Deep Recon

by United Imaging Healthcare  · Based in China → — To Bring Equal Healthcare for All
Radiology

Contact for pricing
Regulatory Status Disclosed

Overview

Deep Recon (also integrated as uAIFI DeepRecon) is an AI-powered, deep learning-based image reconstruction software developed by Shanghai United Imaging Healthcare Co. It is designed to process raw imaging data to enhance diagnostic image quality across both computed tomography (CT) and magnetic resonance imaging (MRI) modalities.

The tool is utilized by radiologists, imaging technologists, and clinical specialists across hospital departments, outpatient imaging centers, and oncology clinics. It supports a broad range of clinical applications, including head, chest, abdomen, cardiac, and vascular imaging.

Deep Recon integrates directly into the reconstruction chain of compatible United Imaging scanners (such as the uCT and uMR series). In clinical workflows, the software operates automatically during the acquisition phase. It utilizes deep neural networks to perform intelligent denoising and K-space-interpolation based image super-resolution. This allows clinical practices to achieve high-resolution anatomical details and improved signal-to-noise ratio (SNR) without requiring longer scan times.

  • Intelligent Denoising: Suppresses unwanted image noise while preserving sharp anatomical boundaries and low-contrast detectability.
  • Super-Resolution Reconstruction: Employs deep learning algorithms to interpolate data, enhancing spatial resolution and image sharpness.
  • Multi-Modality Compatibility: Available as a reconstruction solution for both advanced CT and MRI systems.

Reviewed by Pouyan Golshani, MD — Interventional Radiologist

Key Features

  • Deep learning-based image reconstruction
  • Intelligent denoising for improved signal-to-noise ratio (SNR)
  • Increased image sharpness and resolution
  • Potential for reduced scan time
  • Part of uAIFI technology platform
  • High fidelity with double safety lock design (image and K-space domain)
  • Applicable for multi-bed scanning like whole spine and whole-body imaging
  • Enhances diagnostic accuracy and patient care

Use Cases

  • Enhancing diagnostic accuracy in MRI examinations
  • Improving patient comfort through faster and quieter scans
  • Optimizing workflow and throughput in imaging departments
  • Facilitating high-resolution imaging with ease
  • Supporting comprehensive advanced clinical applications

What Physicians Need to Know

AI Model Architecture (deep learning approach)
Deep Recon utilizes deep learning-based image reconstruction algorithms. It employs neural networks trained on thousands of MRI scans to differentiate true anatomy from artifacts, reconstructing high-quality images from less data. Specifically, DeepRecon uses a convolutional neural network (CNN) trained model for image denoising and K-space interpolation-based image super-resolution. For CT imaging, the DELTA (Deep Learning Trained Algorithm) leverages an extensive training dataset and a sophisticated 3D neural network.
Supported Modalities (CT/MRI/X-ray/US)
Deep Recon (uAI DeepRecon) is primarily integrated into United Imaging's MRI systems, such as the uMR Omega 3T and uMR 680 1.5T. United Imaging also offers deep learning reconstruction for CT (DELTA) and PET (uAI HYPER DLR).
Processing Speed (per study)
Deep learning reconstruction can reconstruct new images in milliseconds, significantly faster than traditional methods. This technology enables MRI scans to be up to 80% faster, with brain imaging achieving up to an 80% reduction in scan time, and cardiac and abdominal imaging seeing 40-60% reductions.
FDA Clearance Pathway (510k/De Novo)
Deep Recon (DELTA) for CT imaging has received 510(k) clearance. United Imaging has also received FDA clearance for uAI HYPER DLR (Deep Learning Reconstruction) for PET images. The uMR 680, which integrates DeepRecon, also received 510(k) clearance.
Reading Room Workflow Impact
Deep Recon aims to improve workflow efficiency by reducing scan times, which can lead to increased patient throughput. Faster scans can also reduce patient anxiety and the need for sedation, particularly in pediatric imaging. The technology also aims to enhance diagnostic confidence by improving image quality, including signal-to-noise ratio (SNR) and image sharpness.
DICOM Support & Standards
While not explicitly detailed for 'Deep Recon,' medical imaging AI tools generally rely on DICOM standards for interoperability. PACS systems, which integrate AI, heavily rely on DICOM for data generation, transmission, and display. Deep learning reconstruction algorithms can be DICOM-based.
RSNA/ACR Validation
United Imaging regularly showcases its AI technologies, including deep learning reconstruction, at RSNA. For example, at RSNA 2025, United Imaging highlighted its native AI technologies and intelligent workflows.
Physician Tip

Deep Recon can significantly reduce MRI scan times, which is beneficial for patients who experience claustrophobia or have difficulty remaining still, such as pediatric patients. The improved image quality, including enhanced SNR and sharpness, may lead to increased diagnostic confidence and potentially better visualization of subtle pathologies. When utilizing Deep Recon, consider the potential for dose reduction in CT imaging while maintaining diagnostic image quality, always in consultation with a radiologist and physicist.

