Let’s start researching the use of thinking machines. The goal: to establish the evolution and severity of the disease
But for the first time all this will be done by an artificial intelligence, i.e. a super server that after receiving all the information on infected patients will be able to predict the evolution of the disease with all its therapeutic implications.
This, in short, the research thatMantua asst is about to start (only the EU’s OK for funding is missing, but if not, we will proceed anyway) together with four other Italian hospitals: Niguarda (leader), Sacco, Monza and Cremona. Principal investigator for the hospital of Mantua he is the primary physician of health physics, Giampietro Barai.
“The answer we are looking for – explains Dr. Barai – is essentially that starting from all the data on Covid that accumulate in the various hospital wards, will go on to tell us how we will have to treat the patient suffering from the coronavirus, how its pathology will evolve and so if we have to assist him at home or if hospitalization will be necessary for him. And this will be told by artificial intelligence, even if but first the health worker will have to enter all the necessary data “.
We are talking about what experts call data lake or big data, or an ocean of data made of images such as x-rays, CT scans, clinical data and more to define the bio-markers that will tell if the disease will develop seriously or less severely.
The procedure involves first loading all the data collected on the disease and then those for each individual patient. AI, artificial intelligence, will develop a model that will be able to predict the progress of the pathology on the individual patient.
“In two words – continues the head of the Poma Healthcare Physics – a sophisticated program will perform the differential diagnosis of the disease. Our research project will start with Covid-19 but the idea is that it can also be exploited for other diseases. The idea is in fact to transfer it to a second phase by applying the same logic for other diseases “.
A further development of the study is that which sees the 5 hospitals involved in a network to implement forms of professional exchange between health professionals. Therefore, not only the use of artificial intelligence but also the ability to compare experiences. In its overall articulation, all this will form the basis of a short-term tool in support of teleconsultation: hospital professionals will be able to evaluate, on the basis of radiological images and other data, the clinical picture of the patients and then, after comparison, complete the diagnosis. It will also be possible to establish the severity of the case and, therefore, to guarantee a better quality and homogeneity of the treatments, with evident improvement in the patient’s care path.
In the long term, the established data lake will be the basis for the possible development of algorithms and these systems will be a useful support for all the medical figures that revolve around the management of Covid-19 patients, providing a quantitative and objective evaluation of the lung areas affected by Covid-19 interstitial pneumonia and by correlating‘imaging radiological with clinical and laboratory data to predict the evolution of the disease. In particular: it will be possible to discriminate Covid-19 from other morphologically similar pathologies (for example pulmonary interstitiopathies, lepid growth adenocarcinoma, other infectious diseases with interstitial lung involvement) by improving the ability of early diagnosis.
In the study that is about to take off and for which a duration of about 6-7 months is expected, the following departments of the AST of Mantua are involved: Health Physics, Pulmonology, Infectious Diseases, Radiology Anesthesia – Resuscitation.