Improved Resolution and Image Quality of Musculoskeletal Magnetic Resonance Imaging using Deep Learning-based Denoising Reconstruction: A Prospective Clinical Study.

OBJECTIVE: To prospectively evaluate a deep learning-based denoising reconstruction (DLR) for improved resolution and image quality in musculoskeletal (MSK) magnetic resonance imaging (MRI). METHODS: Images from 137 contrast-weighted sequences in 40 MSK patients were evaluated. Each sequence was performed twice, first with the routine parameters and reconstructed with a routine reconstruction filter (REF), then with higher resolution and reconstructed with DLR, and with three conventional reconstruction filters (NL2, GA43, GA53). The five reconstructions (REF, DLR, NL2, GA43, and GA53) were de-identified, randomized, and blindly reviewed by three MSK radiologists using eight scoring criteria and a forced ranking. Quantitative SNR, CNR, and structure's full width at half maximum (FWHM) for resolution assessment were measured and compared. To account for repeated measures, Generalized Estimating Equations (GEE) with Bonferroni adjustment was used to compare the reader's scores, SNR, CNR, and FWHM between DLR vs. NL2, GA43, GA53, and REF. RESULTS: Compared to the routine REF images, the resolution was improved by 47.61% with DLR from 0.39 ± 0.15 mm2 to 0.20 ± 0.06 mm2 (p < 0.001). Per-sequence average scan time was shortened by 7.93% with DLR from 165.58 ± 21.86 s to 152.45 ± 25.65 s (p < 0.001). Based on the average scores, DLR images were rated significantly higher in all image quality criteria and the forced ranking (p < 0.001). CONCLUSION: This prospective clinical evaluation demonstrated that DLR allows approximately two times finer resolution and improved image quality compared to the standard-of-care images.

Transforming medical education to strengthen the health professional training in Viet Nam: A case study.

The competency-based undergraduate curriculum reform at the University of Medicine and Pharmacy at Ho Chi Minh City, Faculty of Medicine (UMP-FM) is detailed and reviewed in reference to the instructional and institutional reforms, and enabling actions recommended by the Lancet 2010 Commission for Health Professional Education. Key objectives are to: revise the overall 6-year curriculum to be more integrated and competency-based; reinforce students' knowledge application, problem-solving, clinical competence, self-directed learning and soft skills; develop a comprehensive and performance-based student assessment programme; and establish a comprehensive quality monitoring programme to facilitate changes and improvements. New features include early introduction to the practice of medicine, family- and community-based medicine, professionalism, interprofessional education, electives experiences, and a scholarly project. Institutional reform introduces a faculty development programme, joint planning mechanism, a "culture of critical inquiry", and a transparent faculty reward system. Lessons learnt from the curriculum reform at UMP-FM could be helpful to medical schools from low- and middle-income countries considering transitioning from a traditional to a competency-based curriculum. Funding: This work receives no external funding.

Accuracy, uncertainty, and adaptability of automatic myocardial ASL segmentation using deep CNN.

PURPOSE: To apply deep convolution neural network to the segmentation task in myocardial arterial spin labeled perfusion imaging and to develop methods that measure uncertainty and that adapt the convolution neural network model to a specific false-positive versus false-negative tradeoff. METHODS: The Monte Carlo dropout U-Net was trained on data from 22 subjects and tested on data from 6 heart transplant recipients. Manual segmentation and regional myocardial blood flow were available for comparison. We consider 2 global uncertainty measures, named "Dice uncertainty" and "Monte Carlo dropout uncertainty," which were calculated with and without the use of manual segmentation, respectively. Tversky loss function with a hyperparameter ß was used to adapt the model to a specific false-positive versus false-negative tradeoff. RESULTS: The Monte Carlo dropout U-Net achieved a Dice coefficient of 0.91 ± 0.04 on the test set. Myocardial blood flow measured using automatic segmentations was highly correlated to that measured using the manual segmentation (R2 = 0.96). Dice uncertainty and Monte Carlo dropout uncertainty were in good agreement (R2 = 0.64). As ß increased, the false-positive rate systematically decreased and false-negative rate systematically increased. CONCLUSION: We demonstrate the feasibility of deep convolution neural network for automatic segmentation of myocardial arterial spin labeling, with good accuracy. We also introduce 2 simple methods for assessing model uncertainty. Finally, we demonstrate the ability to adapt the convolution neural network model to a specific false-positive versus false-negative tradeoff. These findings are directly relevant to automatic segmentation in quantitative cardiac MRI and are broadly applicable to automatic segmentation problems in diagnostic imaging.

Allorecognition upon Fungal Cell-Cell Contact Determines Social Cooperation and Impacts the Acquisition of Multicellularity.

