Post-hoc standardisation of parametric T1 maps in cardiovascular magnetic resonance imaging: a proof-of-concept.

BACKGROUND: In cardiovascular magnetic resonance imaging parametric T1 mapping lacks universally valid reference values. This limits its extensive use in the clinical routine. The aim of this work was the introduction of our self-developed Magnetic Resonance Imaging Software for Standardization (MARISSA) as a post-hoc standardisation approach. METHODS: Our standardisation approach minimises the bias of confounding parameters (CPs) on the base of regression models. 214 healthy subjects with 814 parametric T1 maps were used for training those models on the CPs: age, gender, scanner and sequence. The training dataset included both sex, eleven different scanners and eight different sequences. The regression model type and four other adjustable standardisation parameters were optimised among 240 tested settings to achieve the lowest coefficient of variation, as measure for the inter-subject variability, in the mean T1 value across the healthy test datasets (HTE, N = 40, 156 T1 maps). The HTE were then compared to 135 patients with left ventricular hypertrophy including hypertrophic cardiomyopathy (HCM, N = 112, 121 T1 maps) and amyloidosis (AMY, N = 24, 24 T1 maps) after applying the best performing standardisation pipeline (BPSP) to evaluate the diagnostic accuracy. FINDINGS: The BPSP reduced the COV of the HTE from 12.47% to 5.81%. Sensitivity and specificity reached 95.83% / 91.67% between HTE and AMY, 71.90% / 72.44% between HTE and HCM, and 87.50% / 98.35% between HCM and AMY. INTERPRETATION: Regarding the BPSP, MARISSA enabled the comparability of T1 maps independently of CPs while keeping the discrimination of healthy and patient groups as found in literature. FUNDING: This study was supported by the BMBF / DZHK.

Lazy Luna: Extendible software for multilevel reader comparison in cardiovascular magnetic resonance imaging.

BACKGROUND AND OBJECTIVES: Cardiovascular Magnetic Resonance (CMR) imaging is a growing field with increasing diagnostic utility in clinical routine. Quantitative diagnostic parameters are typically calculated based on contours or points provided by readers, e.g. natural intelligences (NI) such as clinicians or researchers, and artificial intelligences (AI). As clinical applications multiply, evaluating the precision and reproducibility of quantitative parameters becomes increasingly important. Although segmentation challenges for AIs and guidelines for clinicians provide quality assessments and regulation, the methods ought to be combined and streamlined for clinical applications. The goal of the developed software, Lazy Luna (LL), is to offer a flexible evaluation tool that is readily extendible to new sequences and scientific endeavours. METHODS: An interface was designed for LL, which allows for comparing annotated CMR images. Geometric objects ensure precise calculations of metric values and clinical results regardless of whether annotations originate from AIs or NIs. A graphical user interface (GUI) is provided to make the software available to non-programmers. The GUI allows for an interactive inspection of image datasets as well as implementing tracing procedures, which follow statistical reader differences in clinical results to their origins in individual image contours. The backend software builds on a set of meta-classes, which can be extended to new imaging sequences and clinical parameters. Following an agile development procedure with clinical feedback allows for a quick implementation of new classes, figures and tables for evaluation. RESULTS: Two application cases present LL's extendibility to clinical evaluation and AI development contexts. The first concerns T1 parametric mapping images segmented by two expert readers. Quantitative result differences are traced to reveal typical segmentation dissimilarities from which these differences originate. The meta-classes are extended to this new application scenario. The second applies to the open source Late Gadolinium Enhancement (LGE) quantification challenge for AI developers "Emidec", which illustrates LL's usability as open source software. CONCLUSION: The presented software Lazy Luna allows for an automated multilevel comparison of readers as well as identifying qualitative reasons for statistical reader differences. The open source software LL can be extended to new application cases in the future.

Multilevel comparison of deep learning models for function quantification in cardiovascular magnetic resonance: On the redundancy of architectural variations.

Background: Cardiac function quantification in cardiovascular magnetic resonance requires precise contouring of the heart chambers. This time-consuming task is increasingly being addressed by a plethora of ever more complex deep learning methods. However, only a small fraction of these have made their way from academia into clinical practice. In the quality assessment and control of medical artificial intelligence, the opaque reasoning and associated distinctive errors of neural networks meet an extraordinarily low tolerance for failure. Aim: The aim of this study is a multilevel analysis and comparison of the performance of three popular convolutional neural network (CNN) models for cardiac function quantification. Methods: U-Net, FCN, and MultiResUNet were trained for the segmentation of the left and right ventricles on short-axis cine images of 119 patients from clinical routine. The training pipeline and hyperparameters were kept constant to isolate the influence of network architecture. CNN performance was evaluated against expert segmentations for 29 test cases on contour level and in terms of quantitative clinical parameters. Multilevel analysis included breakdown of results by slice position, as well as visualization of segmentation deviations and linkage of volume differences to segmentation metrics via correlation plots for qualitative analysis. Results: All models showed strong correlation to the expert with respect to quantitative clinical parameters (rz ' = 0.978, 0.977, 0.978 for U-Net, FCN, MultiResUNet respectively). The MultiResUNet significantly underestimated ventricular volumes and left ventricular myocardial mass. Segmentation difficulties and failures clustered in basal and apical slices for all CNNs, with the largest volume differences in the basal slices (mean absolute error per slice: 4.2 ± 4.5 ml for basal, 0.9 ± 1.3 ml for midventricular, 0.9 ± 0.9 ml for apical slices). Results for the right ventricle had higher variance and more outliers compared to the left ventricle. Intraclass correlation for clinical parameters was excellent (≥0.91) among the CNNs. Conclusion: Modifications to CNN architecture were not critical to the quality of error for our dataset. Despite good overall agreement with the expert, errors accumulated in basal and apical slices for all models.

