Low-Field MRI for Knee Imaging: Current Insights and Evidence From a Systematic Review.
Authors
Affiliations (3)
Affiliations (3)
- Radiography and Diagnostic Imaging, School of Medicine, University College Dublin, Dublin, Ireland.
- School of Health Sciences (HESAV), University of Applied Sciences and Arts Western Switzerland (HES-SO), Lausanne, Switzerland.
- Department of Diagnostic & Interventional Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland.
Abstract
Low-field magnetic resonance imaging (MRI) has regained interest for musculoskeletal imaging due to advances in magnet design, hardware optimization, and artificial intelligence (AI)-based reconstruction, offering improved accessibility, patient comfort, and sustainability. This systematic review aimed to explore and summarize the current status of low-field MRI for knee investigations. Systematic review on MEDLINE, Cinhal, EMBASE, Web of Science, Scopus, and Google Scholar databases. PROSPERO protocol (CRD42024544354). Approximately 3400 participants (8-89 years) encompassing healthy volunteers, community-based cohorts, and patients with diverse knee pathologies were evaluated. Few studies incorporate phantom experiments. This review covered a range of very-low-field (< 0.1 T), conventional low-field (0.1-< 0.29 T), and upper-range low-field (> 0.3 T-0.55 T) with standard, extremity, open, and weight-bearing/standing systems. SE/FSE and GRE sequences, 3D GRE quantitative protocols, and AI-based were used. Two reviewers independently extracted data on knee protocols, technical parameters, diagnostic performance, and reported advantages and limitations. Study quality was assessed using appropriate JBI critical appraisal tools. A narrative synthesis and descriptive analysis of the findings from the included studies are provided due to the heterogeneity of the intervention and outcome. From 3726 records, 31 were included. Low-field MRI demonstrated high diagnostic performance for meniscal and ligamentous injuries but lower sensitivity for early cartilage degeneration. Weight-bearing MRI provided distinct functional information but not direct equivalence to standard MRI. AI-enhanced reconstruction was evaluated in a limited number of single-center, vendor-specific studies and was associated with improved image quality and diagnostic confidence. Important limitations included lower signal-to-noise ratio, reduced spatial resolution, protocol heterogeneity, and limited pathology-validated evidence. Low-field MRI may provide a clinically useful complementary option for selected knee meniscal and ligamentous indications. However, heterogeneous evidence, influenced by field strength, scanner design, sequences, and reference standards, limits claims of equivalence with high-field MRI, particularly for subtle cartilage lesions. Advances in hardware and AI have improved image quality and efficiency; however, evidence is limited and larger standardized multicenter studies are required. 5. Stage 5.