousands of bits of information, in the form of chest X-rays, each described with a diagnosis. It’s called Corsa and it’s an experiment that promises to recognize covid by analyzing patients’ x-rays. The idea arose during the pandemic, when the radiologists of the Molinette Hospital in Turin sent the X-rays to the Computer Science Department of the University of the capital of Piedmont, led by Prof. Marco Grangetto. At the time, computer scientists, forced by covid, developed this software capable of storing and scientifically analyzing data. But it is in 2022, at the Pinerolo Hospital, that the real experimentation begins, extended to the entire ASL To3.
How does it work
“Everything is based on deep learning – explains Agostino De Pascale, director of the Radiology Department of ASL To3 – the more information we feed into the algorithm, the more we refine its diagnostic ability. We need to add different X-rays, from classic , interstitial, of covid and specific cases with complications so artificial intelligence, learning from this experience will be able to be more and more accurate it’s like in a very short time we gave a resident the experience of a retired doctor, who in his life has seen thousands of patients We are currently at 78 percent of correct diagnoses, an already excellent rate, which will continue to increase as the experiment progresses.”
The advantages offered by the algorithm
Yes, because the clinical trial has reached its final stage and will be validated and used by the entire scientific community by the summer. “There are two other advantages that this algorithm offers us – continues De Pascale – the first is that by allowing it to analyze more records from our patients, it indicates in a very short time who are the ones to be treated with priority compared to the others identify the most severe cases The second is that this platform, which we use today to diagnose codiv, can actually be used for any other pathology, even in cancer screening, it is enough to insert images, ie x-rays, and instruct the software what to look for and how.However, we must always remember that artificial intelligence is and must remain at the service of the doctor, who remains the protagonist of diagnosis and treatment, and can never replace him.
Source: TG 24 Sky
I am Barbara Redford, a professional journalist and writer with extensive experience in news reporting. I have been writing for The News Dept since 2019, covering topics related to health and wellness. My passion is to keep people informed about the latest developments in healthcare and the medical industry. With my articles, I strive to create awareness on various diseases while also highlighting their remedies or treatments.
Aside from writing for The News Dept, I also conduct interviews with renowned doctors and medical practitioners who provide valuable insight into different illnesses or conditions. My articles are often highlighted by several leading health websites as well as magazines due to their quality of information and accuracy of facts.
