Solution Processing Silicon Heterojunction Photocathode for Efficient and Stable Hydrogen Production.

Efficient and stable photocathodes are crucial for the development of photoelectrochemical (PEC) water-splitting devices. Silicon heterojunction (SHJ) solar cell is one of the most advanced photovoltaic cells. However, due to the instability of its outermost indium tin oxide (ITO) layers in the electrolyte, a protective layer needs to be introduced on its surface. Previously reported high-quality protective layers almost all involved the use of expensive thin film manufacturing techniques such as atomic layer deposition (ALD). In this work, for the first time, a new strategy is proposed of modifying SHJ-based photocathode with yttrium hydroxide (Y(OH)3) through two-step solution methods to simultaneously improve the stability and activity. The optimized SHJ photocathode exhibits a high applied bias photon-to-current efficiency (ABPE) of 8.4% under simulated 100 mW cm-2 (1 Sun) with an AM 1.5G filter in 0.5 m KOH. Furthermore, the obtained SHJ photocathode demonstrates excellent stability of at least 110 h at 0.3 V versus RHE. In this work, combining facile direct current magnetron sputtering with a solution treatment technique provides a novel design strategy, which lowers the threshold for preparing high-quality protective layer, and paves the way for developing economic, efficient, and stable SHJ-based PEC devices.

N6-methyladenosine modification positively regulate Japanese encephalitis virus replication.

N6-methyladenosine (m6A) is present in diverse viral RNA and plays important regulatory roles in virus replication and host antiviral innate immunity. However, the role of m6A in regulating JEV replication has not been investigated. Here, we show that the JEV genome contains m6A modification upon infection of mouse neuroblast cells (neuro2a). JEV infection results in a decrease in the expression of m6A writer METTL3 in mouse brain tissue. METTL3 knockdown by siRNA leads to a substantial decrease in JEV replication and the production of progeny viruses at 48 hpi. Mechanically, JEV triggered a considerable increase in the innate immune response of METTL3 knockdown neuro2a cells compared to the control cells. Our study has revealed the distinctive m6A signatures of both the virus and host in neuro2a cells infected with JEV, illustrating the positive role of m6A modification in JEV infection. Our study further enhances understanding of the role of m6A modification in Flaviviridae viruses.

Precisely Assembling a CoO Cocatalyst onto Tb4O7/CN and Pt-Tb4O7/CN for Promoting Photocatalytic Overall Water Splitting.

Photocatalytic overall water splitting (POWS) is a promising approach for solar-to-hydrogen conversion. For achieving this target, it is urgent to develop efficient photocatalysts. Constructing a heterojunction and loading a cocatalyst are two effective strategies for enhancing POWS. However, how to achieve the cooperation of loading the cocatalyst site with the charge separation of a heterojunction remains a huge challenge. Herein, we present an ingenious method: precisely assembling a H2O2-producing cocatalyst CoO on Tb4O7/CN. Assembling CoO on CN of Tb4O7/CN improves the photoinduced electron-hole pair separation and promotes the POWS performance. Inversely, engineering CoO on Tb4O7 leads to production of Co, deactivating POWS performance with a H2-evolution rate 5.2 times lower than that of Tb4O7/CN. Furthermore, we precisely assemble CoO on the CN section of Pt-oriented Pt-Tb4O7/CN. The bioriented CoO and Pt cooperatively promote photogenerated carrier separation. Consequently, the prepared Pt-Tb4O7/CN-CoO exhibits spectacularly high POWS activity. The H2-evolution rate reaches 450 µmol h-1 g-1, which is about 9.4 times higher than that of the initial Tb4O7/CN. The apparent quantum yield (AQY) for H2 evolution at 420 nm reaches 14.1%, surpassing those of most reported CN-based photocatalysts. This work offers an approach to precisely load cocatalysts on heterojunctions. These findings provide insights for designing cocatalyst-decorated heterojunctions for POWS.

Lower-extremity fatigue fracture detection and grading based on deep learning models of radiographs.

