Assessment of Deep Learning-Based Triage Application for Acute Ischemic Stroke on Brain MRI in the ER.

RATIONALE AND OBJECTIVES: To assess a deep learning application (DLA) for acute ischemic stroke (AIS) detection on brain magnetic resonance imaging (MRI) in the emergency room (ER) and the effect of T2-weighted imaging (T2WI) on its performance. MATERIALS AND METHODS: We retrospectively analyzed brain MRIs taken through the ER from March to October 2021 that included diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) sequences. MRIs were processed by the DLA, and sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUROC) were evaluated, with three neuroradiologists establishing the gold standard for detection performance. In addition, we examined the impact of axial T2WI, when available, on the accuracy and processing time of DLA. RESULTS: The study included 947 individuals (mean age ± standard deviation, 64 years ± 16; 461 men, 486 women), with 239 (25%) positive for AIS. The overall performance of DLA was as follows: sensitivity, 90%; specificity, 89%; accuracy, 89%; and AUROC, 0.95. The average processing time was 24 s. In the subgroup with T2WI, T2WI did not significantly impact MRI assessments but did result in longer processing times (35 s without T2WI compared to 48 s with T2WI, p < 0.001). CONCLUSION: The DLA successfully identified AIS in the ER setting with an average processing time of 24 s. The absence of performance acquire with axial T2WI suggests that the DLA can diagnose AIS with just axial DWI and FLAIR sequences, potentially shortening the exam duration in the ER.

Generation of a TMEM43 knockout human induced pluripotent stem cell line (HDZi003-A-1) using CRISPR/Cas9.

TMEM43 (LUMA) is a ubiquitously expressed protein with unknown function. The protein is phylogenetically highly conserved and also found in Drosophila melanogaster (Klinke et al., 2022). TMEM43-p.S358L is a rare, fully penetrant mutation that leads to arrhythmogenic right ventricular cardiomyopathy type 5 (ARVC5). To understand the function of the ARVC5-associated mutation it is first important to understand the function of the TMEM43 protein. Therefore, a TMEM43 knockout induced pluripotent stem cell (iPSC) line was generated using the CRISPR/Cas9 genome editing system. The resulting cell line had a deficiency of TMEM43 and showed normal morphology and a stable karyotype. The colonies were positive for pluripotency markers and could be differentiated into the three germ layers.

Differentiation and function of cardiac valves in the adult Drosophila heart.

Drosophila, like all insects, has an open circulatory system for the distribution of haemolymph and its components. The circulation of the haemolymph is essentially driven by the pumping activity of the linear heart. The heart is constructed as a tube into which the haemolymph is sucked and pumped forward by rhythmic contractions running from the posterior to the anterior, where it leaves the heart tube. The heart harbours cardiac valves to regulate flow directionality, with a single heart valve differentiating during larval development to separate the heart tube into two chambers. During metamorphosis, the heart is partially restructured, with the linear heart tube with one terminal wide-lumen heart chamber being converted into a linear four-chambered heart tube with three valves. As in all metazoan circulatory systems, the cardiac valves play an essential role in regulating the direction of blood flow. We provide evidence that the valves in adult flies arise via transdifferentiation, converting lumen-forming contractile cardiomyocytes into differently structured valve cells. Interestingly, adult cardiac valves exhibit a similar morphology to their larval counterparts, but act differently upon heart beating. Applying calcium imaging in living specimens to analyse activity in valve cells, we show that adult cardiac valves operate owing to muscle contraction. However, valve cell shape dynamics are altered compared with larval valves, which led us to propose our current model of the opening and closing mechanisms in the fly heart.

Neprilysin 4: an essential peptidase with multifaceted physiological relevance.

