We have developed a novel service (https://ms.quantim.eu), which from routinely acquired MR images extracts reliable and clinically valuable data on the number, volumes and sites of WML, and brain volume loss (atrophy) measurements, and changes of these parameters through time. The service enables direct communication of anonymized MR images of patients with multiple sclerosis from a hospital to the server for image analysis, which is currently at the Faculty of Electrical Engineering. Jointly with neurologists and neuroradiologists from several slovenian hospitals we have also contributed to the standardization of MR image acquisition. The MR images are processed and analysed within max 3 hours after their receipt with the image analysis software that has been developed and validated within this project. Quantitative measurements about brain structures and their changes that are important for multiple sclerosis are included in the report (see attachment), that is communicated to the neurologist and/or neuroradiologist in the relevant hospital. Within this project our server has been connected with all slovenian hospitals (UKC Ljubljana, UKC Maribor, Jesenice, Celje, Novo Mesto, Izola, Murska Sobota, Slovenj Gradec, Šempeter ter UKC LJ pediatrija) that treat MS patients. Since now we have processed and analysed about 900 MR images of cca 600 MS patients. Responses form neurologists and neuroradiologists are very positive since all they consider the additional quantitative data very useful.
F.11 Development of a new service
The developed service https://ms.quantim.eu is currently free of charge for the slovenian hospitals. For the purpose of future marketing of this and other services and products at home and abroad we have founded the company INTELITEH d.o.o. (http://www.bizi.si/INTELITEH-D-O-O/) in early 2016. The company is mainly focused on R&D and upgrading of this service as well as with R&D of novel services in the field of image-guided diagnosis, therapy planning and simulation, image-guided interventions, and image-guided monitoring of disease progression and efficacy of therapies.
F.20 Company spin off
In Slovenia there are more than a dozen different MR scanners in use, which all have specific settings and protocols of MR image acquisition. To improve and harmonize the quality of automated quantification of MR images (https://ms.quantim.eu) across different scanners we have established, as part of this project, a direct communications channel with the radiology departments of two major clinical centres (Lj and Mb) and seven regional hospitals. Based on our feedback on MR image quality we have optimized the parameters of each scanner, such that the acquired image was optimal both for the automated analysis and visual reading. The newly setup MR protocols are isotropic 3D, compared to previous anisotropic 2D, and better depict the brain structures, which are most often affected by neurological disorders like multiple sclerosis. The benefits of the new MR imaging protocol were verified by the neuroradiologists and neurologists, and based on their positive feedback the Health Insurance Institute of Slovenia introduced a new MR acquisition code ID number 11007 named Multiple sclerosis with 3D.
F.22 Improvement to existing health/diagnostic methods/procedures
Up to now we have processed about 900 MR images of approximately 600 MS patients. The huge database of MR images acquired on Philips, Siemens, and General Electric MR scanners of different generations and types represents an extremely valuable image database that can and will be explored for further improvements of the image analysis algorithms and for longitudinal accession of MS patients.
F.15 Development of a new information system/databases
Two PhD dissertations have been completed on the subject of this research project. Alfiia galimzianova finished her PhD studies with the dissertation entitled Automated segmentation of magnetic resonance brain images into normal and pathological structures, while Žiga Lesjak completed his PhD with the dissertation entitled Automated detection of white matter lesion changes in magnetic resonance images. Dr. Galimzianova has published 4 peer reviewed journal papers while dr. lesjak published two. Currently, dr. Glimzianova z a post-doc at Stanford University (https://profiles.stanford.edu/alfiia-galimzianova), while dr. Lesjak is an entrepreneur.
D.09 Tutoring for postgraduate students
COBISS.SI-ID: 11167316