OBJECTIVES: To identify the feasibility of deep learning-based diagnostic models for detecting and assessing lower-extremity fatigue fracture severity on plain radiographs. METHODS: This retrospective study enrolled 1151 X-ray images (tibiofibula/foot: 682/469) of fatigue fractures and 2842 X-ray images (tibiofibula/foot: 2000/842) without abnormal presentations from two clinical centers. After labeling the lesions, images in a center (tibiofibula/foot: 2539/1180) were allocated at 7:1:2 for model construction, and the remaining images from another center (tibiofibula/foot: 143/131) for external validation. A ResNet-50 and a triplet branch network were adopted to construct diagnostic models for detecting and grading. The performances of detection models were evaluated with sensitivity, specificity, and area under the receiver operating characteristic curve (AUC), while grading models were evaluated with accuracy by confusion matrix. Visual estimations by radiologists were performed for comparisons with models. RESULTS: For the detection model on tibiofibula, a sensitivity of 95.4%/85.5%, a specificity of 80.1%/77.0%, and an AUC of 0.965/0.877 were achieved in the internal testing/external validation set. The detection model on foot reached a sensitivity of 96.4%/90.8%, a specificity of 76.0%/66.7%, and an AUC of 0.947/0.911. The detection models showed superior performance to the junior radiologist, comparable to the intermediate or senior radiologist. The overall accuracy of the diagnostic model was 78.5%/62.9% for tibiofibula and 74.7%/61.1% for foot in the internal testing/external validation set. CONCLUSIONS: The deep learning-based models could be applied to the radiological diagnosis of plain radiographs for assisting in the detection and grading of fatigue fractures on tibiofibula and foot. KEY POINTS: • Fatigue fractures on radiographs are relatively difficult to detect, and apt to be misdiagnosed. • Detection and grading models based on deep learning were constructed on a large cohort of radiographs with lower-extremity fatigue fractures. • The detection model with high sensitivity would help to reduce the misdiagnosis of lower-extremity fatigue fractures.

Immature permanent incisors with complicated crown fractures treated with partial pulpotomy using white mineral trioxide aggregate and IRoot BP plus-a retrospective long-term study.

BACKGROUND/AIMS: Calcium silicate cements have been widely used for pulpotomies in immature permanent teeth with complicated crown fractures due to their superior properties. However, few studies have evaluated the long-term outcomes of white mineral trioxide aggregate (WMTA) and iRoot BP Plus for partial pulpotomies. The aim of this study was to investigate the long-term clinical and radiographic outcomes of WMTA and iRoot BP Plus for partial pulpotomies in immature permanent incisors with complicated crown fractures. MATERIALS AND METHODS: Children who had partial pulpotomies of immature permanent incisors with complicated crown fractures using WMTA or iRoot BP Plus as capping agents were enrolled. Eighty immature permanent incisors in 68 children (aged 8-13 years) were included. They were divided into two groups (WMTA and iRoot BP Plus) according to the capping agents. Clinical and radiographic information was collected during a 5-year follow-up period. Study data were analyzed using Chi-square tests or Fisher exact tests. RESULTS: The clinical and radiographic success rates in the WMTA (n = 36) and iRoot BP Plus groups (n = 44) were 94.4% versus 97.7% and 88.9% versus 97.7%, respectively (both p < .05). The average observation period was 74.5 ± 13.2 months and 61.9 ± 1.6 months in the WMTA and iRoot BP Plus groups, respectively (p < .01). Five cases presented with periapical radiolucencies. The WMTA group had four cases of pulp canal calcification (11.1%), while the iRoot BP Plus group had two cases (4.6%). There was crown discolouration in all cases in the WMTA group, but none in the iRoot BP Plus group. CONCLUSION: Both WMTA and iRoot BP Plus had favorable outcomes in promoting physiological development and maintaining the basic functions of immature permanent incisors with complicated crown fractures. As a partial pulpotomy material, iRoot BP Plus may be more suitable for the esthetic zone than WMTA.