Neprilysins are highly conserved ectoenzymes that hydrolyze and thus inactivate signaling peptides in the extracellular space. Herein, we focus on Neprilysin 4 from Drosophila melanogaster and evaluate the existing knowledge on the physiological relevance of the peptidase. Particular attention is paid to the role of the neprilysin in regulating feeding behavior and the expression of insulin-like peptides in the central nervous system. In addition, we assess the function of the peptidase in controlling the activity of the sarcoplasmic and endoplasmic reticulum Ca2+ ATPase in myocytes, as well as the underlying molecular mechanism in detail.

Deep learning-based motion quantification from k-space for fast model-based magnetic resonance imaging motion correction.

BACKGROUND: Intra-scan rigid-body motion is a costly and ubiquitous problem in clinical magnetic resonance imaging (MRI) of the head. PURPOSE: State-of-the-art methods for retrospective motion correction in MRI are often computationally expensive or in the case of image-to-image deep learning (DL) based methods can be prone to undesired alterations of the image (hallucinations'). In this work we introduce a novel rigid-body motion correction method which combines the advantages of classical model-driven and data-consistency (DC) preserving approaches with a novel DL algorithm, to provide fast and robust retrospective motion correction. METHODS: The proposed Motion Parameter Estimating Densenet (MoPED) retrospectively estimates subject head motion during MRI acquisitions using a DL network with DenseBlocks and multitask learning. It quantifies the 2D rigid in-plane motion parameters slice-wise for each echo train (ET) of a Cartesian T2-weighted 2D Turbo-Spin-Echo sequence. The network receives a center patch of the motion corrupted k-space as well as an additional motion-free low-resolution reference scan to provide the ground truth orientation. The supervised training utilizes motion simulations based on 28 acquisitions with subject-wise training, validation, and test data splits of 70%, 23%, and 7%. During inference, MoPED is embedded in an iterative DC-driven motion correction algorithm which alternatingly updates estimates of the motion parameters and motion-corrected low-resolution k-space data. The estimated motion parameters are then used to reconstruct the final motion corrected image. The mean absolute/squared error and the Pearson correlation coefficient were used to analyze the motion parameter estimation quality on in-silico data in a quantitative evaluation. Structural similarity (SSIM), DC error and root mean squared error (RMSE) were used as metrics of image quality improvement. Furthermore, the generalization capability of the network was analyzed on two in-vivo motion volumes with 28 slices each and on one simulated T1-weighted volume. RESULTS: The motion estimation achieves a Pearson correlation of 0.968 to the simulated ground-truth of the 2433 test data slices used. In-silico results indicate that MoPED decreases the time for the optimization by a factor of around 27 compared to a conventional method and is able to reduce the RMSE of the reconstructions and average DC error by more than a factor of two compared to uncorrected images. In-vivo experiments show a decrease in computation time by a factor of around 20, a RMSE decrease from 0.055 to 0.033 and an SSIM increase from 0.795 to 0.862. Furthermore, contrast independence is demonstrated as MoPED is also able to correct T1-weighted images in simulations without retraining. Due to the model-based correction, no hallucinations were observed. CONCLUSIONS: Incorporating DL in a model-based motion correction algorithm shows great benefit on the optimization and computation time. The k-space-based estimation also allows a data consistent correction and therefore avoids the risk of hallucinations of image-to-image approaches.

Formation and function of a highly specialised type of organelle in cardiac valve cells.

Within a cell, vesicles play a crucial role in the transport of membrane material and proteins to a given target membrane, and thus regulate a variety of cellular functions. Vesicular transport occurs by means of, among others, endocytosis, where cargoes are taken up by the cell and are processed further upon vesicular trafficking, i.e. transported back to the plasma membrane via recycling endosomes or the degraded by fusion of the vesicles with lysosomes. During evolution, a variety of vesicles with individual functions arose, with some of them building up highly specialised subcellular compartments. In this study, we have analysed the biosynthesis of a new vesicular compartment present in the valve cells of Drosophila melanogaster. We show that the compartment is formed by invaginations of the plasma membrane and grows via re-routing of the recycling endosomal pathway. This is achieved by inactivation of other membrane-consuming pathways and a plasma membrane-like molecular signature of the compartment in these highly specialised heart cells.

