
==== Front
Neurooncol Adv
Neurooncol Adv
noa
Neuro-Oncology Advances
2632-2498
Oxford University Press US

10.1093/noajnl/vdae090.035
vdae090.035
Final Category: Data Sciences/AI Advances
AcademicSubjects/MED00300
AcademicSubjects/MED00310
DSAI-02 AN AUTOMATED TRAJECTORY OPTIMIZATION ALGORITHM FOR LASER INDUCED THERMAL THERAPY (LITT) FOR BRAIN TUMORS
Lu Yonggang Medical College of Wisconsin Department of Radiology, Milwaukee/WI, USA

Krucoff Max Medical College of Wisconsin Department of Neurosurgery, Milwaukee/WI, USA
Marquette University & Medical College of Wisconsin Department of Biomedical Engineering, Milwaukee/WI, USA

8 2024
02 8 2024
02 8 2024
6 Suppl 1 2024 SNO/ASCO CNS Metastases Conference i11i12
© The Author(s) 2024. Published by Oxford University Press, the Society for Neuro-Oncology and the European Association of Neuro-Oncology.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com

Abstract

PURPOSE

To plan laser probe trajectories for magnetic resonance imaging (MRI)-guided laser induced thermal therapy (MRIgLITT), current clinical practice relies on manual inspection on MRI images to find a path with minimum safety risk and maximal treatment impact. This process is empirical and time-consuming with potentially suboptimal solutions. The study aims to develop an automated trajectory optimization algorithm for reducing planning time and improving patient safety.

METHODS

Seven brain tumor cases treated with MRIgLITT were retrospectively reviewed. For each case, preoperative MRIs were auto-segmented for 16 brain tissue types (e.g., skull, artery, vein, etc.) using a deep learning-based head modeling algorithm (Sim4life v8.0, ZMT Zurich MedTech AG, Switzerland); Then an optimize trajectory was generated by searching potential trajectories based on a priori established criteria, such as trajectory feasibility (angle to the skull, not entering in the face or neck, etc.), safety risk (avoiding blood vessels and optical cranial nerves, etc.), and expected treatment efficacy (100% tumor coverage with assumed 1.5 cm laser radiation radius). The predicted optimal trajectories were finally compared with the trajectories prescribed from clinical procedures, and differences were analyzed descriptively.

RESULTS

For each case, the algorithm took approximately 4 minutes for brain tissue segmentation and 2 minutes for trajectory optimization on a desktop computer. The predicted trajectories from 7 cases were visually matched with prescribed trajectories with offsets of entry point (9.94±5.97 mm), target point (4.59±2.23 mm) and trajectory length (5.90±3.69 mm). Qualitative review of these trajectories confirmed their adherence to the above principles, meaning they would likely have been usable for their respective cases.

CONCLUSION

We demonstrated the feasibility and efficiency of an automated trajectory optimization algorithm for MRIgLITT. Future studies will improve the performance of brain tissue segmentation, assess anticipated versus achieved tissue treatment volumes, and prospectively evaluate application into case planning.
==== Body
pmc
