A Hybrid Approach for Precise Prediction of Brain Shift in Image-Guided Neurosurgery Using Finite Element (FE) Methods and Artificial Intelligence (AI) (digibrAIn)

Programme

FTI-Strategy Lower Austria 2021-2027

Project coordination

ACMIT - Austrian Center for Medical Innovation and Technology, Dr. Gernot Kronreif

Project partners

• Danube Private University 
• Technical University Vienna
• Medical University Vienna

Researchers involved at DPU 

• Ass.-Prof. Dr. Sepideh Hatamikia
• Ansar Rahman, MSc

Abstract

During brain tumor surgery, image-guided neurosurgical systems—also known as neuro-navigation systems—facilitate tumor localization by aligning preoperative imaging data with the patient’s coordinate system. However, intraoperative deformations of the brain, referred to as “brain shift,” can lead to a displacement of the tumor position and thereby increase the risk of damage to healthy brain tissue. Existing methods for compensating brain shift typically require additional intraoperative imaging (such as MRI or ultrasound) or rely on the manual acquisition of landmarks—both of which can disrupt the surgical workflow. In the digibrAIn project, we propose a novel approach that utilizes intraoperative image data from the surgical microscope to enable an automatic, AI-based estimation of brain shift without interfering with established clinical procedures.

Our approach leverages advanced deep learning techniques for non-rigid 3D point cloud registration, combined with synthetic training data generated using the finite element method (FEM) to realistically model potential deformations of brain structures. In addition, the advantages of transfer learning are exploited to perform the computationally intensive FEM simulations and AI model training in advance, allowing efficient updates based on current imaging data on the day before surgery. The proposed approach will be validated within the project using a novel instrumented brain phantom. With digibrAIn, brain deformations can be accurately predicted in real time, enabling improved planning of optimal surgical trajectories and providing surgeons with reliable guidance throughout the procedure—without interrupting the workflow. Ultimately, our approach aims to enhance surgical quality, increase patient safety, and improve patients’ quality of life. Lebensqualität der Patient*innen positiv zu beeinflussen.

Read more about the project