<p>Artificial Intelligence in Surgery: From Anatomy to Ethics</p>

Abstract

<p>Artificial intelligence (AI) is transforming surgical care across preoperative planning, intraoperative navigation, and postoperative decision-making. This narrative review synthesizes advances in surgical imaging, robotic assistance, oncologic surgery, immuno-oncology, workflow optimization, and global surgery, emphasizing validated algorithms, performance metrics, and clinical outcomes. In imaging, deep-learning segmentation models (e.g., U-Net, nnU-Net) routinely achieve >90% Dice similarity for organ and tumor delineation, enabling patient-specific anatomical modeling and precision planning. In robotics, reinforcement-learning controllers refine instrument trajectories and have been associated with lower conversion-to-open rates in colorectal procedures. Oncologic applications span AI-guided fluorescence imaging with submillimeter margin resolution and recurrence-risk prediction models with AUCs exceeding 0.9. Emerging immuno-oncology platforms integrate multi-omic immune profiling to personalize surgical timing and improve disease-free survival. Natural-language processing and large language models streamline documentation and enable real-time, context-aware workflow support. In low-resource settings, AI augments task-sharing, triage, and resource allocation. Despite rapid progress, gaps remain in prospective validation, generalizability, transparency, and bias mitigation. Integrating rigorous evaluation and ethical safeguards with technical innovation will position AI not as a replacement for surgeons but as a cognitive partner that augments expertise and advances individualized, data-driven surgical care.</p><p><br></p>

Ähnliche Arbeiten