TRABIT is training the following 15 PhD students (click on the pictures to get detailed information):
WP1: Multiple Sclerosis
WP3: Brain Tumors
Maria Ines Meyer
Francesco La Rosa
WP2: Fetal Brain Disorders
WP4: Stroke/Neurovascular Disease
Carmen Moreno Genis
Ezequiel de la Rosa
Maria Ines Meyer (ESR-1 - Icometrix - Belgium)
Bio: My name is Ines and I come from Sines, a small city in the Portuguese coast. I have a masters degree in Biomedical Engineering and Biophysics from the University of Lisbon. During my master's thesis I worked with EEG signal analysis, which I continued to do for some months at a Neurosystems Laboratory of the University of Chile. I have some experience as a researcher in the medical imaging field, which I gained at the Biomedical Imaging Lab at INESC TEC Porto, one of the most prominent research institutions in Portugal. There, most of my research was related to the development of computational systems that could perform automatic retinal image analysis, with a focus on deep neural networks.
Project: The progression of Multiple Sclerosis (MS) is related to brain atrophy and the development of brain lesions. In order to track the evolution of the disease patients are generally followed-up over the course of years, and brain volume is calculated from Magnetic Resonance Images (MRI), which can be performed in more than one scanner or center over time. This can represent a problem, because it is known that different scanners produce images with different characteristics which influence the volumetric measurements performed on them. Even scanner updates or changes in protocol can result in different volume measurements, which decrease the reliability of the atrophy evaluation. My project focuses on this problem and aims to investigate ways to improve this issue both as a post-processing harmonization step, and at the image level.
Bio: I am Francesco and I come from Bologna, a university city in northern Italy. I have a bachelor degree in Physics and a master in Applied Physics, both from the University of Bologna. My master thesis was done in collaboration with Fraunhofer Mevis in Bremen, Germany. There, I worked on a project for an automatic segmentation of bones in abdominal CT scans with deep learning methods.
Project: Multiple sclerosis (MS) is the most common demyelinating disease of the central nervous system, causing inflammatory lesions in both the white and gray matter. Magnetic resonance imaging (MRI) is the gold standard imaging tool for diagnosing MS and monitoring it over time. The demonstration of lesions dissemination in space and time is currently one of the diagnostic criteria of MS. In my project, I focus on the development of automatic methods for an accurate and robust segmentation of lesions (both cortical and white matter lesions) in MS patients based on advanced and clinical MRI imaging protocols. This will include integrating partial volume algorithms with novel deep learning approaches, investigating the added value of advanced MRI sequences, and validating the novel segmentation framework on data acquired at different imaging centers.
Bio: I'm Stefano Cerri. I'm from Inzago, a small town close to Milan, in Italy. I have a bachelor degree in Computer Science and a master degree in Artificial Intelligence and Robotics, both from Politecnico di Milano. My master thesis was done in collaboration with a AI startup called Horus Technology and it was focused in develop deep learning descriptors to recognize objects from a small device for blind people.
Project: Multiple Sclerosis (MS) is the most common inflammatory disorder of the central nervous system and it's characterized by the formation of lesions and a general loss of brain tissue (atrophy). Due to its ability to visualize lesion formation, Magnetic Resonance (MR) imaging has emerged as the central tool to diagnose and monitor MS disease. Automated analysis of MR images of MS patients is complicated by the lack of specificity and sensitivity of the signal in conventional MR sequences, the lack of a standardized clinical protocol to acquire MR images across centers and the need to track tiny changes in longitudinally acquired data. My project is based on develop novel computational models to both detect white matter lesions and dozens of neuroanatomical structures in clinical MR scans of MS patients, and to optimally combine those measurements into sensitive biomarkers of brain disease.
