Professor Nasir Rajpoot is a computer scientist at the University of Warwick, but he also holds an Honorary Scientist position at the Department of Pathology, University Hospitals Coventry & Warwickshire NHS Trust. How did these two seemingly unrelated areas of expertise come together? The answer is through personal experience.

A graduate of the University of Warwick, Professor Rajpoot completed his PhD in image processing at the University in 2001 and began working as a member of the academic staff within the Computer Sciences Department later that year. But when cancer affected his own family, Professor Rajpoot began to consider how he may help the research and diagnosis procedure.

Persistent pursuit

He explains: “Fifteen years ago I had a number of back to back experiences with cancer in my extended family and this was a major trigger for me to want to work in the field of cancer research.

“My work prior to my personal experiences had nothing to do with cancer,” continues Professor Rajpoot. “But when we went through these experiences, I knew there were aspects to my work which could be useful and I wanted to try and use my skills in image processing to probe questions in medical and life sciences. I had no track record and it was difficult to get people interested at first. So I used the persistent pursuit tactic – I kept knocking on doors.”

In the end it was a chance encounter whilst working with a colleague at Yale that Professor Rajpoot was at last introduced to the area of cancer image processing.

Professor Rajpoot began work involving cancer image analysis and set a couple of PhD students in his lab onto projects within the field. The work snowballed and he founded and is now head of the Tissue Analytics Lab (TIA) within the Computer Sciences department at Warwick.

The digital revolution

One of the breakthroughs Professor Rajpoot and the team at the TIA lab have made is sophisticated analysis of digitised high-resolution microscopic images of tissue sample slides. Working alongside colleagues at The University of Warwick, Coventry and Warwickshire Pathology Services (CWPS) and University Hospitals Coventry and Warwickshire (UHCW) NHS Trust, the TIA lab has helped to remove the need for physical slides, which had to be filed and stored and transported between sites upon request. Now technology allows the slides to be scanned in as high quality digital images, which is hugely advantageous to medical experts and patients alike.

Professor Rajpoot explains: “Digitised slides have a number of important advantages. Pathologists can securely access them from any location, get second opinions from colleagues in other sites, or even other countries, and can explore the role of computer-assisted analysis that is not feasible with conventional microscopes. It is also much more efficient - using a digital system avoids any physical samples being lost or damaged, especially during transport and it also has the ability to cut slide processing time by 12 per cent.”

The next steps

In 2014, UHCW, became the first pathology department in the country to use software and digitised slides of tissues for clinical diagnosis, allowing pathologists to view samples in high resolution on their computer screens and from any location.

A year later the University of Warwick and the UHCW NHS Trust were together named as a Centre of Excellence for Digital Pathology.

Now Professor Rajpoot is continuing to further the advantages of having the tissue sample slides scanned.

He says: “Now we have the images digitised, we open up a whole new world of possibilities. A modern day digital slide scanner produces image data of approximately 50 gigabytes each. We now scan about 1,000 samples a day at University Hospital Coventry and Warwickshire which amounts to a huge amount of data, with clues for clinical outcomes for patients hidden within.

“If we can put together algorithms which match image based signatures to patient pathways and outcomes, we can help clinicians better diagnose disease and recommended treatment courses for more patients. Using the predictive power of the data we can look for patterns – and where we see something which fits a known pattern and looks aggressive, the disease can be nipped in the bud early on. If we find something which registers as not so aggressive, clinicians can be less heavy handed with treatment – which is better for the patient.”

International collaborations

The TIA lab at Warwick is one of the leaders in digital pathology image analysis research in the UK, and computerised methods such as those developed in the lab are not only benefitting patients at UHCW but could transform healthcare globally. Professor Rajpoot and colleagues are currently working in collaboration with Shaukat Khanum Memorial Cancer Hospital & Research Centre (SKM), the largest cancer hospital in Pakistan, gathering samples and data from patients with oral cancer – one of the deadliest and most prevalent cancers in the Indo-Pakistan region.

The project will involve UHCW digitally scanning slides of human tissue samples from SKM’s oral cancer patients. Sophisticated computerised image analysis at Warwick will then generate a repository of important information about the different types of cancer cells which can be found in oral cancers in the subcontinent. Again the aim is to make treatment more targeted and efficient by harnessing the power of the data.

Algorithms of the avant-garde

So through his persistent pursuit, Professor Rajpoot has become a revolutionary and he is sure his digital revolution will lead to better understanding and better treatment of cancer in the years to come.

He says: “I am confident that in the next five years the algorithms we are developing will be deployed in a clinical setting. They will be available to empower oncologists to make decisions on individual patients using their microscopic disease footprint and comparing it to the powerful data gathered from hundreds of thousands of digital samples.”

Published
19 April 2017

Terms for republishing
The text in this article is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).

 

Share