Innovation in Patient Diagnosis: Senior AI Engineer David Armstrong Presents Paper at IMVIP 2024

Cirdan are delighted to announce that our colleague and Senior AI Engineer David Armstrong has submitted a research paper which has been accepted and is on display at the Irish Machine Vision and Image Processing Conference 2024!

David Armstrong, who has been working with Cirdan for over 3 years, has submitted a research paper as part of his PhD with Queen’s University Belfast to the 26th Annual IMVIP Conference, which has been accepted and is on display at the Meeting in Limerick on 23rd August!

Cirdan are a leading global provider of informatics and imaging solutions that enhance and speed up patient diagnosis. Our focus on constantly innovating to revolutionise the future of healthcare means that working with innovators such as David in the AI spheres helps us to reach new heights in healthcare innovation. This project specifically has seen David work as part of the Cirdan  team and in collaboration with Dr Niall McLaughlin and Dr Hui Wang, from the School of Electronics, Electrical Engineering and Computer Science, as part of his PhD studies.

David’s project focuses on using AI as part of the patient diagnosis journey; using generic colourisation to assist in virtual slide staining. Stained images are then used to examine specimens as a key part of the diagnosis process.

The project saw David study the effect of batch size on batch normalisation during inference. The paper analyses various interesting aspects of the common deep learning technique of batch normalisation in image colourisation, a related field to virtual staining. Specifically, it looks at how the colourfulness of the output is affected by the inference parameters. This research also presents a method for controlling the output by using the normalisation statistics collected during training. Whilst the paper discusses the subject of image colourisation, normalisation poses a significant problem for virtual staining. Due to the size of whole slide images, inference is performed using tiles. Normalisation must be smooth across the entire image, otherwise clear tiling artefacts present themselves. This is an active area of research, based on this paper.

We are thrilled that David is representing both Cirdan and QUB at the IMVIP in Limerick, where he is discussing his paper and discussing his research. He has already received some fantastic feedback and comments on his research, and how it truly holds the power to streamline workflows in a critical moment of each patient’s journey.

Commenting on the success of having his research paper accepted for the IMVIP, David said: ‘I am delighted to have my paper accepted for the IMVIP 2024 conference. I enjoyed the opportunity Cirdan has given me of studying this subject in detail and believe the knowledge gained will be useful in the extremely important problem of virtual staining of histological tissue. While colourisation is a slightly different subject, the main issue studied in this paper has also presented itself in subsequent work on virtual staining, making this work invaluable to future research‘. 

Dr Niall McLaughlin, Lecturer and PhD supervisor at QUB added: ‘David has done excellent work to have his first paper accepted at IMVIP 2024. David’s paper provides a strong foundation for the rest of his PhD investigating virtual staining, which has the potential to positively impact healthcare. As his supervisor, I’m very excited to be part of this meaningful collaboration between Cirdan and Queen’s’.

At Cirdan, we are truly passionate about improving patient wellbeing through innovation. Undertaking projects focusing on digital health technology is one method of assisting clinicians in expediting patient diagnosis. David’s project, alongside QUB and Cirdan, has demonstrated real benefits for patients and clinicians, which will be a true testament to the fantastic work achieved from our partnership.

Congratulations on the acceptance of your paper, David! We are delighted to see our relationship with Queen’s University Belfast strengthen year on year, and produce new innovations which ultimately help to expedite and improve the patient diagnosis process.