Iran (IMNA) - In an interview with IMNA, Mersa Gholamzadeh, a faculty member at Tehran University of Medical Sciences, explained that within the medical scientific community, the concept of "augmented intelligence" has increasingly replaced the term "artificial intelligence" in the context of healthcare.
She noted that the term augmented intelligence underscores the primary objective of these technologies: to enhance, rather than replace, the cognitive capabilities of physicians. AI can significantly improve the speed and accuracy with which clinicians analyze complex datasets, including radiological imaging and genomic sequencing results. Nevertheless, clinical judgment, ethical decision-making, and patient-centered care remain uniquely human responsibilities that cannot be delegated to algorithms.
Addressing concerns about AI-generated inaccuracies, Gholamzadeh explained that general-purpose large language models (LLMs) are primarily trained for natural language generation rather than rigorous medical reasoning. Consequently, these systems may produce statistically plausible but clinically incorrect responses because they generate outputs based on learned language patterns rather than an understanding of human physiology, disease mechanisms, or evidence-based clinical practice.
She further distinguished between general AI models and medical-grade AI systems approved for clinical use. According to Gholamzadeh, validated clinical AI tools—such as diagnostic algorithms deployed in healthcare settings—are trained using carefully curated, medically verified datasets and undergo extensive validation to ensure compliance with stringent evidence-based performance and safety standards, unlike consumer-oriented AI chatbots.
Gholamzadeh also emphasized that AI systems cannot assume legal, ethical, or professional responsibility for the consequences of an incorrect diagnosis or inappropriate therapeutic recommendation. The physician-patient relationship is fundamentally grounded in trust, accountability, professional ethics, and legal responsibility. Moreover, AI lacks essential human attributes such as conscience, empathy, and the ability to understand the psychological, social, and emotional contexts that are integral to effective patient care.
She concluded that AI should be regarded not as a substitute for physicians but as an advanced clinical support tool, comparable to transformative medical technologies such as magnetic resonance imaging (MRI) or X-ray imaging. Just as the discovery of X-rays revolutionized medical diagnosis without replacing physicians, AI is expected to expand clinicians' capabilities by facilitating intelligent management of increasingly complex medical information, minimizing human error, and supporting more precise and evidence-informed clinical decision-making.
Your Comment