Bio: My name is Lucas and I come from Rouen, a medium-sized town in Normandy, France. I am graduated from CentraleSupélec, one of France’s leading engineering schools, where I specialised in Applied Mathematics and Computer Science (2014-2018). In parallel, I also received a Master degree from ENS Paris-Saclay in Applied Mathematics and Machine Learning (2017-2018) and a Master degree from Paris-Sud University in Core Mathematics (2015-2016). I have been working in research in medical imaging from the beginning of my studies at CentraleSupélec in 2014, where I pursued a specific elite Research Track under the supervision of Prof. Nikos Paragios. I had the opportunity to strengthen my research skills during an internship at University College London (UCL) in 2015-2016 under the supervision of Prof. Tom Vercauteren, and an internship in the french start-up TheraPanacea lead by Prof. Nikos Paragios in 2018.
Project: I am developing new computational tools for brain fetal Magnetic Resonance Images (MRI) analysis to improve treatment for fetus suffering from Spina Bifida. ◉ Advisor: Prof. Tom Vercauteren. ◉ Research Interests: MRI reconstruction and super resolution, robust and accurate biomarkers measure and longitudinal data analysis.
Bio: My name is Thomas Yu, and I come from Chicago, IL in the U.S.A. I have a B.S./B.A. in Mathematics/Physics from the University of Chicago. In addition, I have a joint Msc in Applied Mathematics from the University of Hamburg, the Autonomous University of Barcelona, and the University of L'Aquila through an Erasmus Mundus Joint master program for which I received a full scholarship. My master thesis subject was treatment planning in HDR brachytherapy with robotic needle insertion.
Project: My project is in fetal brain MRI (super-resolution/ motion compensation) and parameter estimation in biophysical models.
Bio: My name is Athena and I received my BSc and MSc degree in Biomedical Engineering from Amirkabir University of Technology (Tehran Polytechnic), which is known as the center of excellence in biomedical studies in Iran. After graduation, I had great opportunity to continue my studies as a research assistant there. During this experience, I especially focused on the application of the blind source separation and compressive sensing on the brain signal/image processing, and extending it to other areas like connectivity analysis and brain-computer interface.
Project: The aim of my project is to model the normal development of fetal functional brain connectome using in utero MRI data. We are particularly interested in finding timing patterns of functional cortical development, its relationship to structural evolution, and the effects of disease that influence brain maturation in utero.
Bio: My name is Ivan and I am originally from Saint-Petersburg, Russia. I was trained as a physicist at both Bachelor's (at Saint-Petersburg Polytechnical University) and Master's (within Erasmus program at different universities Europe-wide) levels. My work during academic research projects focused on theoretical aspects of quantum devices' operation.
Project: The PhD project is dedicated to model-based optimization of treatment margins in the radio-therapy of glioma patients. The objective of this work is to develop and validate a new radio-treatment approach that replaces the 2cm safety margin, that is currently irradiated as a fixed safety area around the visible tumor, with a patient-specific margin that is adapted to anatomy, tumor location and shape. This will include further development of the biophysical model of tumor growth as well as the inference schemes, integrating image information with the model.
Bio: I have a bachelor degree in electrical and computer engineering from the University of Iceland. I worked for one year in signal processing and machine learning research for sleep diagnostics at Nox Medical. I did my master degree in information technology and electrical engineering at ETH in Zurich. My master thesis was about separation of audio signals with deep learning models.
Project: The goal of my PhD project is to develop accurate and interpretable prediction models for brain tumor grading and tumor recurrence detection from joint PET/MRI scans, taking advantage of the increase in sensitivity and specificity of dynamic FET PET data compared to conventional MRI.
Bio: My name is Andrey. Originally I come from Minsk, the capital of Belarus, where I obtained Specialist and Master degrees in Mathematics and Computer Science from Faculty of Applied Mathematics and Computer Science of Belarusian State University (BSU). My graduation project was about cerebrovascular segmentation and analysis of 3D Magnetic Resonance Angiography images. And my master thesis was focused on breast X-ray image analysis using deep learning. In spring 2015 I was a DAAD intern at University Hospital of Jena where I worked on vascular OCT video analysis. I also have some teaching and industrial experience, which overall includes such spheres as discrete mathematics, software engineering, stereo image reconstruction, computer vision, machine learning and deep learning.
