Postdoc in AI-based image synthesis for BoneMRI reconstruction

You have a PhD in image sciences, software engineering, applied physics, bio-medical engineering or a comparable training. Deep learning and MRI experience is a must. You learn quickly and you are a real team player in a multidisciplinary environment.

PROJECT

A 3-year Postdoc position supported by the Health Holland TKI funded project entitled “3D SPINE: Generic BoneMRI to guide growth and regeneration of the juvenile scoliotic spine” (3DSpine, HH-TKI, coordinated by Dr. Peter Seevinck, co-supervised by orthopedic surgeon dr T. Schlosser and biomechanics expert dr. H. Weinans). The project is collaboration between the Image Sciences Institute of the imaging division and the Spine unit of the department of Orthopedics of the UMC Utrecht, MRIguidance BV and Inspine BV.

The Project description: Adolescent idiopathic scoliosis (AIS) is a progressive 3D deformity of the spine affecting previously healthy children, substantially reducing their quality of life and creating a life-long burden of disease. Children with spinal deformities undergo frequent radiation-based radiographic follow-up at the vulnerable age of puberty. Besides the increased risk of cancer at later age, this 2D monitoring focuses management of the progressive spine curvature, most commonly with brace treatment, to correction only in the coronal plane.

3D-Spine will create a paradigm shift in AIS brace treatment. We will examine the growing juvenile scoliotic spine using the new radiation-free BoneMRI technique, allowing – for the first time – detailed longitudinal monitoring of both IVD and vertebral 3D deformity during brace treatment. This unique database will elucidate how the action of the braces effect both IVD and vertebral 3D deformity progression. This mechanobiological knowledge will then be used to develop our novel adaptive patient specific bracing concept. The ability to longitudinally and frequently monitor AIS patients using the safe BoneMRI technique, will allow patient-specific adaptation of the personalized brace, significantly increasing brace efficacy and minimizing the life-long burden for these adolescent children.

JOB DESCRIPTION

Your contribution will be to design, develop and test a new generic AI-based image synthesis methodology for BoneMRI reconstruction, making use of highly variable spine CT/MRI databases obtained from four different institutions, which were selected based on the different MRI scanner types and availability of data. This requires theoretical and experimental work to establish a learning algorithm and perform in silico pre-processing experiments to register CT and MRI data and set up a ‘learning’ algorithm. Novel in the approach will be the use of more than just imaging data, as well as the incorporation of model based information (MRI bloch simulations) in the algorithm development. This work will primarily be performed in close collaboration with the deep learning experts in the imaging division and the developers of MRIguidance BV.

During the project the developed method will be utilized in the clinical study to be initiated by the clinical PhD student, focusing on the investigation and monitoring of scoliosis patients undergoing brace treatment.

REQUIREMENTS

We are looking for a candidate that has:

• a PhD degree in image sciences, (applied) physics, (bio-)medical engineering or a comparable degree.
• a strong interest in image analysis/machine learning and MRI.
• an expertise that is also proven by your track record and peer-reviewed scientific publications.
• experience with deep learning
• experience with MRI data
• the ability to work in a multidisciplinary research environment to contribute to team efforts and projects, and guide the research efforts of students within the group.
• the ability to communicate effectively scientific ideas, foster collaboration and independent thinking.
• a proven proficiency in written and spoken English.

The initial contract will be for a period of one year, which will be extended to a total of max 3 years upon successful assessment. The salary is determined within the Dutch Collective Labour Agreement (CAO) for the University Medical Center Utrecht. Please provide your updated CV and a motivation letter and the contact details of two referees (to be send to P.Seevinck@umcutrecht.nl). Start date: as soon as possible.

CONDITIONS OF EMPLOYMENT

The maximum salary for this position is € 5.383,00 gross per month based on full-time employment. In addition, we offer an annual benefit of 8.3%, holiday allowance, travel expenses and career opportunities. The terms of employment are in accordance with the Cao University Medical Centers (UMC).

DEPARTMENT

Image Sciences Institute This position is hosted in the Image Sciences Institute, Imaging & Oncology Division and appointed at the Dept. of Orthopaedics, University Medical Center Utrecht (UMC Utrecht). The Division has access to >10 clinical MR (1.5, 3 and 7T) and CT scanners. With one of the largest concentration of medical imaging researchers in Europe, it is internationally renowned for its research and technology development in image acquisition, image analysis, high precision imaging, image-based prognosis, and image-guided therapy. You will work in a team of medical image analysis PhD students and staff, incl. radiologists. Development of clinical grade software under MDR guidelines will be done in collaboration with MRIguidance BV. For more information, see www.isi.uu.nl, www.imago.uu.nl and www.mriguidance.com

The Spine Unit in the Dept of Orthopedics is a global leader in scoliosis research. They have introduced an internationally recognized etio-pathogenic theory of adolescent idiopathic scoliosis focussed on the intervertebral disc, leading to many scoliosis focused publications in top 10% journals and dissertations in the last several years. The clinical practice of the department’s spine surgeons is focussed on all forms of scoliosis. In addition to the clinical studies run under this HH-TKI project, other clinical studies pertaining to scoliosis are currently in progress which will provide ample opportunities and data for development of machine learning-based imaging solutions.