An interview with Dr. Peter Caie, Senior Principal Scientist, AI Diagnostics – Indica Labs

Dr. Peter Caie
Senior Principal Scientist, AI Diagnostics – Indica Labs
What did you study at the University of Edinburgh during your PhD in Digital and Quantitative Pathology and how did that prepare you for your role today?
My first job after university was with AstraZeneca where I pioneered work in high content biology and phenotypic screening. This involved the image analysis of multiplexed immunofluorescence cells. I wanted to expand this work into the field of pathology and so took up a PhD studentship at the University of Edinburgh in Prof. David Harrison’s lab. During my PhD I learned about histopathology, prognostic and predictive features within tissue, tissue staining and imaging, and further developed my image analysis skills. My research specialized in colorectal cancer and performing image analysis to quantify tumor budding, including their number, density, and protein expression. I also applied image analysis to quantify the vasculature, lymphovascular invasion, and immune cell infiltrate. Prof. Harrison and I also installed the first whole slide scanner in Edinburgh and began our journey into clinical reporting with digital pathology.
Could you elaborate on your collaboration with Indica Labs while you were at the University of St. Andrews as a Senior Research Fellow and a Lecturer?
After starting my position at St Andrews, I wrote my first ever PhD studentship grant in collaboration with Indica Labs. It was a studentship that required an industry partner and Kate, our CSO, was happy for Indica to support the grant and study. We were successful with our grant and as a result my group had access to the HALO® and HALO AI image analysis software to further our research into prognostic features of colorectal cancer. This included quantifying tumor budding and their spatial relationship to the immune cell infiltrate in multiplexed immunofluorescence and using HALO AI to quantify histopathological features like the desmoplastic reaction patterns of the stroma using H&E-stained whole slide images. The studentship and collaboration were a great success in both research output as well as the fantastic working relationship we had with Indica Labs. In fact, so much so that the PhD student, Ines Nearchou, published 6 research papers as a direct result of the collaboration and we now both work for Indica Labs.
How did you first learn about Indica Labs and what led you to join the company in 2021?
I first met Kate Lillard Tunstall while at conferences during my own PhD studentship and I learned about Indica Labs’ image analysis software through her. We kept in contact when I started my position at the University of St Andrews and as stated above, we collaborated for the full 6 years that I was there. I always enjoyed working with Indica Labs and found everyone personable and knowledgeable and knew the company to have a great work ethic and culture. An opportunity arose for me to apply to Indica Labs for a position as Principal Scientist who would be in charge of both developing and deploying our clinical AI products. Although I enjoyed my time at St Andrews, I felt that this was a fantastic next challenge for me, working in a company I enjoyed collaborating with and being given the chance to have our research translate into software products that directly impact and benefit both pathologists and patients.
What is your role here at Indica Labs?
I am currently Senior Principal Scientist at Indica Labs and manage the AI Diagnostics team. We develop pre-built commercial AI solutions for the field of digital pathology. Our work results in products designed for the clinical and life science fields. We develop these products in direct and close contact with clinicians and life science researchers from multiple global institutes throughout the entire process, from concept and feasibility design to sourcing data and generating annotations, developing the algorithms, and testing and deployment of the final product. In this way we can ensure that the final product is both generalizable and clinically useful to the end users.
What types of projects does the AI Diagnostics team work on?
Our main focus is developing AI-based software to assist pathologists in their routine clinical reporting, where we design products to impact across the field of pathology. These can be in the form of software for pre-clinical use, such as our automated artifact detection software (SlideQC BF), for primary diagnosis, such as our prostate diagnostic and grading software (HALO Prostate AI), for the reporting of prognostic and predictive biomarker quantification (Breast IHC AI and Lung PD-L1 AI), and for use in the field of molecular pathology, such as our software to assist in selecting regions in samples requiring macrodissection and in standardizing the quantification of tumor content (Lung Macrodissect AI). We also develop AI-based software for our new HALO AI Apps product range. These AI Apps are designed for the field of life science and are pre-built to help researchers develop their own image analysis solutions more quickly and with fewer resources.
What do you enjoy most about your role?
