So-called artificial intelligence (AI) is having a huge impact on public health overall due to its capacity for organization, communication and care in the daily practice of medicine.

In his blog Dissemination of Science, Manuel Alfonseca Moreno, doctor of Telecommunications Engineering, Computer Science graduate and Professor at the Autonomous University of Madrid recalls some interesting issues with respect to terminology that are worth noting. What is now called AI is what has always been called computing, a term that has been replaced due to the greater impact created by the word “intelligence”. The term artificial intelligence was coined in 1956 at a seminar on computers at Dartmouth College, a private university in New Hampshire in the USA, at which intelligent programs were discussed.

Since then, AI has been defined in relation to computer programs that process symbolic information through empirical or inquiry norms, not based on precise mathematical deductions, but on the accumulation of data and experience. Indeed, Alfonseca queries the appropriateness of the name adopted, since calling it AI raises a fundamental problem. If the goal is to achieve an AI that even surpasses natural intelligence, we shall have to start by understanding the nature of what we want to imitate and surpass. Do we really know what natural intelligence is? In other words, what the mind is?

It does not seem right to compare AI with human intelligence, or to think that our mind works like computer hardware. Simply put, thought, the mind, is not an epiphenomenon of the brain, nor is it equivalent to it. It is not made up of matter, nor do the chips or their connections function like our neural networks. Neurophysiological and metaphysical dualism, in accordance with the Christian tradition regarding the concept of person, considers that the body and soul, brain and mind, are distinct realities, although hypostatically united in each human being. Having said that, we normally talk about weak AI and strong AI.

So-called weak AI (also known as narrow AI) is the AI of computer systems, which is evolving. It is the AI that we use to effectively, specifically, and automatically solve problems that obey routines restricted to logical algorithms that man himself has provided to the machines by training them to answer questions or address issues based on experiences for which the programs are trained (deep learning). It is not intelligence comparable to human intelligence, given that machines do not think for themselves, but rather they react to what is asked of them by responding specifically and automatically to commands previously provided by their designers.

The many applications of weak AI include some that are extremely important in medicine for: sorting large volumes of data (creating databases); searching for patterns and supporting personalized diagnoses; recognizing images (radio-ultrasound-mammography, etc.); providing remote medical care (telemedicine); and assisting surgery (robot-assisted surgery), etc. As well as these more direct applications in medicine, there are others of special interest in medical research, such as:  data analysis and problem-solving; discovering new drugs; translating texts; word processing; recognizing sounds or the spoken word, etc.

All these applications represent great achievements and new resources, which have made it possible to facilitate human intellectual and manual work with even greater accuracy. In any case, machines or computers do not work by themselves, nor do they run independently, but rather depend on algorithms and previous experiences that their creators have provided. Therefore, in a field as sensitive as that of healthcare, the decisions ultimately have to be human; in applications in medicine, they have to be taken by the clinician.

As to strong AI, which some think would be equated to human natural intelligence, this is still dependent on algorithms and previous information accumulated in the computer memory. Machines do not think for themselves, like a human with all their capabilities and emotions. Their intelligence is not abstract, like human intelligence, but concrete; they are able to manage, recognize and coordinate data according to previously accumulated records and offer possible answers to problems that arise. Many computer scientists dispute that AI will ever be comparable to human natural intelligence, at most conceding it some differences, such as its large capacity to store and associate cumulative data more effectively.

However, followers of transhumanist and posthumanist movements believe that there will come a time when what they call a “point of singularity” will be reached, a point of equivalence between AI and natural intelligence. For those who support these notions, the battle is in full swing, and while human intelligence remains in its natural state, with no advances other than those of the accumulation of knowledge, AI progresses exponentially.

Realistic computer scientists, however, do not believe that independence of thought of AI will be achieved. For example, computer engineer Jeff Hawkins, one of the pioneers of mobile telephony, says that: “scientists in the field of artificial intelligence have claimed that computers will be intelligent when they are powerful enough. I don’t think so. Brains and computers do fundamentally different things”.

Similarly, Dr. Ramón López Mántaras, director of the CSIC’s Artificial Intelligence Research Institute, says that: “The great challenge of artificial intelligence is to provide machines with common sense… No matter how sophisticated some artificial intelligences may be in the future, 100,000 or 200,000 years from now, they will be different from human intelligence”.

Telehealth/Telemedicine

Shortly before its last renewal in June 2022, the Spanish Bioethics Committee (CBE) issued a report on the subject of Bioethical aspects of telemedicine in the context of the clinical relationship [1].

Although the current golden age of health sciences has made possible specific, effective and radical treatments with the growth of research and clinical trials, which have enabled the development of new technologies (chemotherapy, imaging techniques, genomics, gene editing, etc.), the traditional body of the medical profession remains the doctor-patient relationship, in which principles such as compassion, listening, care, encouragement, respect for decisions made, accompaniment in the disease process, and emotional support must take precedence.

Nonetheless, in order to meet the increasingly complex needs of health care, anything the world of so-called information and communication technologies (ICTs) has to offer is of great help. The World Economic Forum talks about the Fourth Industrial Revolution as a fusion of the physical, biological and digital worlds, which is rapidly changing society at a global level and affecting all systems, including healthcare. ICTs have become useful tools in the healthcare setting, focused on the best patient care, including the possibility of transferring part of the patient’s healthcare to their home. AI is vital to move towards not only more efficient medicine, but specifically towards more personalized, participatory, preventive and precision medicine. According to the CBE reportAI plays a major role in the development of so-called personalized medicine, with solutions tailored to each patient’s health profile.