Neprilysins regulate muscle contraction and heart function via cleavage of SERCA-inhibitory micropeptides.

Muscle contraction depends on strictly controlled Ca2+ transients within myocytes. A major player maintaining these transients is the sarcoplasmic/endoplasmic reticulum Ca2+ ATPase, SERCA. Activity of SERCA is regulated by binding of micropeptides and impaired expression or function of these peptides results in cardiomyopathy. To date, it is not known how homeostasis or turnover of the micropeptides is regulated. Herein, we find that the Drosophila endopeptidase Neprilysin 4 hydrolyzes SERCA-inhibitory Sarcolamban peptides in membranes of the sarcoplasmic reticulum, thereby ensuring proper regulation of SERCA. Cleavage is necessary and sufficient to maintain homeostasis and function of the micropeptides. Analyses on human Neprilysin, sarcolipin, and ventricular cardiomyocytes indicates that the regulatory mechanism is evolutionarily conserved. By identifying a neprilysin as essential regulator of SERCA activity and Ca2+ homeostasis in cardiomyocytes, these data contribute to a more comprehensive understanding of the complex mechanisms that control muscle contraction and heart function in health and disease.

A Drosophila melanogaster model for TMEM43-related arrhythmogenic right ventricular cardiomyopathy type 5.

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is a severe cardiac disease that leads to heart failure or sudden cardiac death (SCD). For the pathogenesis of ARVC, various mutations in at least eight different genes have been identified. A rare form of ARVC is associated with the mutation TMEM43 p.S358L, which is a fully penetrant variant in male carriers. TMEM43 p.S358 is homologous to CG8111 p.S333 in Drosophila melanogaster. We established CRISPR/Cas9-mediated CG8111 knock-out mutants in Drosophila, as well as transgenic fly lines carrying an overexpression construct of the CG8111 p.S333L substitution. Knock-out flies developed normally, whereas the overexpression of CG8111 p.S333L caused growth defects, loss of body weight, cardiac arrhythmias, and premature death. An evaluation of a series of model mutants that replaced S333 by selected amino acids proved that the conserved serine is critical for the physiological function of CG8111. Metabolomic and proteomic analyses revealed that the S333 in CG8111 is essential to proper energy homeostasis and lipid metabolism in the fly. Of note, metabolic impairments were also found in the murine Tmem43 disease model, and fibrofatty replacement is a hallmark of human ARVC5. These findings contribute to a more comprehensive understanding of the molecular functions of CG8111 in Drosophila, and can represent a valuable basis to assess the aetiology of the human TMEM43 p.S358L variant in more detail.

Interplay between SERCA, 4E-BP, and eIF4E in the Drosophila heart.

Appropriate cardiac performance depends on a tightly controlled handling of Ca2+ in a broad range of species, from invertebrates to mammals. The role of the Ca2+ ATPase, SERCA, in Ca2+ handling is pivotal, and its activity is regulated, inter alia, by interacting with distinct proteins. Herein, we give evidence that 4E binding protein (4E-BP) is a novel regulator of SERCA activity in Drosophila melanogaster during cardiac function. Flies over-expressing 4E-BP showed improved cardiac performance in young individuals associated with incremented SERCA activity. Moreover, we demonstrate that SERCA interacts with translation initiation factors eIF4E-1, eIF4E-2 and eIF4E-4 in a yeast two-hybrid assay. The specific identification of eIF4E-4 in cardiac tissue leads us to propose that the interaction of elF4E-4 with SERCA may be the basis of the cardiac effects observed in 4E-BP over-expressing flies associated with incremented SERCA activity.

A parallel spatial and Bloch manifold regularized iterative reconstruction method for MR Fingerprinting.

