2023-10-17 13:17:02
“The integration of artificial intelligence (AI) into clinical practice is still quite far away. For now there are indirect applications. For example, robotic-orthopedic engineering in haemophilia, or the management of some elements of genetic research in rare diseases “. This was said by Angelo Claudio Molinari, head of the Hemophilia Center at the Gaslini Hospital and professor of the Specialization School of Hematology of Genoa, on the sidelines of the first appointment of the new series of ‘Sobi Talk’ meetings, dedicated to the topic of AI and its possible applications in the field of health and rare diseases.
“In the specific clinical practice of rare diseases – explains Molinari – the integration of AI is in the development phase. However, there are already examples, mainly coming from the United States, in which AI has seen application both in the field of research for gene therapy and the search for specific drugs for certain metabolic diseases, for which special software has been developed that uses artificial intelligence to select the most suitable drugs for the pathologies. Currently – specifies the hematologist – one of the The most frequently recurring fears are those relating to the ‘derailment’ of AI, i.e. the fear that it may not guarantee safety standards, for example in the field of privacy”.
However, the role of man remains central. “Applied to diagnostics – states the expert – to the therapy of any disease and to the therapy of rare diseases in particular, artificial intelligence will always need human personnel, who should be trained in order to give the correct programming to the learning algorithms “. The latter is an approach that in some ways already exists in some contexts, such as that of anticoagulant therapies. “Perhaps not everyone knows – recalls Molinari – that in Italy for many years, I think for more than 30, patients who receive oral anticoagulant therapy have been managed by a computer, at least as regards the dose. Then it is the human task to control that this is actually appropriate. Therefore – he concludes – the evolution of the digitalisation of medicine can only improve this type of interaction”.
#integration #clinical #practice #long