Somatic cell fusion and conspecific cooperation are crucial social traits for microbial unicellular-to-multicellular transitions, colony expansion, and substrate foraging but are also associated with risks of parasitism. We identified a cell wall remodeling (cwr) checkpoint that acts upon cell contact to assess genetic compatibility and regulate cell wall dissolution during somatic cell fusion in a wild population of the filamentous fungus Neurospora crassa. Non-allelic interactions between two linked loci, cwr-1 and cwr-2, were necessary and sufficient to block cell fusion: cwr-1 encodes a polysaccharide monooxygenase (PMO), a class of enzymes associated with extracellular degradative capacities, and cwr-2 encodes a predicted transmembrane protein. Mutations of sites in CWR-1 essential for PMO catalytic activity abolished the block in cell fusion between formerly incompatible strains. In Neurospora, alleles cwr-1 and cwr-2 were highly polymorphic, fell into distinct haplogroups, and showed trans-species polymorphisms. Distinct haplogroups and trans-species polymorphisms at cwr-1 and cwr-2 were also identified in the distantly related genus Fusarium, suggesting convergent evolution. Proteins involved in chemotropic processes showed extended localization at contact sites, suggesting that cwr regulates the transition between chemotropic growth and cell wall dissolution. Our work revealed an allorecognition surveillance system based on kind discrimination that inhibits cooperative behavior in fungi by blocking cell fusion upon contact, contributing to fungal immunity by preventing formation of chimeras between genetically non-identical colonies.

Non-contrast assessment of microvascular integrity using arterial spin labeled cardiovascular magnetic resonance in a porcine model of acute myocardial infarction.

BACKGROUND: Following acute myocardial infarction (AMI), microvascular integrity and function may be compromised as a result of microvascular obstruction (MVO) and vasodilator dysfunction. It has been observed that both infarcted and remote myocardial territories may exhibit impaired myocardial blood flow (MBF) patterns associated with an abnormal vasodilator response. Arterial spin labeled (ASL) CMR is a novel non-contrast technique that can quantitatively measure MBF. This study investigates the feasibility of ASL-CMR to assess MVO and vasodilator response in swine. METHODS: Thirty-one swine were included in this study. Resting ASL-CMR was performed on 24 healthy swine (baseline group). A subset of 13 swine from the baseline group underwent stress ASL-CMR to assess vasodilator response. Fifteen swine were subjected to a 90-min left anterior descending (LAD) coronary artery occlusion followed by reperfusion. Resting ASL-CMR was performed post-AMI at 1-2 days (N = 9, of which 6 were from the baseline group), 1-2 weeks (N = 8, of which 4 were from the day 1-2 group), and 4 weeks (N = 4, of which 2 were from the week 1-2 group). Resting first-pass CMR and late gadolinium enhancement (LGE) were performed post-AMI for reference. RESULTS: At rest, regional MBF and physiological noise measured from ASL-CMR were 1.08 ± 0.62 and 0.15 ± 0.10 ml/g/min, respectively. Regional MBF increased to 1.47 ± 0.62 ml/g/min with dipyridamole vasodilation (P < 0.001). Significant reduction in MBF was found in the infarcted region 1-2 days, 1-2 weeks, and 4 weeks post-AMI compared to baseline (P < 0.03). This was consistent with perfusion deficit seen on first-pass CMR and with MVO seen on LGE. There were no significant differences between measured MBF in the remote regions pre and post-AMI (P > 0.60). CONCLUSIONS: ASL-CMR can assess vasodilator response in healthy swine and detect significant reduction in regional MBF at rest following AMI. ASL-CMR is an alternative to gadolinium-based techniques for assessment of MVO and microvascular integrity within infarcted, as well as salvageable and remote myocardium. This has the potential to provide early indications of adverse remodeling processes post-ischemia.

Chemotropism and Cell Fusion in Neurospora crassa Relies on the Formation of Distinct Protein Complexes by HAM-5 and a Novel Protein HAM-14.

In filamentous fungi, communication is essential for the formation of an interconnected, multinucleate, syncytial network, which is constructed via hyphal fusion or fusion of germinated asexual spores (germlings). Anastomosis in filamentous fungi is comparable to other somatic cell fusion events resulting in syncytia, including myoblast fusion during muscle differentiation, macrophage fusion, and fusion of trophoblasts during placental development. In Neurospora crassa, fusion of genetically identical germlings is a highly dynamic and regulated process that requires components of a MAP kinase signal transduction pathway. The kinase pathway components (NRC-1, MEK-2 and MAK-2) and the scaffold protein HAM-5 are recruited to hyphae and germling tips undergoing chemotropic interactions. The MAK-2/HAM-5 protein complex shows dynamic oscillation to hyphae/germling tips during chemotropic interactions, and which is out-of-phase to the dynamic localization of SOFT, which is a scaffold protein for components of the cell wall integrity MAP kinase pathway. In this study, we functionally characterize HAM-5 by generating ham-5 truncation constructs and show that the N-terminal half of HAM-5 was essential for function. This region is required for MAK-2 and MEK-2 interaction and for correct cellular localization of HAM-5 to "fusion puncta." The localization of HAM-5 to puncta was not perturbed in 21 different fusion mutants, nor did these puncta colocalize with components of the secretory pathway. We also identified HAM-14 as a novel member of the HAM-5/MAK-2 pathway by mining MAK-2 phosphoproteomics data. HAM-14 was essential for germling fusion, but not for hyphal fusion. Colocalization and coimmunoprecipitation data indicate that HAM-14 interacts with MAK-2 and MEK-2 and may be involved in recruiting MAK-2 (and MEK-2) to complexes containing HAM-5.