Artificial intelligence analysis of minimally invasive surgery data.

To synthesize the expanding literature at the intersection of minimally invasive surgery (MIS) and artificial intelligence (AI) and to delineate the developmental patterns that are shaping the future surgical landscape. A narrative review of current evidence examining the integration of AI into MIS was conducted, focusing on technological evolution, the use of high-dimensional surgical data streams, and the emergence of multimodal datasets combining surgical images, real-time kinematics, and live video feed. The literature demonstrates a clear shift from traditional machine-learning algorithms to advanced deep-learning architectures capable of processing big data without latency. Multimodal datasets are increasingly enabling the creation of smart surgical environments with high-fidelity context awareness. As a result, AI systems are evolving from passive observers to explainable digital assistants capable of identifying anatomical structures, predicting surgical phases, providing real-time guidance, and supporting surgical education through objective, data-driven assessments that reduce the learning curve for novice surgeons. However, current models remain largely confined to experimental “sandbox” settings due to substantial ethical, regulatory, and safety constraints. AI is becoming an integral component of modern MIS, with the potential to augment surgeon performance and enhance surgical training. Yet, meaningful clinical integration will require addressing the ethical, regulatory, and safety challenges that currently limit translation from experimental environments to real-world practice.