Studies on the Effect of Lipofectamine and Cell-Penetrating Peptide on the Properties of 10-23 DNAzyme.

Cationic polymeric materials and cell-penetrating peptides (CPPs) were often used as the delivery vectors in the evaluation of nucleic acid therapeutics. 10-23 DNAzyme is a kind of potential antisense therapeutics by catalytic cleavage of the disease-related RNAs. Here, lipofectamine 2000 and Tat peptide were evaluated for their effect on the catalytic activity of 10-23 DNAzyme, with the observed rate constant, thermal stability, CD spectra, and PAGE analysis, with a duplex DNA mimicking DNAzyme-substrate as a control. It was shown that the cationic carriers had a negative effect on the catalytic performance of the 10-23 DNAzyme. Significantly, the destabilizing effect of the cationic carriers on the duplex formation was noteworthy, as a duplex formation is an essential prerequisite in the silencing mechanisms of antisense and RNAi.

Theoretical and Experimental Studies of Gallate Melilite Electrides from Topotactic Reduction of Interstitial Oxide Ion Conductors.

A nonstoichiometric La1.5Sr0.5Ga3O7.25 melilite oxide ion conductor features active interstitial oxygen defects in its pentagonal rings with high mobility. In this study, electron localization function calculated by density functional theory indicated that the interstitial oxide ions located in the pentagonal rings of gallate melilites may be removed and replaced by electron anions that are confined within the pentagonal rings, which would therefore convert the melilite interstitial oxide ion conductor into a zero-dimensional (0D) electride. The more active interstitial oxide ions, compared to the framework oxide ions, make the La1.5Sr0.5Ga3O7.25 melilite structure more reducible by CaH2 using topotactic reduction, in contrast to the hardly reducible nature of parent LaSrGa3O7. The topotactic reduction enhances the bulk electronic conduction (σ ∼ 0.003 S/cm at 400 °C) by ∼ 1 order of magnitude for La1.5Sr0.5Ga3O7.25. The oxygen loss in the melilite structure was verified and most likely took place on the active interstitial oxide ions. The identified confinement space for electronic anions in melilite interstitial oxide ion conductors presented here provides a strategy to access inorganic electrides from interstitial oxide ion conductor electrolytes.

Comparing methods of detecting and segmenting unruptured intracranial aneurysms on TOF-MRAS: The ADAM challenge.

Accurate detection and quantification of unruptured intracranial aneurysms (UIAs) is important for rupture risk assessment and to allow an informed treatment decision to be made. Currently, 2D manual measures used to assess UIAs on Time-of-Flight magnetic resonance angiographies (TOF-MRAs) lack 3D information and there is substantial inter-observer variability for both aneurysm detection and assessment of aneurysm size and growth. 3D measures could be helpful to improve aneurysm detection and quantification but are time-consuming and would therefore benefit from a reliable automatic UIA detection and segmentation method. The Aneurysm Detection and segMentation (ADAM) challenge was organised in which methods for automatic UIA detection and segmentation were developed and submitted to be evaluated on a diverse clinical TOF-MRA dataset. A training set (113 cases with a total of 129 UIAs) was released, each case including a TOF-MRA, a structural MR image (T1, T2 or FLAIR), annotation of any present UIA(s) and the centre voxel of the UIA(s). A test set of 141 cases (with 153 UIAs) was used for evaluation. Two tasks were proposed: (1) detection and (2) segmentation of UIAs on TOF-MRAs. Teams developed and submitted containerised methods to be evaluated on the test set. Task 1 was evaluated using metrics of sensitivity and false positive count. Task 2 was evaluated using dice similarity coefficient, modified hausdorff distance (95th percentile) and volumetric similarity. For each task, a ranking was made based on the average of the metrics. In total, eleven teams participated in task 1 and nine of those teams participated in task 2. Task 1 was won by a method specifically designed for the detection task (i.e. not participating in task 2). Based on segmentation metrics, the top two methods for task 2 performed statistically significantly better than all other methods. The detection performance of the top-ranking methods was comparable to visual inspection for larger aneurysms. Segmentation performance of the top ranking method, after selection of true UIAs, was similar to interobserver performance. The ADAM challenge remains open for future submissions and improved submissions, with a live leaderboard to provide benchmarking for method developments at https://adam.isi.uu.nl/.

