US-based medical software provider Subtle Medical has received the US Food and Drug Administration (FDA) approval for its MRI acceleration and enhancement software, Subtle-HD.
With the FDA approval, Subtle Medical officially launched its Subtle-ELITE suite of AI-powered software tools, comprising SubtleHD, SubtleSYNTH, and SubtleALIGN solutions.
All three software solutions are designed to improve image quality and reduce MRI scan times with an 80% scan time reduction in certain sequences.
Subtle-ELITE provides enhanced value to radiologists and imaging institutions by better optimising the MRI workflow than other image enhancement solutions, said Subtle Medical.
Subtle Medical chief product officer Ajit Shankaranarayanan said: “The FDA clearance of Subtle-HD marks a significant milestone in AI-driven radiology, giving hospitals and imaging centres a powerful solution to increase efficiency, improve image quality, and optimise workflows, all without additional hardware costs.
“By combining SubtleHD, SubtleSYNTH, and SubtleALIGN into a single, comprehensive package, Subtle-ELITE is setting a new benchmark for the future of MRI.”
SubtleHD uses advanced denoising and sharpening techniques to produce ultra-clear images, enabling significantly faster scan protocols applicable to all body regions.
SubtleSYNTH leverages deep learning to generate MR imaging contrasts from existing sequences, with 100% acceleration and producing STIR images in less time.
It is offered for spine STIR with plans to expand to other body regions.
SubtleALIGN is designed to automatically align brain MR to the ideal anatomical position, reducing variability and improving consistency across longitudinal studies.
Subtle Medical launched its SubtleMR and SubtlePET solutions as the first third-party AI-powered image enhancement solutions, currently deployed in more than 16 countries.
Its SubtleHD builds on the same foundation, using convolutional neural networks to deliver sharper images with greater time savings compared to SubtleMR, said the company.
