At TUM Klinikum Rechts der Isar in Munich, the radiotherapy team faces a growing number of complex cranial cases. From patients with dozens of brain metastases to those requiring retreatment, manual contouring can be time-intensive and variable across operators. As case volumes rise, the team has been exploring new ways to maintain consistency and efficiency without compromising clinical accuracy.

The challenge: segmentation at scale

For radiosurgery teams, delineating multiple metastases or large primary tumors is one of the most demanding steps in planning. Each contour must be precise, consistent and reproducible. As the number of lesions increases, so does the time required for manual work and the risk of inter-operator variability. The team at TUM Klinikum Rechts der Isar needed a way to standardize segmentation while keeping pace with growing treatment demands.

The innovation: AI-powered segmentation integrated into routine workflows

TUM Klinikum Rechts der Isar became one of the first clinical centers to implement Brainlab Elements AI Tumor Segmentation in daily practice. Utilizing contrast-enhanced T1-weighted MRI scans, the software automatically identifies enhancing tumor tissue using a neural-network-based model trained on more than 4,000 scans and 11,000 expert annotations.

The workflow supports:

• Semi-automatic segmentation of multiple brain metastases for SRS
• Semi-automatic segmentation of other cranial tumors, including meningiomas, cranial nerve tumors, gliomas, glioneuronal and neuronal tumors
• Fast processing with a segmentation duration of under one minute per contrast-enhanced T1-weighted MR image set*

Within just one month, the team incorporated the AI tool into all cranial planning workflows, including Gamma Knife (Elekta AB, Sweden) and retreatment reviews.

The shift: efficiency that reinforces precision

Early observations at TUM Klinikum Rechts der Isar show strong performance for brain metastases larger than 3 mm and primary tumors over 10 mm. By automating the most time-consuming parts of segmentation, the team can focus more attention on clinical decision-making and the nuances of each case. As one clinician noted, the practical impact becomes clear when treating patients with numerous lesions. Elements AI Tumor Segmentation helps standardize the foundation of the plan, allowing experts to build on it with confidence.

In their words: “Using Elements AI Tumor Segmentation enhances our multiple brain metastases and retreatment review workflows. We are very happy with the results after implementing this innovative AI method in our clinic.” — Jan Peeken, MD, PhD, Senior Managing Radiation Oncologist at TUM University Hospital.

The impact: AI as a meaningful part of clinical routine

For a center treating complex cranial cases at high volume, automated segmentation is becoming an important tool that supports consistency and efficiency across the radiosurgery workflow. By integrating Elements AI Tumor Segmentation directly into daily practice, the team is shaping a future where technology strengthens—not replaces—clinical expertise.

Please join us in recognizing TUM Klinikum Rechts der Isar for helping pave the way for practical, clinically grounded AI adoption in cranial radiosurgery.

* Valid in a widely available platform like Brainlab Planning Station 9.0.1 or higher.