On 16 August 2026, the journal Bioengineering published an editorial by Daniele Giansanti titled "When Bioengineering Met Artificial Intelligence: From Machine Learning to Intelligent Bioengineering Systems." It examines the gradual convergence of two fields that long developed in parallel but are now actively intersecting.
The author notes that artificial intelligence and bioengineering initially followed different paths: the former through algorithms and data, the latter through physical and biological materials. However, their interests are increasingly coinciding in tasks involving the modeling of complex biological systems, the design of biomaterials, and the creation of adaptive medical devices.
The article highlights the transition from traditional machine learning to more advanced forms—hybrid neural architectures, generative models, and multimodal approaches. These methods allow not only for the analysis of data but also for the generation of new solutions in real time, taking into account the dynamics of living systems.
Special attention is paid to the challenges: the need for large, high-quality biomedical datasets, issues of model interpretability in a clinical context, and the ethical aspects of applying AI in bioengineering. Giansanti points out that many current approaches are still limited to narrow tasks and require significant adaptation to real-world biological conditions.
Compared to earlier works on the application of AI in medicine, this publication stands out for its emphasis on the systems level: not individual tools, but the creation of "intelligent bioengineering systems" capable of self-learning and adaptation. This distinguishes it from narrowly specialized reviews on deep learning in imaging or diagnostics.
The development of such systems opens up new research directions—from designing organs-on-a-chip with embedded AI to personalized bioprosthetics that react to changes in a patient's body. However, widespread implementation will require independent validation and the standardization of evaluation protocols.
It remains unclear how quickly the gap between laboratory results and clinical practice can be bridged, and what new risks will emerge when scaling intelligent biosystems. The community will likely focus on creating open benchmarks and interdisciplinary collaborations.
The key conclusion of the article is that the true value of this convergence will manifest not in isolated breakthroughs, but in the ability to create closed-loop cycles for designing, testing, and optimizing biological solutions with the help of AI.
