top of page
  • Karl Turley

Championing the 3Rs with 3D imaging and AI

The world is rapidly changing with innovations in science, medicine and technology being made at a faster rate in the past two decades than at any time before in history. With the rise of technology and digital transformation, artificial intelligence (AI) has become more than the latest buzzword and is beginning to play a powerful role in powering science and drug discovery.

Reflecting the growing interest in all things AI, we have been sharing our experience in using 3D imaging and AI to measure and monitor subcutaneous tumours in pre-clinical oncology trials at key life science industry events including the recent Bio-IT conference in Boston and the AALAS national meeting in Baltimore.

Dr Juan Delgado, our lead data scientist, shared highlights from our research with our pharmaceutical industry co-development partner as part of the AALAS conference programme. In his presentation ‘Championing the 3Rs with 3D imaging and AI’ he explained how our BioVolume subcutaneous tumour monitoring solution – which combines 3D imaging, thermography and deep learning – improves experimental design, ensures better science through greater accuracy and reproducible measurements and tackles Refinement and Reduction head-on.

Refinement and Reduction

For animal welfare, the greatest benefits of combining 3D imaging with AI is greater confidence in when to stop testing and the reduction in the number of animals used in trials due to greater precision and accuracy of data. With the addition of thermography into our BioVolume solution, it is now possible to identify and measure tumours that are very small and that are pre-palpable, which means studies and the tracking of tumour volume can be started much earlier.

BioVolume 3D imaging is non-invasive and removes the potential for physical trauma to be caused to mice by the manual manipulation of tumours with callipers. Unlike other alternatives to callipers such as medical imaging, no sedation is required, reducing potential distress. And because tumours can be measured quickly and effectively, the overall handling of mice is reduced.

AI and the future  

We see a lot of potential in using AI and deep learning not just to identify tumours but to predict tumour symptoms.

Research carried out with our co-development partner highlights that a number of symptoms or tumour conditions can be observed using BioVolume, including redness, pallor, necrosis and ulceration. By developing our solution in the cloud, and thanks to machine learning tools, we are able to deploy our algorithms and analyse thousands of scans to train our system to identify key symptoms and predict conditions such as ulceration before they happen.

The early detection of tumours and symptoms bring significant 3R benefits and will allow more science to be achieved from the same number or even fewer animals. In addition to championing the 3Rs, ultimately, the hope is that thanks to AI, BioVolume will learn to spot the earliest signs of success and failure in drug development, so that a decision can be taken more quickly on whether to accelerate or terminate a trial so that crucial drugs can be brought to market more quickly.

bottom of page