Development of Novel Anti-influenza Thiazolides with Relatively Broad-Spectrum Antiviral Potentials.

Seasonal and pandemic influenza causes 650,000 deaths annually in the world. The emergence of drug resistance to specific anti-influenza virus drugs such as oseltamivir and baloxavir marboxil highlights the urgency of novel anti-influenza chemical entity discovery. In this study, we report a series of novel thiazolides derived from an FDA-approved drug, nitazoxanide, with antiviral activity against influenza and a broad range of viruses. The preferred candidates 4a and 4d showed significantly enhanced anti-influenza virus potentials, with 10-fold improvement compared to results with nitazoxanide, and were effective against a variety of influenza virus subtypes including oseltamivir-resistant strains. Notably, the combination using compounds 4a/4d and oseltamivir carboxylate or zanamivir displayed synergistic antiviral effects against oseltamivir-resistant strains. Mode-of-action analysis demonstrated that compounds 4a/4d acted at the late phase of the viral infection cycle through inhibiting viral RNA transcription and replication. Further experiments showed that treatment with compounds 4a/4d significantly inhibited influenza virus infection in human lung organoids, suggesting the druggability of the novel thiazolides. In-depth transcriptome analysis revealed a series of upregulated cellular genes that may contribute to the antiviral activities of 4a/4d. Together, the results of our study indicated the direction to optimize nitazoxanide as an anti-influenza drug and discovered two candidates with novel structures, compounds 4a/4d, that have relatively broad-spectrum antiviral potentials.

Development and validation of an artificial intelligence system for grading colposcopic impressions and guiding biopsies.

BACKGROUND: Colposcopy diagnosis and directed biopsy are the key components in cervical cancer screening programs. However, their performance is limited by the requirement for experienced colposcopists. This study aimed to develop and validate a Colposcopic Artificial Intelligence Auxiliary Diagnostic System (CAIADS) for grading colposcopic impressions and guiding biopsies. METHODS: Anonymized digital records of 19,435 patients were obtained from six hospitals across China. These records included colposcopic images, clinical information, and pathological results (gold standard). The data were randomly assigned (7:1:2) to a training and a tuning set for developing CAIADS and to a validation set for evaluating performance. RESULTS: The agreement between CAIADS-graded colposcopic impressions and pathology findings was higher than that of colposcopies interpreted by colposcopists (82.2% versus 65.9%, kappa 0.750 versus 0.516, p < 0.001). For detecting pathological high-grade squamous intraepithelial lesion or worse (HSIL+), CAIADS showed higher sensitivity than the use of colposcopies interpreted by colposcopists at either biopsy threshold (low-grade or worse 90.5%, 95% CI 88.9-91.4% versus 83.5%, 81.5-85.3%; high-grade or worse 71.9%, 69.5-74.2% versus 60.4%, 57.9-62.9%; all p < 0.001), whereas the specificities were similar (low-grade or worse 51.8%, 49.8-53.8% versus 52.0%, 50.0-54.1%; high-grade or worse 93.9%, 92.9-94.9% versus 94.9%, 93.9-95.7%; all p > 0.05). The CAIADS also demonstrated a superior ability in predicting biopsy sites, with a median mean-intersection-over-union (mIoU) of 0.758. CONCLUSIONS: The CAIADS has potential in assisting beginners and for improving the diagnostic quality of colposcopy and biopsy in the detection of cervical precancer/cancer.

An objective comparison of cell-tracking algorithms.

