NOW CLOSED- Machine Learning augmented hierarchical tomographic image reconstruction of human organs

Full home student tuition fees and stipend of ca. £17,609 per annum (for up to 3 years, with possible 4th year if required.) NOW CLOSED

University College London working with ESRF, and part of the Centre for Doctoral Training in Intelligent, Integrated Imaging In Healthcare (i4health).

Supervision Team:
UCL: Prof. Peter D. Lee and Prof Sandro Olivo
ESRF: Dr Paul Tafforeau

Project Background: 

Biological organisms are hierarchically structured, spanning from molecules to whole organs via cells and organotypic units. For humans these scales span from the nanometre to the metre. There have been significant advances in imaging at both extremes, with developments in electron microscopy imaging at the nano-scale of tiny biopsies, to clinical CT and MRI at the whole body scale. However, there is a gap in imaging techniques to span the scale from histology (micron) to clinical CT/MRI (100’s of microns). Therefore, there is a drive to develop imaging techniques which achieve cellular resolution deep in soft-tissue to correlate cellular level function to whole organ processes.  

X-ray tomography provides high resolution when the sample size is limited, enabling imaging of millimetre-sized mouse organs at a cellular level. However, in human organs which can be 100’s of millimetres thick, tiny tissue biopsies must be taken to achieve cellular resolution. These biopsies distort, making correlation of these high-resolution results to low-resolution images challenging.  

You will be part of a team developing a new technique that decouples resolution from specimen size using the exceptional coherence and high energy provided by the European Synchrotron Research Facility’s Extremely Brilliant Source (ESRF-EBS) upgrade to a 4th generation source,. This technique, Hierarchical Phase-Contrast Tomography2 (HiP-CT), can achieve local cellular resolution of soft-tissues using phase information obtained through beam propagation. Specifically, 1 um voxels were achieved locally in a range of intact 150 mm diameter human organs, as shown for a human lung lobe in Fig. 1) https://mecheng.ucl.ac.uk/HiP-CT) 

Your role will be developing Machine Learning based techniques for improving the tomographic reconstructions using information from high resolution scans in the coarse scans. You will also explore other techniques like dual-energy and phase contrast technqiues.  

The PhD project is jointly supervised by Profs. Peter Lee (Mech. Eng.) and Sandro Olivo (BioMed.Phy), with Dr. Paul Tafforeau from ESRF (Grenoble France) where you will spend spend a portion of your time. The results will form part of the Human-Organ-Atlas.esrf.eu, a public database we’re compiling of our organs in health and disease, funded by the Chan Zuckerberg Initiative with a goal of eradicating disease. 

Person Specification:

Applicants should ideally have a first class undergraduate degree (or equivalent) in Physical Sciences (Computer Science, Engineering, Mathematics or Physics). Knowledge of image processing and strong computer programming skills are required. Applicants should have an interest in bio-medical imaging, Machine Learning, and synchrotron x-ray tomography. Excellent organisational, interpersonal and communication skills, along with a stated interest in interdisciplinary research, are essential.

The position is open to students on Home Fees and applicants whose first language is not English are usually required to provide evidence of proficiency in English by UCL. Please do not enquire about this studentship if you are ineligible. Please refer to the following website for eligibility criteria: https://www.ucl.ac.uk/prospective-students/graduate/research-degrees/mechanical-engineering-mphil-phd 

How to apply:

Please complete the following steps to apply.

•    Send an expression of interest and current CV to: Prof peter.lee@ucl.ac.uk and cdtadmin@ucl.ac.uk
Please quote Project Code: 22007 in the email subject line.
•    Make a formal application to via the UCL application portal https://www.ucl.ac.uk/prospective-students/graduate/apply . Please select the programme code MPhil Medical Imaging RRDMEISING01 and enter Project Code 22007 under ‘Name of Award 1’ 

Application Deadline:

Applications considered on a rolling basis until position is filled. Latest start date available Sept 2022.
If shortlisted, you will be invited for an interview.