Artificial Intelligence and Health: Chronicles of a World to Come
Predicting the evolution of diseases to better treat patients, supporting doctors in managing clinical documentation, contributing to the search for new drugs. At the base, an extraordinary ability to quickly and simultaneously grind a huge amount of data. The possible virtual applications of Artificial Intelligence (AI) in the field of health are many and promise, with algorithms, a revolution. To be governed with great caution due to the scope of application and the multiple implications, from health rights to data privacy through the doctor-patient relationship. A technology that we must deal with today: it comes with many promises but with just as many concerns, requires a multidisciplinary approach and also raises ethical questions.
In a landscape where everything moves very quickly, we have identified a series of significant cases and consulted some of the major experts in the sector, including doctors, researchers and psychologists.
Ask ChatGpt
There are those who ask ChatGpt questions about pathologies and those who attach their clinical reports asking for an interpretation, regardless of the reliability of the answers (the machine invents when it doesn’t know) and of feeding sensitive data to the ether.
AI interacts with those who ask questions, unlike search engines that return a list of links, creating the illusion of an empathetic relationship. “AI can provide support in some aspects of care, but it cannot replace human empathy and understanding. Integration into clinical practice presents potential but also challenges, in particular the possibility of providing incorrect information,” he writes. Charles Alfred Clericiprofessor of clinical psychology at the University of Milan and medical director of pediatrics at the Fondazione IRCCS Istituto Nazionale dei Tumori also in Milan. As a doctor and patient he has studied what role algorithms can have when one falls ill with a rare disease.
Support in diagnosis
One of the strengths of AI is the ability to read radiological images to support (without replacing) the radiologist. Some hospitals, such as the emergency radiology department of San Camillo-Forlanini in Rome, are experimenting with software for the diagnosis of osteoarticular lesions and lung diseases. The goal? To improve outcomes, times and use of resources.
At Humanitas in Rozzano (Milan), software is used to identify fractures or cerebral hemorrhages: rapid diagnosis in emergency rooms allows for priority to be given to the most serious cases. “AI is an important support but it cannot replace the radiologist,” he specifies Letterio Politihead of High Field Neuroradiology and Functional Diagnostics at Humanitas.
Personalized Medicine
Researchers at the Mario Negri Institute in Milan also study drug efficacy through advanced data analysis. By examining molecular structures, clinical and sociodemographic data from patients, and the results of studies in the medical literature, they identify complex correlations that are not immediately obvious.
AI, they explain, has the potential to evaluate the effectiveness of a drug on a single person. They are studying, he explains Alexander Augustus“digital twins” or digital twins, that is, virtual replicas of an individual, an organ, a process or a population. The virtual twin of a patient simulates the characteristics and biological behavior of the patient, allowing to virtually test the effectiveness of different treatments and to identify the most effective one with the least side effects.
Predicting Diseases
To recognize the possible evolution of a disease, it is necessary to integrate clinical, genomic, social and environmental data and understand the relationships of dependence and causality that connect them. In this context, AI can help structure this information, cross-reference it and transform it into predictive tools.
Identifying early signs of a disease, more sensitively than conventional investigations, allows the doctor to indicate a targeted treatment path early on.
Many projects underway. The neurologist Federica Augusta and his multidisciplinary staff at San Raffaele in Milan are working on a project that focuses on the early analysis of Alzheimer’s. Mark Of The Doorhead of Leukemia at the Humanitas Clinical Institute, to another project to identify the timing of stem cell transplantation in leukemia patients.
How the doctor/patient relationship is changing
With the arrival of AI in healthcare settings, the doctor/patient relationship is changing: new words are needed to address the patient when algorithms are used in the diagnostic and therapeutic process. “When the decision-making process sees the support of the machine, with all the advantages it can bring in identifying the diagnostic range and the type of treatment, this must be brought to the patient’s attention,” he says. Elena Vegniprofessor of Clinical Psychology at the University of Milan and director of Clinical Psychology at ASST Santi Paolo e Carlo. The relationship, the psychologist emphasizes, must continue to be transparent.