Deep Learning Applications in Magnetic Resonance Imaging: Has the Future Become Present?

Area of Science:

• Radiology
• Artificial Intelligence
• Medical Imaging

Background:

• Deep learning (DL) represents a significant advancement in radiology.
• Its applications are poised to transform daily clinical practice in image acquisition and reporting.

Purpose of the Study:

• To review current deep learning technologies, focusing on magnetic resonance (MR) image reconstruction.
• To examine the technical principles, clinical translation, and literature on DL in MR image reconstruction and post-processing.

Main Methods:

• Literature review of deep learning applications in magnetic resonance imaging reconstruction.
• Analysis of technical principles and clinical translation of DL techniques.
• Synthesis of current research on DL for MR image reconstruction and post-processing.

Main Results:

• Deep learning is a major emerging force in radiology.
• DL-based MR image reconstruction offers reduced acquisition times and enhanced image quality.
• These techniques can accelerate workflows and potentially alleviate scanner availability issues.

Conclusions:

• Deep learning technologies, especially in MR image reconstruction, hold immense promise for radiology.
• The integration of DL is expected to permanently alter daily routines and workflows.
• DL offers significant advantages in diagnostic support, efficiency, and image quality.