Künstliche Intelligenz in der Strahlentherapie

Radiotherapy is a cornerstone of cancer treatment, and approximately half of all cancer patients receive radiotherapy during their course of disease. Artificial intelligence (AI) can be employed at all stages of this process to enhance precision, efficiency, and personalization.

This study represents a critical assessment of the current use of AI in radiotherapy.

Based on a selective literature review, the current state of knowledge regarding AI in radiotherapy was summarized and critically discussed.

First approved software solutions enable automated contouring of organs at risk which is crucial for radiation treatment planning. The quality of AI-generated contours for these organs is generally high and typically requires minimal human adjustment. AI applications also produce contours of locoregional lymphatic drainage pathways as part of the therapeutic target volume which are comparable to those created by experts in many cases. Especially in adaptive radiotherapy, AI-based applications are essential for swiftly adjusting treatment plans to accommodate daily changes in patient anatomy. AI-based prognosis and prediction tools have also shown promising results, but in most cases, prospective validation in large multicenter cohorts is still pending before these can be used in clinical routine.

The integration of AI into radiotherapy has advanced significantly in recent years and is already being clinically applied in several domains. AI-based applications have the potential to significantly accelerate radiotherapy processes, thereby reducing personnel workload. Moreover, they can contribute to standardizing tumor and organ at risk contouring, potentially enhancing quality when used correctly.