Artificial intelligence and medicine: applications can be varied, particularly in the field of diagnostics. A market destined to grow further over the next few years. Artificial intelligence and medicine are the two worlds that have decidedly distant origins but which have already started a close relationship of collaboration, with the fundamental objective of improving the health and life expectancy of the world population.
A relationship that is destined to be further strengthened over the next few decades, also because the challenges facing in medicine are incredibly demanding. The scientific evidence is available in an ever-increasing amount of scientific studies often difficult to interpret, find the limited application. And, last but not least, a growing expectation on the part of the sick and their loved ones, in the idea that diseases are unpleasant obstacles that stand between us and our right to health and the doctors are obliged to remove efficiently.
Artificial intelligence and medicine: what this binomial means
Artificial intelligence can play a fundamental role in responding to these challenges, thanks to the ever-increasing availability of data and information in the medical field. Just think of the wearable that can trace various parameters of our state of health (pressures, frequency cardiac, etc.). In general, the evolution of technology has now made it possible to store a series of data concerning the patient's medical history in digital format: X-rays, laboratory results, reports on patient pathologies, family history, and genome.
The point, as always happens, is to be able to transform these data into something useful, in this case from the health point of view and here - as well as in other sectors - AI applications - Artificial Intelligence comes into play. So much so that, according to a study, the market of artificial intelligence for health care should reach 6.6 billion dollars by 2021, with a growth rate of 40%.
Artificial intelligence in diagnostics
But what are the concrete applications of artificial intelligence in the world of medicine? In general, it can be said that the application of AI in the health field offers several advantages over traditional clinical analyzes and subsequent medical decisions.
AI learning algorithms, in an increasingly precise and refined way, can allow obtaining ultra-accurate information on diagnostics, treatment processes, treatment variability, and patient results. The Cognitive Computing systems typical of the AI, in particular, constitute a useful tool for the realization of the Clinical Decision Support System (CDSS, support systems for clinical decision) of new generation, which can play a fundamental role in the processes of prevention, diagnosis and treatment, as well as in the reduction of medical errors.
For example, artificial intelligence is already today able to identify diseases (such as tumors) in radiological images with great precision, due to the possibility of having instant access to a historical heritage of data that is not within reach of a single doctor. This, even for reasons linked to fatigue or non-total experience (especially in developing countries), can make mistakes in this delicate phase, with dramatic repercussions on patients' state of health.
An example of this type is Watson for Oncology, developed by IBM, a platform that has an advanced ability to analyze the context and meaning of unstructured and structured data contained within clinical notes and reports. Subsequently, due to the combination of the characteristics from the patient's file with the clinical expertise, but also with external research and other data, the program can identify the potential treatment plan.
Artificial intelligence in cardiology
Much discussed in recent weeks, was a study published in the journal Nature that involved Google and, in particular, its technicians of the Medical brain team. Using AI techniques, Google was able to predict the probability of a patient's death with 95 percent accuracy, due to the ability to automatically access PDFs from previous medical analyzes and the notes of doctors and nurses.
The applications of AI in cardiology are increasingly promising, research by the Medical Research Council in London published in the scientific journal Radiology. It highlighted how a latest-generation software equipped with AI, starting from the data provided by cardiac magnetic resonance, can predict which patients are at greater risk of experiencing a cardiac event, first of all, the heart attack.
Artificial intelligence and medicine: when the binomial means "patient health."
The ability to perform accurate and anticipated analyzes can allow AI to play an important role even in a particularly "hot" issue in the health world, such as the growing resistance to antibiotics in the populations. The machine learning and artificial intelligence can help identify patterns of infection and to identify patients at risk even before they start to show symptoms of the disease, thereby reducing the use of antibiotic therapy.
Ai can also be beneficial for improving patient care: a San Francisco startup, Sense.ly, has developed a virtual nurse. Basically, it is an avatar that uses machine learning to support chronic patients between visits to the doctor. AiCure developed a similar solution, an app supported by the National Institutes of Health, which uses the smartphone's webcam and AI to confirm that patients are following the doctor's instructions independently.
Returning instead to hospitals, the use of smart devices plays a crucial role in monitoring the condition of patients in intensive care. The integration of artificial intelligence in these devices can improve the ability to identify possible deteriorations or improvements, thus allowing health professionals to intervene more quickly and effectively.
The applications of artificial intelligence in medicine are many and destined to change in depth the world of medicine, perhaps even too much. The risk that may arise in the long term is that of excessively dependent medical workers who are attached to the recommendations of artificial intelligence systems.
This technology has been compared by many to a stethoscope. Just as this last tool, not everyone has immediately grasped its importance and usefulness. In the beginning, Artificial Intelligence in medicine was perceived as a threat. This negative attitude has now, fortunately, largely been overcome.
Precision and speed: the two characteristics of Artificial Intelligence
If one wanted to compare the AI to a doctor, two are the main differences that emerge. Artificial Intelligence in the health sector guarantees greater precision in the development of a diagnosis and even more speed in identifying pathology.
At Oxford, for example, researchers developed an AI-based diagnostic system. Thanks to technology, it is possible to diagnose heart disease with greater accuracy in 80% of cases.
The computing power and the deep learning, moreover, are not absolutely comparable to the capacity of any human being. This feature determines the speed in developing a diagnosis that helps the doctors make the best decision. In the case of a patient with internal injuries, for example, the AI allows the doctor to know immediately what the problem is. It is then up to his knowledge and skills to decide how to act to save a person's life.
Predictive algorithms and Big Data in medicine
The AI can predict events even before they happen and even predict with what percentage in terms of probability. In the case of a pandemic, for example, he can understand the next development. All this is possible due to machine learning and the creation of predictive algorithms.
Artificial Intelligence in the health sector is a real valuable ally for all those involved. The AI stands as a solution and support in a world in which all health systems are in crisis both for lack of doctors and for a lengthening of life expectancy that is leading to the affirmation and spread of numerous chronic diseases due to the advancing years. The impact of Artificial Intelligence in the health sector is already visible to everyone. It is moving towards an increasingly personalized medicine in which the patient is at the center.