We present a combined report on the results of three editions of the Cell Tracking Challenge, an ongoing initiative aimed at promoting the development and objective evaluation of cell segmentation and tracking algorithms. With 21 participating algorithms and a data repository consisting of 13 data sets from various microscopy modalities, the challenge displays today's state-of-the-art methodology in the field. We analyzed the challenge results using performance measures for segmentation and tracking that rank all participating methods. We also analyzed the performance of all of the algorithms in terms of biological measures and practical usability. Although some methods scored high in all technical aspects, none obtained fully correct solutions. We found that methods that either take prior information into account using learning strategies or analyze cells in a global spatiotemporal video context performed better than other methods under the segmentation and tracking scenarios included in the challenge.

A light- and heat-driven glycal diazidation approach to nitrogenous carbohydrate derivatives with antiviral activity.

The aminated mimetics of 2-keto-3-deoxy-sugar acids such as the anti-influenza clinical drugs oseltamivir (Tamiflu) and zanamivir (Relenza) are important bioactive molecules. Development of synthetic methodologies for accessing such compound collections is highly desirable. Herein, we describe a simple, catalyst-free glycal diazidation protocol enabled by visible light-driven conditions. This new method requires neither acid promoters nor transition-metal catalysts and takes place at ambient temperature within 1-2 hours. Notably, the desired transformations could be promoted by thermal conditions as well, albeit with lower efficacy compared to the light-induced conditions. Different sugar acid-derived glycal templates have been converted into a range of 2,3-diazido carbohydrate analogs by harnessing this mild and scalable approach, leading to the discovery of new antiviral agents.

Reverse active learning based atrous DenseNet for pathological image classification.

BACKGROUND: Due to the recent advances in deep learning, this model attracted researchers who have applied it to medical image analysis. However, pathological image analysis based on deep learning networks faces a number of challenges, such as the high resolution (gigapixel) of pathological images and the lack of annotation capabilities. To address these challenges, we propose a training strategy called deep-reverse active learning (DRAL) and atrous DenseNet (ADN) for pathological image classification. The proposed DRAL can improve the classification accuracy of widely used deep learning networks such as VGG-16 and ResNet by removing mislabeled patches in the training set. As the size of a cancer area varies widely in pathological images, the proposed ADN integrates the atrous convolutions with the dense block for multiscale feature extraction. RESULTS: The proposed DRAL and ADN are evaluated using the following three pathological datasets: BACH, CCG, and UCSB. The experiment results demonstrate the excellent performance of the proposed DRAL + ADN framework, achieving patch-level average classification accuracies (ACA) of 94.10%, 92.05% and 97.63% on the BACH, CCG, and UCSB validation sets, respectively. CONCLUSIONS: The DRAL + ADN framework is a potential candidate for boosting the performance of deep learning models for partially mislabeled training datasets.

Skin Lesion Analysis towards Melanoma Detection Using Deep Learning Network.

Skin lesions are a severe disease globally. Early detection of melanoma in dermoscopy images significantly increases the survival rate. However, the accurate recognition of melanoma is extremely challenging due to the following reasons: low contrast between lesions and skin, visual similarity between melanoma and non-melanoma lesions, etc. Hence, reliable automatic detection of skin tumors is very useful to increase the accuracy and efficiency of pathologists. In this paper, we proposed two deep learning methods to address three main tasks emerging in the area of skin lesion image processing, i.e., lesion segmentation (task 1), lesion dermoscopic feature extraction (task 2) and lesion classification (task 3). A deep learning framework consisting of two fully convolutional residual networks (FCRN) is proposed to simultaneously produce the segmentation result and the coarse classification result. A lesion index calculation unit (LICU) is developed to refine the coarse classification results by calculating the distance heat-map. A straight-forward CNN is proposed for the dermoscopic feature extraction task. The proposed deep learning frameworks were evaluated on the ISIC 2017 dataset. Experimental results show the promising accuracies of our frameworks, i.e., 0.753 for task 1, 0.848 for task 2 and 0.912 for task 3 were achieved.

Design, synthesis and evaluation of 2,2-dimethyl-1,3-dioxolane derivatives as human rhinovirus 3C protease inhibitors.

