Limitations of artificial intelligence in tissue engineering and regenerative medicine

Designing and understanding the organizational principles of living tissues requires the integration of diverse knowledge and skills owing to the extreme complexity of biological systems. This complexity makes predicting tissue behavior and dynamic properties particularly challenging. Artificial intelligence (AI) and machine learning (ML) have emerged as powerful tools for addressing these challenges. They have demonstrated success in fields including natural language processing, computer vision, fraud detection, autonomous systems, and robotics. However, their application to tissue engineering and regenerative medicine (TERM) presents unique domain-specific challenges that remain unresolved. This narrative review identifies and discusses five critical limitations that currently hinder the effective integration of AI and ML into TERM, namely: (1) imprecise terminology in biology, (2) lack of formalized concepts and terminology in TERM, (3) absence of clear logical constructs in the field, (4) limited development of quantitative physiology, and (5) lack of quantitative requirements for applying engineering principles.