The development of artificial intelligence (AI) has created opportunities for various industries to enhance, streamline and speed-up workflows. In healthcare, AI-driven technology is being designed to support diagnoses, simplify treatment delivery, optimise the predictability of outcomes and increase clinical efficiency. It achieves this through a deep learning approach.
The basics of deep learning
Deep learning is a special form of machine learning that teaches computers to learn by example. It is the technology that powers driverless cars and smart speakers. The software learns directly from large sets of labelled data and neural network architectures, without the need for manual feature extraction. The result is a super-human performance and highly accurate results in whatever task it is completing.
Deep learning in medical imaging
One of the areas of healthcare that deep learning is being applied to is medical imaging, specifically PET and MRI scanning. The aim of the game has been to enhance the speed, comfort, and safety for patients – without reducing image quality and therefore impacting the clinician’s ability to diagnose issues and develop treatment plans effectively.
Though the technology is still being developed, the results so far have been extremely promising, and some solutions are already FDA-approved and being implemented in hospital and imaging centre environments. There is even evidencei to show that technology AI-driven software can deliver PET images of a quality comparable to normal scans, but at half the typical scanning time. This has important ramifications for healthcare services, as it enables much faster scanning times for quicker treatment pathways, improved patient care and greater efficiency and productivity for the professional team.
Other research has found similarly beneficial results when using deep learning-based software with whole-body PET scans, both with reduced scan time and radiation dosage.ii
The technology in question
What was the technology that made this possible? The SubtlePET™ and SubtleMR™ software from Subtle Medical – a trusted partner of Getz Healthcare – is an innovative solution that has been developed by clinicians. It is already being utilised in hospitals and imaging centres, with professionals and patients alike benefiting from the faster, more comfortable and highly accurate scans it facilitates.
Easy to integrate within existing workflows, the SubtlePET and SutbleMR software are vendor-neutral options that can be used alongside a wide range of imaging technologies. For more details about our AI Efficiency Solutions, visit our product page or Contact Us today.
i Weyts, K., Lasnon, C., Ciappuccini, R. et al. Artificial intelligence-based PET denoising could allow a two-fold reduction in [18F]FDG PET acquisition time in digital PET/CT. Eur J Nucl Med Mol Imaging 49, 3750–3760 (2022). https://doi.org/10.1007/s00259-022-05800-1
ii Chaudhari, A.S., Mittra, E., Davidzon, G.A. et al. Low-count whole-body PET with deep learning in a multicenter and externally validated study. npj Digit. Med. 4, 127 (2021). https://doi.org/10.1038/s41746-021-00497-2