Magnetic resonance imaging (MRI) is the preferred method for evaluating internal knee injuries, particularly in sports-related cases. However, traditional MRI protocols often involve long scan times, high operational costs and energy consumption. To address these limitations, recent advancements have focused on acceleration techniques. A recent prospective study explored the clinical impact of integrating deep learning (DL)-enhanced parallel acquisition technique (PAT) with simultaneous multislice (SMS) imaging in routine knee MRI. The findings indicate that the combined approach, particularly the P4S2 protocol, offers substantial improvements in image quality, acquisition time and sustainability compared to standard two-fold PAT imaging.
Enhanced Imaging Performance
The study included 34 adult patients who underwent knee MRI scans using three protocols: standard two-fold PAT (P2), DL-enhanced four-fold PAT (P4) and a combination of P4 with two-fold SMS (P4S2). Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were evaluated to quantify image clarity and tissue differentiation. P4 imaging significantly outperformed P2 in SNR and CNR, but P4S2 achieved the highest values across all measurements. This translated to clearer anatomical delineation, particularly in soft tissues and bone structures.
Despite P4's improvements in image clarity, it was less effective in evaluating cartilage due to over-smoothing, which impaired the visibility of fine structures. In contrast, P4S2 consistently outperformed both P2 and P4 in all assessed anatomical areas, including cartilage, ligaments and menisci. Subjective image evaluations using a 5-point Likert scale confirmed these findings, with radiologists rating P4S2 significantly higher in all quality metrics. Radiomics analysis further demonstrated that P4S2 yielded a greater range of textural and gray-level differences compared to both P2 and P4, reinforcing its superior diagnostic potential.
Time and Energy Efficiency
The time savings associated with DL-enhanced protocols were significant. Standard P2 imaging required approximately 20 minutes per scan, while P4 reduced this by 31% and P4S2 by 41%. The average scan time decreased to 13.8 minutes with P4 and to 11.8 minutes with P4S2. These gains translated into improved scanner throughput, enabling more patients to be scanned each day. With P4S2, a scanner could perform 22 scans daily, compared to 16 with P2, representing a 37.5% increase in daily capacity.
From an energy perspective, P4S2 demonstrated notable advantages. P2 imaging consumed 0.54 kWh per scan, whereas P4 and P4S2 required only 0.20 kWh and 0.15 kWh, respectively. On an annual basis, assuming 20 scans per day, this results in energy savings of over 2,000 kWh for P4S2 per scanner. These reductions contribute to lower operational costs and a smaller environmental footprint, aligning with emerging healthcare sustainability goals.
Must Read: MRI in Knee Evaluation: ESR’s Best Practices and Recommendations
The economic impact was also substantial. Based on activity-based costing, P4 and P4S2 reduced radiology staff workload, lowering full-time equivalent (FTE) requirements. P4S2 saved 8.2 minutes per scan, equivalent to 683 hours per year, resulting in annual labour savings of over €18,000. When accounting for increased scan capacity and a reimbursement rate of €200 per scan, P4S2 could generate over €300,000 in additional annual revenue per scanner.
Clinical Applicability and Limitations
The integration of DL and SMS in knee MRI offers clear advantages in terms of diagnostic confidence, workflow efficiency and sustainability. The P4S2 protocol achieved superior anatomical delineation across all evaluated structures, eliminated common artefacts and maintained robust image consistency. It also resolved the limitations of over-smoothing, as seen in P4, making it more reliable for clinical decision-making.
While the study findings support clinical adoption of P4S2 imaging, some limitations must be considered. Conducted at a single centre with one MRI system, the generalisability of results to other settings or equipment remains uncertain. Although image evaluations involved three independent radiologists, interreader variability indicated some subjectivity in image quality assessments. Additionally, the small sample size may limit statistical power, and the long-term clinical utility of radiomics in routine practice remains to be determined.
Despite these limitations, the study underscores the transformative potential of DL and SMS acceleration in musculoskeletal imaging. By enabling faster, higher-quality imaging, P4S2 contributes to more patient-centred radiology while advancing sustainability in medical imaging operations.
The combined use of DL-enhanced PAT and SMS imaging, specifically through the P4S2 protocol, marks a significant step forward in knee MRI. It enhances diagnostic accuracy, reduces scan times, lowers energy consumption and supports operational efficiency. These improvements position P4S2 as a promising candidate for routine clinical adoption in radiology departments focused on quality, efficiency and sustainability.
Source: Academic Radiology
Image Credit: iStock
