飞利浦

BACKGROUND:Cardiac Cine Magnetic Resonance Imaging (MRI) provides dynamic visualization of the heart's structure and function but is hindered by slow acquisition, requiring repeated breath-holds that challenge sick patients. Accelerated imaging can mitigate these issues but potentially reduce spatial and temporal resolution. Therefore, innovative approaches are essential to ensure effective performance under high acceleration conditions. Deep learning-based reconstruction methods show promise in enhancing image quality from highly undersampled data, accelerating scans while maintaining diagnostic accuracy. However, they often fail to effectively exploit the spatio-temporal features inherent to cine MRI, which are essential for accurate reconstruction, thereby leaving room for further improvement. PURPOSE:We aim to more effectively exploit the spatio-temporal features inherent in cine MRI sequences by integrating convolutional recurrent operations with a U-Net architecture, enhancing the reconstruction performance of cine MRI. METHODS:We developed a new deep learning model called CRUNet-MR that enhances the extraction of spatio-temporal features by combining convolutional recurrent operations with a U-Net structure. This design ensures continuous extraction of temporal features while fusing fine-grained spatial details with high-level semantic information. Furthermore, dilated convolutions are incorporated to expand the spatial receptive field, and appropriate combinations of dilation factors are explored to further enhance overall performance. RESULTS:Training, validation, and testing were performed on the public CMRxRecon2023 dataset, using two views and four acceleration factors ranging from 4 to 24 with the given Auto-Calibration Signal (ACS) area. The dataset consists of 120 subjects for training, 60 for validation, and 120 for testing. In general, the proposed CRUNet-MR shows statistically significant differences with benchmark models and consistently outperforms them, particularly showcasing better reconstruction quality in dynamic regions, highlighting its effective extraction of spatio-temporal features. Ablation studies further validated the design choices of CRUNet-MR. The model demonstrated strong reconstruction performance, achieving an average SSIM of 0.986 at an acceleration factor of 4 and 0.971 at a factor of 8 across both views. Furthermore, CRUNet-MR was validated on a small in-house LUMC dataset, showing its generalization capability and rapid adaptability through fine-tuning. CONCLUSIONS:The proposed CRUNet-MR model is well-suited for cine MRI reconstruction, effectively leveraging spatio-temporal features to reconstruct high-quality images, especially in dynamic cardiac regions. This capability highlights its potential to support higher acceleration factors, enabling faster and more patient-friendly cardiac imaging.

Since the various contrast-weighted MR images of a given anatomy contain redundant information, one contrast can be used to guide the reconstruction of another undersampled contrast acquired subsequently in the same session. To solve this reconstruction problem leveraging multi-contrast side information, several end-to-end learning-based guided reconstruction methods have been proposed. However, a key challenge is the requirement for large paired training datasets comprising raw k-space data and aligned reference images. We propose a modular plug-and-play approach, which requires no k-space training data and relies solely on partially paired image-domain datasets. In this approach, a content/style model of two-contrast MR data is first learned from a purely image-domain dataset and subsequently applied as a plug-and-play operator in iterative reconstruction. The disentanglement of content and style allows explicit representation of contrast-independent and contrast-specific factors. Consequently, incorporating prior information into the reconstruction reduces to a simple replacement operation on the aliased content of the estimated image using high-quality content derived from the reference scan. Combining this so-called content consistency operation with an MR data consistency step, followed by a corrective procedure for the content estimate, yields an iterative scheme. We name this novel approach PnP-CoSMo. This approach, by design, offers cross-contrast generalizability and provides an explanatory framework based on the shared and non-shared generative factors underlying the two given contrasts. We explore various aspects of PnP-CoSMo, including interpretability and convergence, via simulations. Furthermore, its practicality is demonstrated on the public NYU fastMRI DICOM dataset, showing equivalent or superior quality and greater generalizability compared to end-to-end methods. On two in-house multi-coil datasets, PnP-CoSMo enabled up to 32.6% greater acceleration over non-guided plug-and-play reconstruction at given SSIM.

Objectives Motion and limited compliance compromise diagnostic MR image quality, particularly in pediatric patients who frequently require sedation. Single-shot sequences offer a time-efficient alternative but suffer from reduced image quality. This study aimed to evaluate the diagnostic performance of a deep learning (DL) framework combining compressed sensing (CS) and convolutional neural networks (CNNs) to enhance T2-weighted single-shot MRI (T2-SSHDL) compared with conventional CS-based reconstruction (T2-SSHconv) and routinely acquired high-resolution T2-weighted sequences. Materials and methods This prospective single-center study included 62 pediatric patients (mean age, 7.4 +/- 4.9 years; 36 males, 26 females), who underwent T2-weighted single-shot brain MRI (29 sedated, 33 awake). Raw data were reconstructed using a DL-based pipeline and compared with conventional CS-based reconstructions. Quantitative metrics included apparent contrast-to-noise ratio (aCNR), apparent signal-to-noise ratio (aSNR), and edge rise distance (ERD). Two radiologists rated images for artifacts, sharpness, lesion conspicuity, and overall quality on a 5-point Likert scale. Results T2-SSHDL-sequences showed significantly higher aCNR (29.9 +/- 22.6 vs. 26.7 +/- 16.5; p < 0.001), aSNR (41.6 +/- 27.9 vs. 38.2 +/- 20.8; p = 0.003), and improved sharpness (ERD 0.90 +/- 0.35 mm vs. 1.35 +/- 0.42 mm; p < 0.001). Qualitative assessments confirmed superior image quality, lesion conspicuity, and sharpness (p < 0.001). Compared with high-resolution T2-weighted sequences, T2-SSHDL-sequences showed fewer motion artifacts and comparable lesion conspicuity in non-sedated patients. Conclusion DL-based reconstruction significantly enhances the diagnostic quality of T2-weighted single-shot brain MRI in pediatric patients, enabling clinically usable, ultrafast, motion-robust imaging with potential to reduce the need for sedation.

