Research Projects

This section introduces all current and prospective projects of the Interventional Neuroscience group.

COVID-19 - Diagnosis tool based on artificial intelligence and medical imaging

To assess potential damage to the lungs of COVID-19 patients, hospitals currently use medical imaging such as X-rays or CT scans. Analysis of these images with artificial intelligence solutions could allows for better and faster diagnosis as well as better understanding of COVID-19 effects. It could significantly improve disease treatment and distribution of the resources. Several studies already point in this direction, showing that detection of lesions with chest imaging is key to diagnose and understand the disease. This project, proposed by researchers from the Systems Control and Interventional Neuroscience groups, aims to implement artificial intelligence solutions for computer-aid-diagnosis in Luxembourg as soon as possible in order to aid in the clinal diagnosis and improve diagnosis times. Later on, based on the data collected during the current outbreak, the researchers will build tools to predict the course of the disease in future patients and to evaluate long term sequalae. This project requires a close collaboration between hospitals and researchers from the LCSB and the CHL. (contact: Beatriz Garcia Santa Cruz, Dr Andreas Husch and Prof. Frank Hertel)

Integrated neurosurgical perioperative imaging (INSITU® study)

This study evaluates the diagnostic accuracy of state-of-the-art laser technologies, such as Raman spectroscopy (RS) and optical coherence tomography (OCT), in the evident intraoperative tumour diagnosis. Tissue specimens taken during the surgery of tumours of the nervous system (brain, spinal cord, peripheral nerves) are examined with a novel robotised device (SOLAIS System®, Synaptive, Canada). Recorded data are classified using methods of artificial intelligence/machine learning and correlated with conventional tumour diagnostic methods (histopathology, computed tomography, etc.). 

The aim of this project is to support a clear intraoperative tumour diagnosis, which in the future will help to perform resection control during surgery, as well as to clarify perioperatively if a tumour has been sufficiently removed or if there are still remaining parts. These new techniques may be able to replace or supplement intraoperative MRI and will help the surgeon in direct analysis of the data under the surgical microscope. Further findings will reveal new information on the tumour composition at the protein level. This could improve personalised adjunctive postsurgical treatment and could potentially close the gap between epigenetics, histopathology, imaging and biochemistry. This innovative research project, within a collaborative partnership between the National Service for Neurosurgery at the CHL, the LIH and the LNS, is funded by the Luxembourgish Cancer Foundation.

Deep brain stimulation planning and postoperative analysis with deep learning methods

Deep Brain Stimulation (DBS) is a clinically proven surgical treatment of the symptoms of Parkinson’s disease and other neurological and psychiatric diseases. Patients receiving DBS have electrodes implanted into specific brain structures containing different functional areas. The trajectory for such electrode insertion has to be a straight line through the brain, without injuring brain structures or blood vessels. The post-operative reconstruction of the DBS electrodes is important for an efficient stimulation parameter tuning. A major limitation of existing DBS approaches is that they are manual or semi-automatic, and thus both time-consuming and subjective. High quality 3D images are needed to define the best path in each patient for inserting the electrodes and to decide on a precise electrode positioning. Sophisticated device programming to reach an optimal electrical stimulation for each patient after surgery is currently lacking as well. 

The PaCER ® algorithm developed in our team most accurately detects the electrode position, based on postoperative CT scans. Furthermore, we work on algorithms for automatic trajectory planning for DBS. The aim is to develop an automatic-image processing pipeline offering access to high quality 3D images and guidance for planning and execution of this procedure. To define optimal electrical stimulation for each patient after surgery, we build real-time feedback loops to deliver best electrical stimulation, thus to design personalised feedback systems optimised for each patient.

Sensor-based analysis of patients with movement disorder patients and feedback steering for DBS

In collaboration with the Systems Control group and the NCER-PD-Team at LCSB (see Flagship project page 20), the National Department of Neurosurgery at CHL and the University of Applied Sciences in Trier, Germany, we are elaborating sensor-based tremor analysis of patients with movement disorders, such as Parkinson’s disease, Essential Tremor and Dystonic Tremor. The computed results are used to develop an algorithm for feedback steering in Deep Brain Stimulation. This project receives funding of the FNR. 

Projects in Neuromodulation for Pain

Pathophysiology and clinical analysis of high-frequency and pulsed radio-frequency spinal cord stimulation Together with the teams of Prof. Fernand Anton and Dr Marian van der Meulen from the Department of Behavioural and Cognitive Sciences at UL and our partners from CHL, as well as an industrial partner, we examine the pathophysiology of high frequency spinal cord stimulation (SCS) in chronic pain patients (HF-10 SCS). Today, the exact pathophysiological mechanisms of the HF-10 therapy are not fully understood. Within our project, the HF-10 stimulation effects of patients are examined and pathophysiological conclusions will be drawn, to better understand the treatment modalities. 


Together with the Clinical and Epidemiological Investigation Center at LIH and an industrial research partner, our team at the CHL has been working out the clinical scientific base for pulsed radiofrequency stimulation of the spinal cord for the treatment of chronic pain syndromes within a multicentric retrospective (EPIPULSE-Retro®) and a prospective monocentric trial (EPIPULSE-Pro®).