I am lucky as I enjoy many aspects of my role here at Indica Labs. Firstly, I get to work directly with my fantastic team within AI Diagnostics. Not only are they very hard-working and knowledgeable in their field of expertise, but I get on with all of them on a personal level. This is also the case for my other colleagues at Indica Labs. Additionally, I have the opportunity to work closely with clinical collaborators, where I get to understand more about their workflow and how AI-based software for digital pathology can assist them to improve turnaround times and increase the quality and efficiency of their diagnostic reporting. I greatly enjoy the R&D process, from the very start of product planning to the pride and satisfaction of launching a new product. Finally, I enjoy, and am motivated by, the fact that the research we do, and the products we develop, can directly impact patients and help ensure they get the fastest and most accurate diagnosis and treatment.
What do you appreciate most about working at Indica Labs?
I very much appreciate the opportunity to be able work with fantastic people, to be able to translate research to clinically useful software, and also the work culture. The culture at Indica encourages hard work while being rewarding and having a very personable approach. Senior management are always available to listen to, and are understanding of, everyone at the company, and value the impact each employee makes on Indica’s success as a whole.
What project are you most proud of contributing to at Indica Labs?
I cannot single out a project that I am most proud of. I am very proud of starting AI Diagnostics from scratch and developing it into a highly productive and efficient team that is involved in every aspect of developing clinical AI. I am also proud of the processes and hard work that my team puts into generating the final product. This includes identifying which product to work on next, where to source the data, which clinicians to collaborate with, developing the AI algorithms, and generating an end product. The work also includes clinical and analytical validation studies, generating marketing and product release collateral, and taking our clinical products through regulatory approval.
How do you spend your time outside of work?
I am the father of two fantastic children, which takes up a lot of my time outside of work. I am also lucky to live in a beautiful part of Scotland that gives us access to incredible beaches, harbor towns, woodland, and mountains and I take the family on many adventures in the outdoors. When I get time to myself, I enjoy reading, going to the pub, watching football and films, as well as very amateurish DIY projects in the house and garden. I should probably also mention that I am supported by, and enjoy spending time with, my wife… who is an artist and takes inspiration from the surrounding Scottish scenery for her work.
What are you most excited about in the field of AI in digital pathology?
I think the thing that excites me the most is the incredible pace that the field is developing at. We are in a truly pioneering field. There are constant new developments in computer vision being applied to the field of digital pathology. These new developments ensure that clinical AI continues to improve in performance and allows new features, functions, and approaches to be developed and applied.
Furthermore, as the field of pathology picks up pace in its translation to going fully digital, more institutes will be able to deploy AI-based diagnostic products which can assist with the large workload pathologists are faced with and improve patient care. We are at the beginning of a new era of pathology, that brings many challenges but is at the same time very rewarding and extremely exciting to be a part of.
SlideQC BF is intended to be used as a quality control tool for whole slide images and is not regulated as a medical device in the EU/UK. SlideQC BF is For Research Use Only in the USA and is not FDA cleared for clinical diagnostic use. SlideQC BF is accessed via the HALO AP® enterprise digital pathology platform.
HALO AI Prostate is CE-marked for in-vitro diagnostic use in Europe, the UK, and Switzerland. HALO AI Prostate is For Research Use Only in the USA and is not FDA cleared for clinical diagnostic use. HALO AI Prostate is accessed via the HALO AP® enterprise digital pathology platform.
Breast IHC AI is For Research Use Only and not intended for clinical diagnostic use. Breast IHC AI is accessed via the HALO AP® enterprise digital pathology platform.
Lung PD-L1 AI is For Research Use Only and not intended for clinical diagnostic use. Lung PD-L1 AI is accessed via the HALO AP® enterprise digital pathology platform.
Lung Macrodissect AI is not a medical device in the EU/UK and is not intended to be used for diagnostic purposes. Lung Macrodissect AI is accessed via the HALO AP® enterprise digital pathology platform. Lung Macrodissect AI is For Research Use Only in the USA and is not FDA cleared for clinical diagnostic use.
HALO AP® is CE-IVDR marked for in-vitro diagnostic use in Europe, the UK, and Switzerland. HALO AP® is For Research Use Only in the USA and is not FDA cleared for clinical diagnostic use. In addition, HALO AP® provides built-in compliance with FDA 21 CFR Part 11, HIPAA, and GDPR.