As an un-initiated medical professional about new technologies, retired since long and no more in practice, I may be forgiven if I dare ask a perfectly reasonable question, “What is this artificial intelligence any way”? Assured in my blissful ignorance with the knowledge that I will not have any use of AI, I kept myself away from it. But not for long. My son sent me a poem written by AI. Being a doctor himself, he told me that AI is coming in a big way and doctors eventually will have to get acquainted with it. As an example of its impact, he informed me that AI is more accurate in diagnosing complicated cases when even senior professors might find it difficult. Eventually I persuaded myself to make an effort to find out what it was all about.
Definition of Artificial Intelligence
Many versions of its definitions exist but mainly ‘AI refers to an intricate system, capable of performing complex tasks that historically only humans could do including decision making, reasoning, and problem solving.’ As is common with most definitions, this did not make much of a sense to me.
With the wide use of computers, people became more aware of ‘computer intelligence’ but this was dependent on human skill. Alan Turin, a British mathematician in the 1950s, referring to the computers, asked the question, “Can machines think”? That was the beginning of Artificial Intelligence. John MaCarthy later used the term Artificial Intelligence, instead of computer intelligence. Artificial Intelligence (AI) can be defined simply as “the science and engineering of making intelligent machines.” Since those early days, the phenomenon of AI has engulfed the entire world encompassing every aspect of human life.
How does it work
The 2024 Nobel prize has been awarded this year to two scientists, john Hopfield and Geoffrey Hinton, for their pioneering work spanning over four decades of ‘foundational discoveries and inventions that enable machines to learn’. There were many other scientists involved in this research. Taking cue from nature that how biological cells of the brain (neurons) work, they pursued the nature model. Our brain has neuronal cells, which work by linking together by axons and dendrites, minute thread like structures jutting out of each neuron. One neuron is connected to many others. The connections are made at the junctions called synapses. Thus, neurons are connected together to make networks, which give our brains the power to think and learn, and perform other numerous body functions.
The scientist working on the artificial intelligence created artificial neural networks by making digital neurons and connecting them to one another like the biological neurons. They made and created an artificial neural network called ANN. The digital neurons are capable of learning like the biological neurons. Each digital neuron by itself does not do much but when connected with other digital neurons in various configurations, patterns and strength it becomes possible for artificial intelligence to work. With the knowledge of the basics of how artificial intelligence is created one can broadly deduce its main functions:
- Learning.
- Reasoning and problem solving.
- Self-correction.
Using complex algorithms, many AI systems and platforms have been created that are capable of analyzing complex problems, self- learning, acquisition of information and memorization etc. Chatbot and chat GPT are platforms that are capable of interaction with the consumer through audio or text. AI is also capable of self-correction suitably by fine tuning of algorithms.
This discourse is of course is very basic and brief, and needs much more elaborate detail. One hopes that it will kindle an urge to learn and understand this mind- boggling subject.
AI for the physician
Coming to the individual physician and clinician working on his own or in a health facility, the main benefits of AI are assessment of clinical problems including disease diagnosis, evaluation of related problems and co-morbidities in chronic and complicated cases. AI helps in creating a treatment plan along with monitoring, progress and prognosis.
With more accurate diagnosis, lab. Investigations and imaging procedures can be minimized to the essential. The same applies to the rational use of medicine, and invasive and operative procedures. Not to forget the robotic surgery. Managing data, keeping records, generating reports and communication tasks can be integrated in day today work.
These technological advantages are impressive, but I was still skeptical and wondered about the place of compassion and empathy. Convincingly enunciated in a ted lecture by a Stanford professor of medicine, with examples of clinical settings, touching patients physically is important and should be an integral part of holistic medicine. We are not dealing with tools and instruments but with living beings, who along with their physical problems may also be burdened with emotional and other concerns. For example, breaking the bad news, to the parents of a child in my care with acute lymphatic leukemia proved daunting. I was lost for words and required special counseling skill. Similarly informing parents of retarded or permanently handicapped children is not easy either. Perhaps, to console a dying cancer patient requires special communication skill and courage. Here we are in the domain of compassion and empathy, which I consider beyond the skill of a robot!