Project: Given that our brain contains an enormous amount nerve fibres, often grouped in bundles, which are essential for us to be in command of our senses and our body, it is of high importance that in case of brain tumour surgery they would not be harmed. Thus neurosurgeon should be well informed of where and how close to the tumour nerve fibres are located. This information can be gained from Diffusion Weighted Magnetic Resonance Image. Thus the main focus of my project is to develop robust and accurate algorithms for fibre tracking in brains of the patients with tumours based Diffusion Weighted Images. Another goal is to provide uncertainty and risk estimation in order to inform neurosurgeons about fibres being dangerously close to tumour or about the potential errors in the results of fibre tracking.
Daniel Krahulec (ESR-10 – Philips – The Netherlands)
Bio: My name is Daniel Krahulec and I come from the marvelous mountain-rimmed lands of the Czech Republic. Embarking on university studies, I first pursued my Bachelor's degree in the Electrical Engineering program, consisting of numerous courses in biomedical engineering at the Technical University of Ostrava, and medicine in the Faculty of Medicine at University of Ostrava. In 2014, I spent a five-month exchange period at Tampere University of Technology (Tampereen Teknillinen Yliopisto, Finland), where I peeked into the Master's program of Biomedical Engineering and filled my study plan with courses in radiology and radiotherapy. One year later, I gained admission to the freshly launched study program of Human Neuroscience and Neurotechnology at Aalto University (Aalto-yliopisto, Finland), where I graduated with honours in December 2017.
Project: In the TRABIT network, I have been working on a PhD project focusing on industrial design. As opposed to the usual scientific PhD with the aim of publishing novel research articles, my primary goal is to develop a prototype of a clinical software application that will be applicable in tumor neurosurgery planning, intrasurgical guidance, as well as postoperative neurooncological follow-up. Firstly, a thorough analysis of clinical needs is conducted, during which customers (neurosurgeons and neuroradiologists) are interviewed about clinical workflow in different hospitals. Next, these clinical needs must be transformed into functional blocks of the prototype application running in a Philips internal software development environment. Further steps include designing a user interface and software architecture, as well as testing, in-house verification, technical validation, and clinical evaluation of the final solution.
Bio: My name is Luca and I come from Pietrasanta, a town on the Tuscan coast of Italy. I obtained my bachelor in Biomedical Engineering at the University of Pisa. There, in collaboration with the Center for Computer Assisted Surgery (EndoCAS), I conducted a thesis related to an augmented reality application for the synthesis of long bones fractures. I continued then my education with a master degree in Medical Imaging and Applications (MAIA), an Erasmus Mundus Joint Master organized by three different universities in Europe. During my master thesis at the ViCOROB research group (University of Girona), I focused on the application of deep neural networks to localize, segment and characterize skin lesions on full body images.
Project: In brain surgery, the survival of glioma patients correlates with the extent of resection. However, the intra-operative delineation of pathological tissue and functional brain imaging has not yet been resolved or requires considerable technical and time expenditure during the operating procedures. Moreover, the neuro-navigation based on pre-operative image data is limited by an increasing brain shift during the resection. In this context, intra-operative ultrasound can update the anatomical information, but its use is clinically insufficient if not combined with a high-quality pre-operative imaging technique, such as MRI. The goal of my PhD thesis is to use segmentation and registration techniques to combine intra-operative ultrasound with preoperative MRI to assist brain surgery in glioma removal.
Bio: Hi, my name is Suprosanna. I was born and brought up in Kolkata. I did my Master's in Signal Processing from Indian Institute of Science (IISc), Bangalore, India's one of the most prominent research institute. During my Master's, I explored sparsity aware signal recovery and image reconstruction using deep learning, which I continued as research associate in Spectrum Lab, IISc, for the next couple of months. Prior to that, I studied Electrical Engineering during my Bachelor's from Jadavpur University. I have some experience in both image processing and machine learning which I gained during my post-graduate thesis.