Magnetic Resonance Fingerprinting (MRF) reconstructs tissue maps based on a sequence of very highly undersampled images. In order to be able to perform MRF reconstruction, state-of-the-art MRF methods rely on priors such as the MR physics (Bloch equations) and might also use some additional low-rank or spatial regularization. However to our knowledge these three regularizations are not applied together in a joint reconstruction. The reason is that it is indeed challenging to incorporate effectively multiple regularizations in a single MRF optimization algorithm. As a result most of these methods are not robust to noise especially when the sequence length is short. In this paper, we propose a family of new methods where spatial and low-rank regularizations, in addition to the Bloch manifold regularization, are applied on the images. We show on digital phantom and NIST phantom scans, as well as volunteer scans that the proposed methods bring significant improvement in the quality of the estimated tissue maps.

Identification and bioinformatic analysis of neprilysin and neprilysin-like metalloendopeptidases in Drosophila melanogaster.

The neprilysin (M13) family of metalloendopeptidases comprises highly conserved ectoenzymes that cleave and thereby inactivate many physiologically relevant peptides in the extracellular space. Impaired neprilysin activity is associated with numerous human diseases. Here, we present a comprehensive list and classification of M13 family members in Drosophila melanogaster. Seven Neprilysin (Nep) genes encode active peptidases, while 21 Neprilysin-like (Nepl) genes encode proteins predicted to be catalytically inactive. RNAseq data demonstrate that all 28 genes are expressed during development, often in a tissue-specific pattern. Most Nep proteins possess a transmembrane domain, whereas almost all Nepl proteins are predicted to be secreted.

Automated detection of critical findings in multi-parametric brain MRI using a system of 3D neural networks.

With the rapid growth and increasing use of brain MRI, there is an interest in automated image classification to aid human interpretation and improve workflow. We aimed to train a deep convolutional neural network and assess its performance in identifying abnormal brain MRIs and critical intracranial findings including acute infarction, acute hemorrhage and mass effect. A total of 13,215 clinical brain MRI studies were categorized to training (74%), validation (9%), internal testing (8%) and external testing (8%) datasets. Up to eight contrasts were included from each brain MRI and each image volume was reformatted to common resolution to accommodate for differences between scanners. Following reviewing the radiology reports, three neuroradiologists assigned each study to abnormal vs normal, and identified three critical findings including acute infarction, acute hemorrhage, and mass effect. A deep convolutional neural network was constructed by a combination of localization feature extraction (LFE) modules and global classifiers to identify the presence of 4 variables in brain MRIs including abnormal, acute infarction, acute hemorrhage and mass effect. Training, validation and testing sets were randomly defined on a patient basis. Training was performed on 9845 studies using balanced sampling to address class imbalance. Receiver operating characteristic (ROC) analysis was performed. The ROC analysis of our models for 1050 studies within our internal test data showed AUC/sensitivity/specificity of 0.91/83%/86% for normal versus abnormal brain MRI, 0.95/92%/88% for acute infarction, 0.90/89%/81% for acute hemorrhage, and 0.93/93%/85% for mass effect. For 1072 studies within our external test data, it showed AUC/sensitivity/specificity of 0.88/80%/80% for normal versus abnormal brain MRI, 0.97/90%/97% for acute infarction, 0.83/72%/88% for acute hemorrhage, and 0.87/79%/81% for mass effect. Our proposed deep convolutional network can accurately identify abnormal and critical intracranial findings on individual brain MRIs, while addressing the fact that some MR contrasts might not be available in individual studies.

Human heart-forming organoids recapitulate early heart and foregut development.