The human rhinovirus (HRV) is the most significant cause of the common cold all over the world. The maturation and replication of this virus entirely depend on the activity of a virus-encoded 3C protease. Due to the high conservation among different serotypes and the minimal homology existing between 3C protease and known mammalian enzymes, 3C protease has been regarded as an attractive target for the treatment of HRV infections. In this study, we identified a novel (4R,5R)-N4-(2-((3-methoxyphenyl)amino)ethyl)-2,2-dimethyl-N5-(naphthalen-2-yl)-1,3-dioxolane-4,5-dicarboxamide (7a) to be a HRV 3C protease inhibitor via virtual screening. Further research has been focused on the design, synthesis and in vitro biological evaluation of 7a derivatives. The studies revealed that compound 7d has an IC50 value of 2.50±0.7µM against HRV 3C protease, and it thus could serve as a promising compound for the development of novel anti-rhinoviral medicines.

Immobilization of PDMS-SiO2-TiO2 composite for the photocatalytic degradation of dye AO-7.

Crack-free PDMS-SiO2-TiO2 composite as photocatalyst was prepared for degrading dyes by using thin-film fixed bed reactor. The hydrophobic surface of the photocatalyst loaded with PDMS-SiO2-TiO2 composite could be considered as an extractant for organic pollutants. The effect of different supports including pumice stone, medicinal stone, and fiberglass for photocatalytic efficiency were compared. Under the same condition, it was found that the photocatalytic degradation effect of dyes was best when PDMS-SiO2-TiO2 composite was fixed on pumice stone rather than medicinal stone or fiberglass. Furthermore, when pumice stone was used as the support for PDMS-SiO2-TiO2 composite, the photocatalytic degradation effect of dyes hardly decreased after five cycles.

Generation and characterization of a protective mouse monoclonal antibody against human enterovirus 71.

Human enterovirus 71 (EV71) infection has emerged as a major threat to children; however, no effective antiviral treatment or vaccine is currently available. Antibody-based treatment shows promises to control this growing public health problem of EV71 infection, and a few potent monoclonal antibodies (mAbs) targeting viral capsid protein have been well described. Here, we generated an EV71-specific mouse mAb 2G8 that conferred full protection against lethal EV71 challenge in a suckling mouse model. 2G8 belonged to IgM isotype and neutralized EV71 at the attachment stage. Biochemical assays mapped the binding epitope of 2G8 to the SP70 peptide, which spanning amino acid residues 208-222 on the VP1 protein. Alanine scanning mutagenesis defined the essential roles of multiple residues, including Y208, T210, G212, K215, K218, L220, E221, and Y222, for 2G8 binding. Then, a panel of single mutation was individually introduced into the EV71 infectious clone by reverse genetics, and three mutant viruses, K215A, K218A, and L220A, were successfully recovered and characterized. Biochemical and neutralization assays revealed that K218A mutant partially escaped 2G8 neutralization, while L220A completely abolished 2G8 binding and neutralization. In particular, neutralization assays with human sera demonstrated that K218A and L220A substitutions are also critical for antibody neutralization in natural infection population. These findings not only generate a protective mAb candidate with therapeutic potential but also provide insights into antibody-mediated EV71 neutralization mechanism.

Recombinant tandem multi-linear neutralizing epitopes of human enterovirus 71 elicited protective immunity in mice.

BACKGROUND: Human Enterovirus 71 (EV71) has emerged as the leading cause of viral encephalitis in children, especially in the Asia-Pacific regions. EV71 vaccine development is of high priority at present, and neutralization antibodies have been documented to play critical roles during in vitro and in vivo protection against EV71 infection. RESULTS: In this study, a novel strategy to produce EV71 vaccine candidate based on recombinant multiple tandem linear neutralizing epitopes (mTLNE) was proposed. The three well identified EV71 linear neutralizing epitopes in capsid proteins, VP1-SP55, VP1-SP70 and VP2-SP28, were sequentially linked by a Gly-Ser linker ((G4S)3), and expressed in E.coli in fusion with the Trx and His tag at either terminal. The recombinant protein mTLNE was soluble and could be purified by standard affinity chromatography. Following three dosage of immunization in adult mice, EV71-specific IgG and neutralization antibodies were readily induced by recombinant mTLNE. IgG subtyping demonstrated that lgG1 antibodies dominated the mTLNE-induced humoral immune response. Especially, cytokine profiling in spleen cells from the mTLNE-immunized mice revealed high production of IL-4 and IL-6. Finally, in vivo challenge experiments showed that passive transfer with anti-mTLNE sera conferred full protection against lethal EV71 challenge in neonatal mice. CONCLUSION: Our results demonstrated that this rational designed recombinant mTLNE might have the potential to be further developed as an EV71 vaccine in the future.

