The medical digital images are routinely collected and archived by the NHS. Such images, especially when linked to other routinely collected health data, are extremely useful for research into areas including: early/preclinical diagnosis, disease progression, validation of treatment methods, development of novel computer vision methods for biomarker extraction, validation of novel algorithms and machine learning approaches and discovery and classification of disease types. Research using routinely collected imaging data for research has historically been under-utilised for many reasons including the impracticality of seeking consent of patients, challenges of de-identification of image data, absence of non-proprietary software for efficient handling this type of “big data” for research.
This project will develop and research innovative informatics methods for the processing of routinely collected images which meets the requirements of data governance and also researchers who wish to utilise the images in secure environments. The project will build upon and existing software suit, the Research Data Management Platform (RDMP) software. Challenges include
•Storage: methods to ensure the same image can be utilized for multiple research projects at once without having to create multiple copies •Speed and Scaling: methods to optimise the speed of processing of image big data •Anonymization and Governance: methods to ensure that no identifiable data is release for research
The methods developed will be tested on specific exemplars. The PhD will provide excellent training in applied health informatics and imaging research and software development for research. The project would suit someone from a computer science background
Apply To apply please send a cover letter, curriculum vitae and two references to: [email protected]