Organoid models of early tissue development have been produced for the intestine, brain, kidney and other organs, but similar approaches for the heart have been lacking. Here we generate complex, highly structured, three-dimensional heart-forming organoids (HFOs) by embedding human pluripotent stem cell aggregates in Matrigel followed by directed cardiac differentiation via biphasic WNT pathway modulation with small molecules. HFOs are composed of a myocardial layer lined by endocardial-like cells and surrounded by septum-transversum-like anlagen; they further contain spatially and molecularly distinct anterior versus posterior foregut endoderm tissues and a vascular network. The architecture of HFOs closely resembles aspects of early native heart anlagen before heart tube formation, which is known to require an interplay with foregut endoderm development. We apply HFOs to study genetic defects in vitro by demonstrating that NKX2.5-knockout HFOs show a phenotype reminiscent of cardiac malformations previously observed in transgenic mice.

Reproducibility of an Automated Quantitative MRI Assessment of Low-Grade Knee Articular Cartilage Lesions.

OBJECTIVE: The goal of this study was to assess the reproducibility of an automated knee cartilage segmentation of 21 cartilage regions with a model-based algorithm and to compare the results with manual segmentation. DESIGN: Thirteen patients with low-grade femoral cartilage defects were included in the study and were scanned twice on a 7-T magnetic resonance imaging (MRI) scanner 8 days apart. A 3-dimensional double-echo steady-state (3D-DESS) sequence was used to acquire MR images for automated cartilage segmentation, and T2-mapping was performed using a 3D triple-echo steady-state (3D-TESS) sequence. Cartilage volume, thickness, and T2 and texture features were automatically extracted from each knee for each of the 21 subregions. DESS was used for manual cartilage segmentation and compared with automated segmentation using the Dice coefficient. The reproducibility of each variable was expressed using standard error of measurement (SEM) and smallest detectable change (SDC). RESULTS: The Dice coefficient for the similarity between manual and automated segmentation ranged from 0.83 to 0.88 in different cartilage regions. Test-retest analysis of automated cartilage segmentation and automated quantitative parameter extraction revealed excellent reproducibility for volume measurement (mean SDC for all subregions of 85.6 mm3), for thickness detection (SDC = 0.16 mm) and also for T2 values (SDC = 2.38 ms) and most gray-level co-occurrence matrix features (SDC = 0.1 a.u.). CONCLUSIONS: The proposed technique of automated knee cartilage evaluation based on the segmentation of 3D MR images and correlation with T2 mapping provides highly reproducible results and significantly reduces the segmentation effort required for the analysis of knee articular cartilage in longitudinal studies.

A trimeric metazoan Rab7 GEF complex is crucial for endocytosis and scavenger function.

Endosome biogenesis in eukaryotic cells is critical for nutrient uptake and plasma membrane integrity. Early endosomes initially contain Rab5, which is replaced by Rab7 on late endosomes prior to their fusion with lysosomes. Recruitment of Rab7 to endosomes requires the Mon1-Ccz1 guanine-nucleotide-exchange factor (GEF). Here, we show that full function of the Drosophila Mon1-Ccz1 complex requires a third stoichiometric subunit, termed Bulli (encoded by CG8270). Bulli localises to Rab7-positive endosomes, in agreement with its function in the GEF complex. Using Drosophila nephrocytes as a model system, we observe that absence of Bulli results in (i) reduced endocytosis, (ii) Rab5 accumulation within non-acidified enlarged endosomes, (iii) defective Rab7 localisation and (iv) impaired endosomal maturation. Moreover, longevity of animals lacking bulli is affected. Both the Mon1-Ccz1 dimer and a Bulli-containing trimer display Rab7 GEF activity. In summary, this suggests a key role for Bulli in the Rab5 to Rab7 transition during endosomal maturation rather than a direct influence on the GEF activity of Mon1-Ccz1.

The septate junction protein Tetraspanin 2A is critical to the structure and function of Malpighian tubules in Drosophila melanogaster.