Constructing an Open-Structured J-Type ZnIn2S4/In(OH)3 Heterojunction for Photocatalytical Hydrogen Generation.

Constructing heterojunctions to separate photogenerated carriers is an effective strategy for improving the efficiency of photocatalytic reactions. A J-type heterojunction is a recently reported efficient anisotropic heterojunction. Herein, taking anisotropic ZnIn2S4 (ZIS) nanosheets as an example of a type-II heterojunction, we report for the first time the concept of open and closed structures (O and C structure) of J-type heterojunctions. A simple ammonia-post-treatment method was employed to prepare the O- and C-structured J-type ZnIn2S4/In(OH)3 (ZIS/IOH) heterojunctions. The O-structured J-type ZIS/IOH (OJ-ZIS/IOH) heterojunction exhibits a high hydrogen production activity, reaching 400 µmol·h-1, 2.67 times higher than that of pristine ZIS. However, the activity of the C-structured heterojunction (CJ-ZIS/IOH) is close to that of pristine ZIS. The findings emphasize the importance of the cooperation of photogenerated carrier separation and transport in J-type heterojunctions, providing insights into developing efficient heterojunction photocatalysts.

Automatic view plane prescription for cardiac magnetic resonance imaging via supervision by spatial relationship between views.

BACKGROUND: View planning for the acquisition of cardiac magnetic resonance (CMR) imaging remains a demanding task in clinical practice. PURPOSE: Existing approaches to its automation relied either on an additional volumetric image not typically acquired in clinic routine, or on laborious manual annotations of cardiac structural landmarks. This work presents a clinic-compatible, annotation-free system for automatic CMR view planning. METHODS: The system mines the spatial relationship-more specifically, locates the intersecting lines-between the target planes and source views, and trains U-Net-based deep networks to regress heatmaps defined by distances from the intersecting lines. On the one hand, the intersection lines are the prescription lines prescribed by the technologists at the time of image acquisition using cardiac landmarks, and retrospectively identified from the spatial relationship. On the other hand, as the spatial relationship is self-contained in properly stored data, for example, in the DICOM format, the need for additional manual annotation is eliminated. In addition, the interplay of the multiple target planes predicted in a source view is utilized in a stacked hourglass architecture consisting of repeated U-Net-style building blocks to gradually improve the regression. Then, a multiview planning strategy is proposed to aggregate information from the predicted heatmaps for all the source views of a target plane, for a globally optimal prescription, mimicking the similar strategy practiced by skilled human prescribers. For performance evaluation, the retrospectively identified planes prescribed by the technologists are used as the ground truth, and the plane angle differences and localization distances between the planes prescribed by our system and the ground truth are compared. RESULTS: The retrospective experiments include 181 clinical CMR exams, which are randomly split into training, validation, and test sets in the ratio of 64:16:20. Our system yields the mean angular difference and point-to-plane distance of 5.68 ∘ $^\circ$ and 3.12 mm, respectively, on the held-out test set. It not only achieves superior accuracy to existing approaches including conventional atlas-based and newer deep-learning-based in prescribing the four standard CMR planes but also demonstrates prescription of the first cardiac-anatomy-oriented plane(s) from the body-oriented scout. CONCLUSIONS: The proposed system demonstrates accurate automatic CMR view plane prescription based on deep learning on properly archived data, without the need for further manual annotation. This work opens a new direction for automatic view planning of anatomy-oriented medical imaging beyond CMR.