Digipathconnect - a smarter Meuse-Rhine area
Cross-border collaboration for digital pathology and AI development
Pathology (the study of disease) is crucial in clinical workflows for diagnostics. Yet, there is a shortage of experts across the region (Bychkov A., The Pathologist, 2023). As healthcare becomes more digital, this project aims to bridge the gap between the two by creating a platform for digital pathology. This platform will help pathologists to work together and use artificial intelligence (AI) and new innovative imaging tools to diagnose diseases more effectively.
Digital Pathology involves turning specimen slides into digital images that can be shared and analysed. This method makes diagnostics faster and more efficient. With the help of AI, pathologists can get extra support, especially when dealing with complicated diseases like lung or colorectal cancer. However, developing AI tools needs large volumes of high-quality data and intensive collaboration between pathologists and data scientists. DigiPathConnect will make this collaboration easier, bringing together experts from across the region to enhance diagnostic possibilities.
The project will also integrate a promising imaging technology, namely mass spectrometry imaging (MSI), into the digital pathology platform.
MSI is one of the expertises of the Maastricht MultiModal Imaging institute (M4i), an internationally leading molecular imaging institute affiliated to the Maastricht University. The M4i disposes of state-of-the-art instrumentation and long-standing experience in applying this technology to the field of biomedical research. Only with this technology can metabolites be spatially observed in pathology specimens.
The M4i will complement the data obtained from microscopic images with spatial metabolic data through MSI experiments. The integration of these two data types in the digital pathology platform will allow linking metabolic processes to the specific cell types. These analyses will be performed for a selected cohort of patients with the aim of predicting response to therapy or prognosis.
By integrating spatially resolved metabolic data with AI-assisted workflows in a digital pathology platform, DigiPathConnect will help the hospitals for using these applications, speed up research, and roll out new applications.
The project is funded by the Interreg Meuse-Rhine (NL-BE-DE) programme, an EU-fund for cross-border cooperation.
More information on this project can be found at: https://www.interregmeuserhine.eu/en/projects/digipathc
Can a smart surgical knife help brain surgeons remove tumours?
[Click here for the original story in Dutch]
When removing a tumour, it is not always easy for a surgeon to be sure that all cancer cells have been removed. Removing too much healthy tissue is obviously not desirable, and in fact, sparing organ tissue is vital. Especially when it comes to the brain.
Scientists at the Maastricht MultiModal Molecular Institute (M4I) developed their ‘intelligent knife’, or iKnife, in recent years. In mutual collaboration between M4i and UZ Leuven, a very extensive study of the knife's clinical application has been carried out in recent years. This showed that the iKnife can recognise tissue of brain tumours, among others, within seconds while cutting, with more than 98% precision. The European Interreg Flanders-Netherlands grant programme has made over EUR 2 million available for the further development of this technique.
Molecular brain navigator tool
The iKnife is an electrosurgical tool that generates smoke when cutting tissue during surgery. This smoke contains molecules that act as a kind of fingerprint, revealing key information about the tissue. By analyzing this molecular fingerprint, surgeons can differentiate a tumour from the surrounding tissue—information that is invisible to the naked eye. However, the size of the original analysis device posed challenges in the operating room during earlier tests. To address this, the research team is now developing a more user-friendly handheld screener specifically designed for brain surgeons.
"Making our current iKnife, or molecular navigator, more compact is challenging because we aim to maintain both the sensitivity and accuracy of the existing system," explains M4I researcher Eva Cuypers. "This presents several technical hurdles. For example, we need to use a different type of detector to achieve a compact design. This also means developing a smaller component for the smoke supply. However, the underlying principle will remain the same, allowing us to compare molecular profiles from the new system with those from our previous model. Artificial Intelligence will help us determine if we can successfully translate our current model to this new system."
Brain Tumours
The ‘intelligent surgical knife’ uses mass spectrometry, a special analysis technique that can significantly improve surgical outcomes. Currently, around 60% of brain tumour patients experience a relapse within five years, as residual tumour cells often remain after surgery. How could the more precise tissue recognition offered by the iKnife benefit individual brain tumour patients?
"The iKnife will likely enable quicker identification of tumour cells during brain surgery," explains Maastricht neurosurgeon Olaf Schijns. "By analysing the molecules in the smoke generated by the iKnife, we can remove brain tumours faster and more thoroughly, without damaging ‘healthy’ tissue. Once we have sufficient evidence of its benefits for brain tumour patients, it’s even possible we could use this technique for other patient groups, such as those with severe epilepsy."