Tetraspanin-2A (Tsp2A) is an integral membrane protein of smooth septate junctions in Drosophila melanogaster. To elucidate its structural and functional roles in Malpighian tubules, we used the c42-GAL4/UAS system to selectively knock down Tsp2A in principal cells of the tubule. Tsp2A localizes to smooth septate junctions (sSJ) in Malpighian tubules in a complex shared with partner proteins Snakeskin (Ssk), Mesh, and Discs large (Dlg). Knockdown of Tsp2A led to the intracellular retention of Tsp2A, Ssk, Mesh, and Dlg, gaps and widening spaces in remaining sSJ, and tumorous and cystic tubules. Elevated protein levels together with diminished V-type H+-ATPase activity in Tsp2A knockdown tubules are consistent with cell proliferation and reduced transport activity. Indeed, Malpighian tubules isolated from Tsp2A knockdown flies failed to secrete fluid in vitro. The absence of significant transepithelial voltages and resistances manifests an extremely leaky epithelium that allows secreted solutes and water to leak back to the peritubular side. The tubular failure to excrete fluid leads to extracellular volume expansion in the fly and to death within the first week of adult life. Expression of the c42-GAL4 driver begins in Malpighian tubules in the late embryo and progresses upstream to distal tubules in third instar larvae, which can explain why larvae survive Tsp2A knockdown and adults do not. Uncontrolled cell proliferation upon Tsp2A knockdown confirms the role of Tsp2A as tumor suppressor in addition to its role in sSJ structure and transepithelial transport.

Scanning laser optical tomography resolves developmental neurotoxic effects on pioneer neurons.

Developmental neurotoxic compounds impair the developing human nervous system at lower doses than those affecting adults. Standardized test methods for assessing developmental neurotoxicity (DNT) require the use of high numbers of laboratory animals. Here, we use a novel assay that is based on the development of an intact insect embryo in serum-free culture. Neural pathways in the leg of embryonic locusts are established by a pair of afferent pioneer neurons, extending axons along a well-defined pathway to the central nervous system. After exposure to test chemicals, we analyze pioneer neuron shape with conventional fluorescence microscopy and compare it to 3D images, obtained by scanning laser optical tomography (SLOT) and processed by a segmentation algorithm. The segmented SLOT images resolve the 3D structure of the pioneers, recognize pathfinding defects and are thus advantageous for detecting DNT-positive compounds. The defects in axon elongation and pathfinding of pioneer axons caused by two DNT-positive reference compounds (methylmercury chloride; sodium(meta)arsenite) are compared to the biochemically measured general viability of the embryo. Using conventional fluorescence microscopy to establish concentration-response curves of axon elongation, we show that this assay identifies methylmercury chloride and the pro-apoptotic compound staurosporine as developmental neurotoxicants.

The septate junction protein Mesh is required for epithelial morphogenesis, ion transport, and paracellular permeability in the Drosophila Malpighian tubule.

Septate junctions (SJs) are occluding cell-cell junctions that have roles in paracellular permeability and barrier function in the epithelia of invertebrates. Arthropods have two types of SJs, pleated SJs and smooth SJs (sSJs). In Drosophila melanogaster, sSJs are found in the midgut and Malpighian tubules, but the functions of sSJs and their protein components in the tubule epithelium are unknown. Here we examined the role of the previously identified integral sSJ component, Mesh, in the Malpighian tubule. We genetically manipulated mesh specifically in the principal cells of the tubule at different life stages. Tubules of flies with developmental mesh knockdown revealed defects in epithelial architecture, sSJ molecular and structural organization, and lack of urine production in basal and kinin-stimulated conditions, resulting in edema and early adult lethality. Knockdown of mesh during adulthood did not disrupt tubule epithelial and sSJ integrity but decreased the transepithelial potential, diminished transepithelial fluid and ion transport, and decreased paracellular permeability to 4-kDa dextran. Drosophila kinin decreased transepithelial potential and increased chloride permeability, and it stimulated fluid secretion in both control and adult mesh knockdown tubules but had no effect on 4-kDa dextran flux. Together, these data indicate roles for Mesh in the developmental maturation of the Drosophila Malpighian tubule and in ion and macromolecular transport in the adult tubule.