my timesThe Korea Times
  1. Opinion

Medical AI needs bedside manners

Listen

Summary

Medical AI can support more precise diagnosis, personalized care and prevention, but its effectiveness depends on whether patients and doctors are comfortable using it. The authors argue that developers should design tools around existing examinations and clinical workflows, involve users early and provide doctors with AI training. They also call for transparent communication with patients and ongoing feedback after deployment. They cite screening participation and patient disclosure figures to show why comfort and trust matter.


Key Facts

  • AI applications in health care include estimating disease risk, personalizing monitoring and treatment, accelerating drug development, and helping hospitals manage emergencies and daily operations.
  • Health-care professionals rank compatibility with existing practices among the most important factors in medical AI development, after a clear regulatory framework.
  • A 2013 study found that 25–46% of women who skip breast-cancer screening cite compression pain as the reason.
  • Only 55% of professionals using AI inform their patients, while 80% of patients want clearer information about its benefits and risks.
  • The authors say AI could help companies expand preventive health programs for employees and help doctors assess complex conditions by correlating physiological metrics and test results.
By Aurélie Jean and Sébastien Abgrall
  • Published Nov 18, 2026 11:10 am KST
Aurélie Jean

Aurélie Jean

PARIS — Artificial intelligence (AI) is quickly finding its way into almost every industry, making models more precise and predictive. In medicine, AI tools are already helping patients take a more active role in managing their health care and giving doctors access to information that was previously difficult, if not impossible, to obtain.

But these tools will prove ineffective if doctors and patients are reluctant to use them or reject them altogether. Developers must therefore consider the comfort of everyone involved, starting with patients and practitioners.

Sébastien Abgrall

Sébastien Abgrall

AI is a set of methods for modeling and simulating phenomena, understanding their underlying mechanisms, answering questions, and solving problems. In health care, it can improve the precision of medical examinations and diagnoses, estimate the risk of developing certain diseases, personalize patient monitoring and treatment, accelerate drug development, and help hospitals manage emergencies and daily operations. These applications are helping medicine move beyond treatment to prevention.

It may seem counterintuitive to talk about comfort in medical AI, but even the most accurate tool has little value if it makes patients uneasy or doctors find it too cumbersome. For patients, comfort begins with the examination itself. Some AI solutions require an additional examination, which the patient may never have experienced previously. The procedure must respect the patient’s dignity and consider their physical and psychological well-being, as patients who feel at ease are more likely to understand and consent to it.

This is already well understood by medical practitioners. Mammography machines, for example, are now designed with surfaces at human body temperature to make examinations less unpleasant.

When possible, AI should work with examinations that are already part of routine care. If a new procedure is necessary, changes must be kept to a minimum. Every additional step takes time to explain and can make the examination longer and more daunting for patients.

AI tools must also fit into doctors’ workflow. Health-care professionals rank compatibility with existing practices as the one of the most important factors in medical AI development after a clear regulatory framework. Doctors should not have to switch between systems or spend too much time reading and interpreting new results. Otherwise, they will reject or misuse these tools. Over time, though, overreliance on AI could also erode medical skills.

One way to avoid these pitfalls is to involve patients and doctors from the early stages of product design, so that developers understand the needs and constraints of their most important stakeholders. Working together can improve models’ fairness and explainability, preventing algorithmic discrimination.

Before developing or buying an AI tool, health-care providers should be clear about the problem it needs to solve. Doctors should also help align these tools with medical practice, guiding models toward fairer answers and setting limits on how they operate in different scenarios. To do that, and to keep up with the technology as it evolves, doctors need AI training during medical school and throughout their careers.

Patients need to be able to make sense of the information AI provides, too. Access to real-time health data could facilitate shared decision-making with medical staff and help patients take a more active role in their own care.

Developers also need to follow up regularly with doctors and patients about their experience long after a tool has been introduced. Regulations already require certain forms of monitoring, but user feedback can reveal problems that performance metrics miss. A comfort-by-design approach would make this feedback part of how medical AI is developed and improved.

Patient comfort is a measure of public-health effectiveness. A 2013 study found that 25–46% of women who skip breast-cancer screening cite compression pain as the reason. Tracking patient satisfaction alongside diagnostic sensitivity would help institutions understand whether advances in medical technology can encourage more patients to get screened.

Health-care institutions should consider the patient’s experience from the first appointment to receiving their results and be transparent about how they use AI. Yet only 55% of professionals using AI inform their patients, even though 80% of patients want clearer information about its benefits and risks.

With hospital budgets under pressure, AI tools could make prevention more accessible beyond traditional health-care facilities. Many companies are already introducing preventive health programs for employees and could use AI to expand access to early detection at controlled costs.

Beyond early detection, AI could give doctors a more holistic picture of a patient’s health by correlating physiological metrics, cross-referencing test results, and identifying links between pathological mechanisms. This could help them better understand complex conditions that cannot be fully assessed through a single test.

The AI-driven health care of the future will depend on two requirements long considered separately: model performance and the comfort of the patients and doctors who use these tools. Only by meeting both can medical AI make prevention accessible to all.

Aurélie Jean is CEO and co-founder of Infra, an AI deep-tech startup in women’s preventive medicine. Sébastien Abgrall is director-general of the Pôle Santé Bergère in Paris. This article was distributed by Project Syndicate.

Explore More

  • Q.

  • Q.

  • Q.