Deep Recon is integrated into United Imaging's MRI systems as part of their uAIFI technology platform. This allows for seamless operation within the scanner's native reconstruction pipeline. For CT, the DELTA algorithm is deployed within the uCT acquisition and processing software. These integrations aim to optimize workflow and enhance image quality directly at the point of acquisition and reconstruction.

Details

Category Radiology & Imaging AI
Pricing Contact for pricing
DeploymentIntegrated into United Imaging's MRI systems (e.g., uMR 680, uMR OMEGA)
Compliance
BAA AvailableUnknown
HIPAA CompliantUnknown
FDA Status Cleared AI-estimated
Integrations
EHR Not specified
Specialties Radiology

Ratings & Reviews

No reviews yet. Be the first to review this tool!

Rate Deep Recon

Clinical Value
Ease of Use
Integration
Support & Docs
Value for Money
No press coverage or publications on file yet. Know of some? Suggest an edit →

Videos

Product demos, reviews, and walkthroughs for Deep Recon.

View all on YouTube

Frequently Asked Questions

DeepRecon utilizes deep learning to enhance image sharpness and signal-to-noise ratio (SNR) in MRI scans, leading to clearer anatomical details and improved diagnostic confidence. It is integrated into United Imaging's MRI systems, such as the uMR 680 and uMR OMEGA, to provide high-resolution imaging and accelerate scan times.
While DeepRecon significantly improves image quality and processing time, deep learning-based methods can still lag behind classical numerical optimization techniques in terms of accuracy and generalizability in some complex applications. For instance, in coronary CT angiography, a notable limitation was the model's inability to detect ostially occluded grafts due to the indiscernibility of a collapsed graft lumen.
DeepRecon is a deep-learning based image processing algorithm that has undergone evaluation by American Board of Radiologists certified physicians. These evaluations confirmed that DeepRecon meets the requirements for clinical diagnosis, with images rated equivalently or higher in diagnostic quality.
Other deep learning reconstruction technologies in the market include Siemens Healthineers Deep Resolve, GE Healthcare AIR Recon DL, and Philips SmartSpeed AI. While all aim for faster scans and better image quality, they differ in their approach to applying AI, with some using a hybrid 'physics + AI' method and others leaning more heavily on pure deep learning or integrating AI directly into the acquisition pipeline.
United Imaging's strategy is to provide price transparency and simplicity for its medical imaging equipment and associated software. While specific pricing for DeepRecon is not publicly detailed, AI radiology solutions generally follow a per-study or per-scan pricing model, with costs typically ranging from $3 to $25 per study, influenced by factors like study volume, modalities covered, and integration requirements.

Related Tools

AZmed Rayvolve® (AZtrauma)
AZmed
Radiology & Imaging AI
Exo Iris
Exo Imaging, Inc.
Cardiology AI
PanEcho
Yale School of Medicine (Cardiovascular Data Science Lab)
Cardiology AI
PanEcho is an AI tool from Yale School of Medicine that interprets echocardiograms with high accuracy, performing 39 diagnostic tasks and detecting conditions like severe aortic stenosis and systolic dysfunction.
Pixelplan Medical OS
Pixelplan Medical
Radiology & Imaging AI
Pixelplan Medical transforms surgical planning, execution, and outcomes with GPU-accelerated AI. From pre-op to post-op, one platform sees it all.
Indica Labs HALO AI
Indica Labs
Drug Discovery & Research
HALO AI is a deep-learning add-on for the HALO image analysis platform, helping users solve image analysis challenges in segmentation, classification, and phenotyping for research.
DentRead
DentRead
Clinical Decision Support & Reference
DentRead provides AI-assisted radiograph analysis built for dental workflows, offering quick X-ray analysis to support clinical diagnosis.

See all Radiology & Imaging AI tools →

More from Shanghai United Imaging Healthcare

HYPER AiR
United Imaging Healthcare
Radiology & Imaging AI
HYPER AiR (uAI HYPER DLR/DPR) is an AI-powered deep learning reconstruction technology by United Imaging Healthcare, designed to enhance image quality, reduce noise, and shorten scan times in PET/CT and MR medical imaging.
HYPER DLR
United Imaging Healthcare
Oncology AI
uAngio AVIVA CX
United Imaging Healthcare
Radiology & Imaging AI
uCT ATLAS Astound with uWS-CT-Dual Energy Analysis
United Imaging Healthcare
Radiology & Imaging AI
The uCT ATLAS Astound with uWS-CT-Dual Energy Analysis is an AI-powered multi-slice computed tomography (CT) system by United Imaging Healthcare, offering advanced whole-body imaging, dual-energy analysis, and low-dose capabilities for diverse clinical applications.

View company profile →

Suggest an Edit → | Report an issue | Last Verified: 2026-11-13 | First Added: 2026-04-19
AI Tool Finder
AI-powered search. Results may not be comprehensive.