BACKGROUND:To compare the diagnostic quality of deep learning (DL) super-resolution reconstructed breath-hold (BH) and free-breathing (FB) single-shot (SSH) black-blood T2-weighted short tau inversion recovery (STIR) imaging with standard BH T2-STIR in cardiovascular magnetic resonance (CMR). METHODS:In this prospective study, short-axis BH and FB SSH T2-STIR were added to a standard cardiomyopathy CMR protocol at 1.5T, and DL super-resolution reconstruction was performed. Two readers evaluated diagnostic quality and certainty using a five-point Likert scale. Presence of focal edema was assessed on T2-weighted sequences, including standard T2-STIR and T2 mapping (both used for reference clinical assessment) as well as SSH T2 STIR and DL-SSH T2-STIR. Friedman test and one-way ANOVA were performed. RESULTS:Eighty-one participants (mean age: 54 ± 20 years; 50 men) were included. No difference was found in edema detection between reference assessment and DL-SSH T2-STIR (both 26%, (21/81 participants)). Scan time was reduced by 63% for BH and 86% for FB DL-SSH T2-STIR compared to standard T2-STIR (90 ± 6 s vs. 35 ± 3 s vs. 243 ± 16 s; p<.0001). BH and FB DL-SSH T2-STIR achieved lower artifact burden (5 [IQR, 4-5] vs. 4 [IQR, 4-5] vs. 4 [IQR, 3-5]; p<.0001), superior image contrast and sharpness compared to standard T2-STIR, especially in non-cooperative or arrhythmic participants. BH and FB DL-SSH T2-STIR imaging provided higher diagnostic certainty than standard T2-STIR (5 [IQR, 5-5] vs. 5 [IQR, 5-5] vs. 4 [IQR, 4-5]; p<.0001). Edema visibility was superior in BH DL-SSH compared to BH-SSH and standard T2-STIR (5 [IQR, 4.8-5] vs. 4 [IQR, 3.3-5] vs. 4 [IQR, 3-4.8]; p<.0001). Inter-rater agreement was substantial to excellent in the rating of edema visibility (BH DL-SSH T2-STIR, κ: 0.73 [95% CI: 0.44-1.0]; BH SSH T2-STIR, κ: 0.79 [95% CI: 0.66-0.97]; standard T2-STIR, κ: 0.86 [95% CI: 0.71-1.0]). Slice level-analysis showed that BH DL-SSH T2-STIR consistently provided superior image quality in apical slices compared to BH SSH and standard T2-STIR (4 [IQR, 4-5] vs. 4 [IQR, 4-4] vs. 4 [IQR, 3-4]; p<.0001). CONCLUSION:DL-SSH imaging enabled ultrafast T2-STIR acquisition and robust edema assessment in routine clinical CMR.

SUMMARY & CONCLUSIONSSoftware accounts for an ever-increasing fraction of overall system complexity. It therefore plays a critical role in the safety of medical devices, and managing their complexity is key to preventing system hazards, especially those arising from errors in software requirements specification, design, and implementation. Given its interaction with medical equipment, any software faults must be comprehensively included in risk analysis. In the healthcare technology arena, ensuring software reliability in medical devices is crucial due to its impact on patient and professional safety. Software Design Failure Mode and Effects Analysis (SW DFMEA) is a valuable process for improving software reliability, optimizing design, and ensuring safety in medical device hardware, which interfaces with increasingly complex software systems. While traditionally DFMEA focused on hardware, the complexity of modern software requires a focus on software-specific risks and failures, challenging the identification of nested failure modes. SW DFMEA complements risk management by systematically identifying and mitigating design related failures, thereby improving the safety and effectiveness of medical devices through proactive risk assessments. Integrating DFMEA with the Risk Management Matrix (RMM) strengthens strategic risk management and addresses potential hazards effectively. Despite their importance, SW DFMEA and risk management processes often receive less attention compared to hardware, due to regulatory and awareness gaps. This oversight poses challenges in complex software systems where identifying nested failures is difficult without systematic. Methods like SW DFMEA, which is underutilized, particularly in healthcare and Agile environments. This overview demonstrates the application of SW DFMEA and its integration with risk management to comply with medical software regulations such as FDA and EU MDR. A simplified case study illustrates practical applications in medical software development, emphasizing proactive maintenance, software updates, and adherence to industry regulations. When properly applied, SW DFMEA provides in-depth system understanding and implements proactive risk controls, significantly reducing safety risks in medical device environments. Proper scoping and execution foster robust, reliable, and safe software design, allowing for early identification and mitigation of potential failures during development. Adherence to standards and regulations through SW DFMEA and risk management integration enhances the safety and reliability of medical device software.