Deep Resolve

Faster than ever before

AI-based image reconstruction is the next chapter in the digital future of magnetic resonance imaging. With Deep Resolve technologies, Siemens Healthineers applies AI to image acquisition and uses deep learning for accelerated image reconstruction. With a convolutional neural network at its core, it can radically shorten acquisition time without compromising image resolution. We have a solid foundation of state-of-the-art med-tech and curated data from an international network and continue to drive the digitalization of healthcare and improve the patient experience by securely developing AI-powered solutions in MRI.

Targeted denoising and deep learning methods power Deep Resolve Gain and Deep Resolve Sharp technologies for sharper images acquired in faster scans. Deep Resolve Sharp generates a high-resolution image from low-resolution input. By including raw data in the reconstruction process, clinically robust results are achieved that, in combination with Deep Resolve Gain targeted denoising, outperform classical denoising and interpolation. Deep Resolve Boost¹ is our first raw data-to-image deep learning reconstruction technology that enables high SNR and radically accelerated image acquisition. Deep Resolve Swift Brain¹ neuro acquisition employs a multi-shot echo planar imaging (EPI) approach that achieves all relevant contrasts in just one go, enabling a total neuro exam in two minutes.