(by Giancarlo De Leo) Artificial Intelligence (AI) or AI (Artificial Intelligence, in the Anglo-Saxon diction) is a branch of computer science that, through the programming and design of both software and hardware systems, allows machines to be equipped with tasks that normally they are intended solely for the human mind.
Artificial Intelligence was born with the spread of computers and its date of birth is conventionally 1956. In fact, this year, Artificial Intelligence was first talked about during a conference (Dartmouth Summer Research Project on Artificial Intelligence) which was held in Dartmouth in the United States and which saw the participation of some of the most important names in computer science (McCarthy, Minsky, Rochester, Shannon, Newell, Simon). During this historic conference, some programs capable of carrying out some logical reasoning, particularly related to mathematics, were presented.
Today AI is able to revolutionize entire industrial sectors, as well as the very interaction of citizens with each other and with businesses, administrations and civil society.
Its potential is such that experts agree that AI is an unprecedented opportunity to increase work productivity and enable extraordinary progress.
On the other hand, the development of emerging technologies is one of the points at the center of the action carried out by the Ministry of Economic Development: the final document was published last July with the proposals for the "Italian Strategy for Artificial Intelligence" carried out by the group of experts which accepted the observations received following the public consultation last year.
While recently the Council of Ministers approved the establishment of the Italian Institute of Artificial Intelligence (I3A), a technological research and development structure created with the dual purpose of attracting talents from the international market and becoming a point of reference for the development of '' Artificial Intelligence in Italy, in connection with the main technological trends (including 5G, Industry 4.0, Cybersecurity).
The center, based in Turin, will be able to count on an annual budget of around 80 million euros and will become a reference point with 600 employees, in collaboration with research centers and universities. The sectors involved will include manufacturing and robotics, IoT, mobility, agrifood and energy, Public Administration, culture and digital humanities, aerospace and healthcare.
Artificial Intelligence in Healthcare
Artificial Intelligence applications in Healthcare are becoming more common for the computerization of some diagnostic activities to support doctors' decisions. AI in the medical field is an area that can still be considered underdeveloped even for the regulations in robotic surgery that are not adequate and to be updated.
AI applied to the huge mass of data produced by healthcare companies allows enormous benefits including:
- plan predictive and personalized health prevention policies;
- anticipate requests for assistance;
- improve the accuracy of the detection of clinical signs and symptoms;
- automated use of the results of laboratory analyzes;
- formulate personalized treatment protocols or plans;
- analyze the findings of the side effects of the therapies;
- anticipate emergency situations and calculate readmission rates to health facilities;
- predict the spread of a disease and facilitate coordination of care teams.
Artificial Intelligence therefore allows us to enter a new era of extremely early diagnosis, through the research and identification of symptoms or predisposition to contract a specific disease.
It is also possible to estimate the costs of hospitalization for critical diseases and monitor the impact of possible prevention campaigns.
Medical specialties such as cardiology and radiology are benefiting from this.
It should be noted that the diagnostic imaging sector is undergoing a profound transformation and the advantages of Artificial Intelligence are and will be extremely significant.
Another field of application of Artificial Intelligence in Healthcare is that of robot-assisted surgery. There are already very significant uses that testify to how AI is able to improve techniques already used for several years in the field of robotic surgery, especially in microsurgery.
Another area in which Artificial Intelligence can make a very significant contribution is in improving healthcare by promoting preventive medicine and the identification of new drugs.
In Healthcare IBM Watson is certainly one of the reference software in the world of AI and is currently used to identify specialized treatments for cancer patients.
In addition, Covid-19 has forced everyone to search for answers, putting a strain on companies, government agencies, health organizations, schools and non-profit agencies.
IBM has trained Watson Assistant to answer common questions about Covid-19 by directly leveraging US CDC guidelines and the results of the latest medical therapies.
The algorithms are able to provide results in less time than the analyzes based on the evaluation made by humans and can be used anywhere in the world.
Patients can thus take advantage of simplified processes and personalized assistance, otherwise impossible and much faster and with more time to spend with the treating doctor for more delicate activities.
Currently there is no real synergy between doctors and AI and it should be noted that these professionals still have difficulty understanding how Artificial Intelligence can actually carry out an assessment in complete autonomy.
The algorithm should have the ability to create an analysis that identifies areas of uncertainty by subjecting them to the evaluation of human intervention when it comes to the patient's well-being, because diagnostic and therapeutic decisions will always be in the hands of healthcare professionals who will then have the last word that can be faster and more accurate with the help of AI assistance.
Risks and ethical aspects
The integration of Artificial Intelligence in the Health System can save money by dedicating human intervention to more specifically health-related activities.
The spread of AI-based healthcare technology is also largely due to the IoT (Internet of Things) and wearables that exploit AI to increase the quality of life of users and to predict their problems.
Software using AI can analyze data to alert users and their healthcare professionals to potential health problems and risks, easing their workload and preventing unnecessary hospital visits, reducing costs.
As with everything AI-related, these health technology innovations rely on human-supplied data, which means there is a risk that datasets contain unconscious bias. In the delicate healthcare market, in particular, it will be essential to establish new ethical rules to limit and prevent the potentially “dangerous” propensities of Artificial Intelligence.
Future perspectives
The use of AI in healthcare is growing due to the continuous demand for evaluative and predictive technology, wearable technology, digital healthcare and digital transformation, but the concrete privacy problems have slowed the adoption of mobile and digital tools for manage patient data.
Healthcare facilities and healthcare professionals can therefore become real targets for cybercrime, given that healthcare data is perhaps the most valuable available alongside financial data.
On the other hand, the entry into force of EU Regulation 679/16 (GDPR) has introduced important changes in the approach to data security also in the health sector. This new approach, based on the concepts of accountability and risk management, has pushed and is pushing solutions in the clinical field, towards the implementation of a business management system based on privacy by default and by design, and on the adoption of adequate security measures (training, preventive and predictive maintenance, cybersecurity, business continuity and disaster recovery).
The exponential growth of large data sets (Big Data) and the consequence that there will not be enough trained and certified medical personnel to manage this new flow of information should be taken into consideration.
It will be important that digital transformation, in order to ensure the availability of clinically reliable, safe and cost-effective information, can reduce the uncertainty in the decisions taken in order to create a high value for the improvement of the diagnostic-therapeutic pathways of patients.
Therefore, the support of Artificial Intelligence will be decisive to allow a rapid analysis of the exams together with the reduction of the error rate.
Healthcare facilities that do not equip themselves with these tools will be doomed to lose ground, with very long waiting lists and increasing chances of making incorrect diagnoses.
In conclusion, Artificial Intelligence in Healthcare therefore figures among the enabling factors capable of eliminating distances and promoting social and environmental sustainability for the post-Covid rebirth of the country.
Giancarlo De Leo, Consultant in Medical-Scientific Publishing and Digital Health, Member and Secretary of the Digital Health Observatory of the Italian Digital Revolution Association (AIDR)