Introduction of a cascaded segmentation pipeline for parametric T1 mapping in cardiovascular magnetic resonance to improve segmentation performance.

The manual and often time-consuming segmentation of the myocardium in cardiovascular magnetic resonance is increasingly automated using convolutional neural networks (CNNs). This study proposes a cascaded segmentation (CASEG) approach to improve automatic image segmentation quality. First, an object detection algorithm predicts a bounding box (BB) for the left ventricular myocardium whose 1.5 times enlargement defines the region of interest (ROI). Then, the ROI image section is fed into a U-Net based segmentation. Two CASEG variants were evaluated: one using the ROI cropped image solely (cropU) and the other using a 2-channel-image additionally containing the original BB image section (crinU). Both were compared to a classical U-Net segmentation (refU). All networks share the same hyperparameters and were tested on basal and midventricular slices of native and contrast enhanced (CE) MOLLI T1 maps. Dice Similarity Coefficient improved significantly (p < 0.05) in cropU and crinU compared to refU (81.06%, 81.22%, 72.79% for native and 80.70%, 79.18%, 71.41% for CE data), while no significant improvement (p < 0.05) was achieved in the mean absolute error of the T1 time (11.94 ms, 12.45 ms, 14.22 ms for native and 5.32 ms, 6.07 ms, 5.89 ms for CE data). In conclusion, CASEG provides an improved geometric concordance but needs further improvement in the quantitative outcome.

Fast acquisition of left and right ventricular function parameters applying cardiovascular magnetic resonance in clinical routine - validation of a 2-shot compressed sensing cine sequence.

Objectives. To evaluate if cine sequences accelerated by compressed sensing (CS) are feasible in clinical routine and yield equivalent cardiac morphology in less time. Design. We evaluated 155 consecutive patients with various cardiac diseases scanned during our clinical routine. LV and RV short axis (SAX) cine images were acquired by conventional and prototype 2-shot CS sequences on a 1.5 T CMR. The 2-shot prototype captures the entire heart over a period of 3 beats making the acquisition potentially even faster. Both scans were performed with identical slice parameters and positions. We compared LV and RV morphology with Bland-Altmann plots and weighted the results in relation to pre-defined tolerance intervals. Subjective and objective image quality was evaluated using a 4-point score and adapted standardized criteria. Scan times were evaluated for each sequence. Results. In total, no acquisitions were lost due to non-diagnostic image quality in the subjective image score. Objective image quality analysis showed no statistically significant differences. The scan time of the CS cines was significantly shorter (p < .001) with mean scan times of 178 ± 36 s compared to 313 ± 65 s for the conventional cine. All cardiac function parameters showed excellent correlation (r 0.978-0.996). Both sequences were considered equivalent for the assessment of LV and RV morphology. Conclusions. The 2-shot CS SAX cines can be used in clinical routine to acquire cardiac morphology in less time compared to the conventional method, with no total loss of acquisitions due to nondiagnostic quality. TRIAL REGISTRATION: ISRCTN12344380. Registered 20 November 2020, retrospectively registered.

Introduction of Lazy Luna an automatic software-driven multilevel comparison of ventricular function quantification in cardiovascular magnetic resonance imaging.

Cardiovascular magnetic resonance imaging is the gold standard for cardiac function assessment. Quantification of clinical results (CR) requires precise segmentation. Clinicians statistically compare CRs to ensure reproducibility. Convolutional Neural Network developers compare their results via metrics. Aim: Introducing software capable of automatic multilevel comparison. A multilevel analysis covering segmentations and CRs builds on a generic software backend. Metrics and CRs are calculated with geometric accuracy. Segmentations and CRs are connected to track errors and their effects. An interactive GUI makes the software accessible to different users. The software's multilevel comparison was tested on a use case based on cardiac function assessment. The software shows good reader agreement in CRs and segmentation metrics (Dice > 90%). Decomposing differences by cardiac position revealed excellent agreement in midventricular slices: > 90% but poorer segmentations in apical (> 71%) and basal slices (> 74%). Further decomposition by contour type locates the largest millilitre differences in the basal right cavity (> 3 ml). Visual inspection shows these differences being caused by different basal slice choices. The software illuminated reader differences on several levels. Producing spreadsheets and figures concerning metric values and CR differences was automated. A multilevel reader comparison is feasible and extendable to other cardiac structures in the future.

An agenda-setting paper on data sharing platforms: euCanSHare workshop.

Various data sharing platforms are being developed to enhance the sharing of cohort data by addressing the fragmented state of data storage and access systems. However, policy challenges in several domains remain unresolved. The euCanSHare workshop was organized to identify and discuss these challenges and to set the future research agenda. Concerns over the multiplicity and long-term sustainability of platforms, lack of resources, access of commercial parties to medical data, credit and recognition mechanisms in academia and the organization of data access committees are outlined. Within these areas, solutions need to be devised to ensure an optimal functioning of platforms.