To ensure the successful development of the molecular navigator, a collaborative team of researchers, surgeons and companies has been assembled thanks to the support of Interreg Flanders-Netherlands, the Province of Limburg and Dutch national co-financing companies like Next Generation Sensors (for co-developing the compact navigator) and Aspect Analytics (for statistical analysis and a user-friendly interface). Neurosurgeons from Maastricht UMC+, Zuyderland in Sittard, UZ Leuven, Ziekenhuis Oost-Limburg (Genk), and AZ Groeninge (Kortrijk) will provide input and test the developed brain navigator as end-users throughout the development process.
High throughout spectral imaging
with pixelated detectors
This STW project aims to develop new direct imaging detector systems based on the Medipix/Timepix chip for applications in X-ray, ion and electron imaging. The focus areas in biomedical imaging are molecular histology with mass spectrometry and low-dose, energy dispersive X-ray imaging. We target the development of an ion imaging system that is suitable for the detection of macromolecular ions of either polarity, which can be post-accelerated up to 15kV. In this way, we create a bio-molecular imaging device for molecular histology for use with ion mobility and time-of-flight mass spectrometers. A complimentary imaging approach aims at energy dispersive X-ray imaging. A novel system will be developed with a new MPX chip based device that makes it possible to separate the incoming X-rays based on their energy and use eight thresholds to make seven windows in which one can count the incoming photons. Our overall goal is to improve quality and reduce the measuring time in ion and electron imaging and introduce energy information into X-ray imaging and thereby reduce the required dose.
Microscope mode mass spectrometry imaging using a pixelated detector
Bare chips of the Timepix type in combination with microchannel plates (MCPs) are implemented on a triple focusing time-of-flight (TRIFT) mass spectrometer as a detector for microscope mode mass spectrometry imaging of organic molecules in biological samples at high speed. (Figure 1) Applying a high potential on the Timepix/MCP detector, a wide mass range and an increased signal-to-noise ratio can be achieved compared with other microscope mode instruments. [1-6]
Figure 1 – Schematic overview of SIMS-trift ion microscope setup with implemented MCP/Timepix detector. Mass spectra are acquired without the need of mass selection. The spatial distribution of compounds on the sample surface are maintained during ionization and mass seperation and projected on the Timepix detector as such.
Intra-operative diagnostics for
oncological procedures
Molecular information based surgery - Toward intraoperative MS-based
diagnostics for removal of hepato-pancreato-biliary (HPB) and breast tumors.
Partners: AZM-General Surgery & AZM-Department of Otorhinolaryngology |
Head and Neck Oncology
With more than 3 million new cases and 1.7 million deaths each year, cancer is one of the major health concerns in the modern European society, with continuously increasing incidence. Among the most prevalent forms are hepato-pancreato-biliary (HPB) cancers affecting the liver, pancreas and bile ducts; colorectal cancer (CRC); and breast cancer - the leading type of cancer in women. In most cases, surgery is the only curative treatment possibility, on the condition that the tumor has been diagnosed early enough. The patient, who is facing a life-changing diagnosis, has to undergo an invasive procedure with relatively high risk of recurrence, implying further surgical procedures with a low rate of long-term survival.
Improving the efficiency of surgery is bound to fundamentally improve the prognosis and quality of life for patients with the challenging diagnosis of cancer. The key hypothesis here is that accurate determination of the resection margin between tumor and non-tumor/healthy tissues and thus an improved rate of R0 resections in tumor surgery would reduce the recurrence rates in patients by lowering the risk of residual tumor tissue that remains during the intraoperative process. This would minimize removal and/or damage of healthy tissues as an undesirable effect for the patient and thus would potentially improve patient outcomes and survival rates. In collaboration with the Departments of Surgery and Pathology at MUMC+ we aim to deliver a novel molecular-guided platform for improving ex vivo cancer diagnostics and real-time diagnostics approaches during surgery. The first step will rely in building comprehensive molecular and histopathological classifiers (i.e. e-biobank) for HPB, CRC and breast cancers; which will potentially reveal new candidates for targeted therapies. The second step will be the incorporation of molecular-guided surgery tools for intraoperative MS-based diagnostics.
Identification of cancerous tissue
This innovative approach will present a more precise and accurate identification of cancerous tissue during surgical resection procedures and thus potentially improving the diagnosis and prognosis of cancer patients.