Your role

This position sits at the interface of machine learning, uncertainty quantification and computational biomechanics. You will work within the Legato group and in close collaboration with ARSPECTRA to design and deploy probabilistic surrogate models that run on ARSPECTRA’s devices and provide robust, real-time guidance to neurosurgeons.

Key Responsabilities:

• Design and implement Bayesian graph neural networks and convolutional neural networks as surrogates for high-fidelity biomechanical models
• Quantify and propagate uncertainty, and develop strategies for model selection and switching in real time
• Integrate surrogate models with physics-based solvers, e.g. SOFA, FEniCSx, SOniCS, and clinical or phantom data
• Deploy models on ARSPECTRA hardware, including optimisation for embedded and GPU platforms.
• Collaborate with ARSPECTRA engineers and surgeons to create a complete AR guidance...