The UNESCO International Bioethics Committee also issued a report on Big Data and health in September 2017, in which it mentioned three fundamental ethical challenges to be addressed: autonomy, privacy and justice, the latter in terms of accessibility and solidarity, and stressed the importance of establishing effective guarantees so that both the dignity and freedom of patients, especially the most vulnerable, are protected.

However, if there is one matter that is becoming increasingly important in the use of ICTs, it is that of telemedicine, which consists of the provision of healthcare services in which distance is a critical factor. The use of telemedicine firstly facilitates the doctor-patient relationship (telecare or teleconsultation), and debuted recently with the Covid-19 pandemic. In its 2018 Declaration, the World Medical Association stated that: “Face-to-face consultation between physician and patient remains the gold standard of clinical care”. Today, teleconsultation is accepted as a replacement for face-to-face consultation in certain circumstances, but both types of consultations must be governed by the same principles of medical ethics, namely to: preserve autonomy; respect the dignity of the patient by seeking their well-being and not causing harm; guarantee the security of data, procedures and the right to privacy; and facilitate access to all healthcare services (principle of justice).

Telemedicine also enables communication between physicians, or with other health professionals such as nurses, physical therapists, and pharmacists. Its functions include facilitating the exchange of data to make diagnoses, recommend treatments and prevent diseases, and to mobilize resources. It is also an excellent resource to expand continuing professional development, research and evaluation tasks, etc.

In relation to patients, however, what remains essential is the need to maintain trust in the doctor-patient relationship. Dr. Pedro Laín Entralgo (1908-2001) defined the clinical relationship as a particular and unique type of relationship between people, at the heart of which is trust, which he established in three aspects: trust in the technique to heal; in the professional knowledge to apply it; and in the values of the person of the doctor [2]. Consequently, we must fight so that the dehumanization that is permeating many sectors of society — and in which AI is somehow involved — does not harm the doctor-patient relationship. Trust is intrinsically linked to a close human relationship. Dr. Warner Slack (1933-2018), who conducted pioneering research into computer health care records, said that: “any doctor that can be replaced by a computer deserves to be replaced by a computer.”

In accordance with this, the potential dehumanization associated with telemedicine has become one of its main challenges to overcome and its potential enemy. Consequently, we need to move towards focusing telecare on the patient, conserving the human aspects and the patient’s specific needs. We must avoid what is known as techno-solutionism, one of the pitfalls of a super-technified world, which offers automatic flawless solutions [3].

Telemedicine cannot become an element of convenience that jeopardizes patient safety, but rather an ally of the clinician that aids him in his work to address safety, risks and possible adverse events.

The CBE therefore proposes the following recommendations:

  • Telemedicine should be governed, at the very least, by the same bioethical principles as traditional medicine. Remember the sacred dignity of each person.
  • Telemedicine and teleconsultation should be considered as complementary and never an absolute substitute for face-to-face consultation.
  • Use telemedicine when it is an opportunity to improve the individual’s medical and health care.
  • Have an evaluation plan that allows adjustments to be made based on the results and outcomes of the different modalities.
  • Conduct quality studies on the impact of telemedicine on different diseases and in different populations.
  • Promote professional training in the use of telemedicine.
  • Do not use non-face-to-face consultations in order to make the workday profitable or reduce professional recruitment.
  • Improve the education of the public and make it accessible to the most vulnerable and underserved populations.
  • Provide telemedicine with regulatory and legal support that guarantees its proper use, security, confidentiality and data protection.
  • Promote a rigorous social, ethical and legal analysis of the impact of telemedicine on social and health care.

A fundamental issue in the use of AI in medicine is the protection of confidentiality, an obligation of health ethics. Incorporating personal data on patients’ health into computer systems increases the risk of losing privacy and confidentiality. All technology and data storage used in telemedicine must meet security and certification criteria established by health authorities, which prevent security breaches and unauthorized access to information. Depending on the nature of the information recorded in the computer systems, it may be necessary to use data traceability systems and, where appropriate, the data should be duly anonymized for authorized access to professionals only, for use in institutions or research projects. In any event, all of this requires procedures to be established for confirming the identity of users, legal representatives and professionals with access to medical data, treatment results, medication, etc. but never to patients’ identifying information.

Nicolás Jouve

Member of the Bioethics Observatory

Professor Emeritus of Genetics

Former member of the Spanish Bioethics Committee

 

References:

[1] https://comitedebioetica.isciii.es/wp-content/uploads/2023/10/CBE_Informe-sobre-aspectos-bioeticos-de-la-telemedicina-en-el-contexto-de-la-relacion-clinica.pdf

[2] Laín Entralgo P. La relación médico-enfermo. Madrid: Revista de Occidente; 1964

[3] Evgeny Morozov, La locura del solucionismo tecnológico, Katz, Madrid, 2017

Share This Story, Choose Your Platform!

Subscribe

Subscribe to our newsletter:

We don’t spam! Read our privacy policy for more info.