Project: 4D Flow MRI is an emerging non-invasive imaging technique to measure blood flow dynamics. Local blood flow parameters obtained from 4D Flow MRI can be used to aid in the diagnosis and therapeutic management of neurovascular diseases. The objective of this PhD project is to develop a robust and efficient physics based inference scheme to quantify the local blood flow parameters from 4D flow MRI data.
Maria del Carmen Moreno Genis (ESR-13 - RegionH - Denmark)
Bio: My name is Maria del Carmen Moreno Genis, I am a Mexican girl coming from the beautiful state of Oaxaca. In México, I studied my Bachelor degree in Biomedical Engineering at "Universidad Politécnica de Chiapas". During my studies, I was part of the joint research study "Thermography for Breast Cancer Screening", between my university and the "Centro de Estudios y Prevención del Cáncer" (CEPREC). After graduating, I started working at CEPREC as a research assistant and Clinical Trial Coordinator at the same time. Later on, I formed part of an Erasmus Mundus Joint Master program in Medical Imaging and Applications (MAIA), taking place at 3 universities in France, Italy and Spain, respectively. I obtain my Master degree with my thesis "False positive reduction for lesion detection in breast mammography based on two-views lesion correspondence strategy", which I did at the VICOROB research group from the University of Girona.
Project: Stroke and Transient Ischemic Attack (TIA) are cerebrovascular diseases caused by obstructions of the blood flow, depriving nutrients like oxygen and glucose to brain microstructures. Clinical manifestations after stroke, such as motor and functional deficits, have been studying; however, symptoms like fatigue or reduced attention span, present on chronic stroke and TIA, have barely been explored. Therefore, my project aims to explore novel clinical MRI techniques, such as microstructure imaging based on diffusion MRI and quantitative MRI, to reveal new insights into microstructure features that my underline why chronic stroke and TIA patients experience deficits like fatigue. As result, it is expected to obtain a new clinical neuroimaging tool, which may guide the clinical doctor in identifying patients with a need of interventions for fatigue and/or cognitive dysfunction for a better stroke/TIA recovery.
Bio: My name is Amnah and I come to Germany from Vancouver, Canada and originally from Pakistan. I completed my Masters in Biomedical Sciences from National University of Sciences and Technology, Pakistan. My masters research was focused on the brain effective connectivity analysis and I investigated the modulatory effects of tongue movements onto the motor network by applying dynamic causal modeling. My thesis was co-supervised by Dr Adeel Razi from University College London, UK and Dr Zahra Fazal from Donder’s Institute, Netherlands. Moreover, I also gained research experience in fMRI and functional connectivity of the brain from the UBC Cognitive Neuroscience of Schizophrenia Lab in Vancouver, Canada.
Project: In the brain, the blood-brain barrier (BBB) regulates the vessel permeability which leads to the exchange of water, nutrients and oxygen between the blood and the surrounding tissue. Several diseases are linked to the damage of the BBB like neurodegenerative diseases, brain tumors, stroke, and others. My aim is to develop a robust, efficient and reliable method for the assessment of BBB using a non-invasive MRI technique called Arterial Spin Labeling.
Ezequiel de la Rosa (ESR-15 - Icometrix - Belgium)
Bio: My name is Ezequiel and I come from General Roca, a southern city of Argentina. I have a degree in Biomedical Engineering (National University of Cordoba) and a master’s degree in Medical Imaging & Applications (University of Burgundy, University of Girona, and University of Cassino & Southern Lazio). During my master’s thesis I worked on myocardial infarction quantification using Late Gadolinium Enhancement MRI, applying machine and deep learning techniques.
Project: The objectives of my PhD project are to accelerate and improve the detection of the extent of ischemic stroke on clinically acquired CT images. In particular, we aim to develop and validate techniques with high clinical transfer potential, being able to i) provide fast and robust segmentation results, ii) to deal with multi-center and multi-vendor data, and iii) to cope with old lesions.