A deep-learning model for intracranial aneurysm detection on CT angiography images in China: a stepwise, multicentre, early-stage clinical validation study.

BACKGROUND: Artificial intelligence (AI) models in real-world implementation are scarce. Our study aimed to develop a CT angiography (CTA)-based AI model for intracranial aneurysm detection, assess how it helps clinicians improve diagnostic performance, and validate its application in real-world clinical implementation. METHODS: We developed a deep-learning model using 16 546 head and neck CTA examination images from 14 517 patients at eight Chinese hospitals. Using an adapted, stepwise implementation and evaluation, 120 certified clinicians from 15 geographically different hospitals were recruited. Initially, the AI model was externally validated with images of 900 digital subtraction angiography-verified CTA cases (examinations) and compared with the performance of 24 clinicians who each viewed 300 of these cases (stage 1). Next, as a further external validation a multi-reader multi-case study enrolled 48 clinicians to individually review 298 digital subtraction angiography-verified CTA cases (stage 2). The clinicians reviewed each CTA examination twice (ie, with and without the AI model), separated by a 4-week washout period. Then, a randomised open-label comparison study enrolled 48 clinicians to assess the acceptance and performance of this AI model (stage 3). Finally, the model was prospectively deployed and validated in 1562 real-world clinical CTA cases. FINDINGS: The AI model in the internal dataset achieved a patient-level diagnostic sensitivity of 0·957 (95% CI 0·939-0·971) and a higher patient-level diagnostic sensitivity than clinicians (0·943 [0·921-0·961] vs 0·658 [0·644-0·672]; p<0·0001) in the external dataset. In the multi-reader multi-case study, the AI-assisted strategy improved clinicians' diagnostic performance both on a per-patient basis (the area under the receiver operating characteristic curves [AUCs]; 0·795 [0·761-0·830] without AI vs 0·878 [0·850-0·906] with AI; p<0·0001) and a per-aneurysm basis (the area under the weighted alternative free-response receiver operating characteristic curves; 0·765 [0·732-0·799] vs 0·865 [0·839-0·891]; p<0·0001). Reading time decreased with the aid of the AI model (87·5 s vs 82·7 s, p<0·0001). In the randomised open-label comparison study, clinicians in the AI-assisted group had a high acceptance of the AI model (92·6% adoption rate), and a higher AUC when compared with the control group (0·858 [95% CI 0·850-0·866] vs 0·789 [0·780-0·799]; p<0·0001). In the prospective study, the AI model had a 0·51% (8/1570) error rate due to poor-quality CTA images and recognition failure. The model had a high negative predictive value of 0·998 (0·994-1·000) and significantly improved the diagnostic performance of clinicians; AUC improved from 0·787 (95% CI 0·766-0·808) to 0·909 (0·894-0·923; p<0·0001) and patient-level sensitivity improved from 0·590 (0·511-0·666) to 0·825 (0·759-0·880; p<0·0001). INTERPRETATION: This AI model demonstrated strong clinical potential for intracranial aneurysm detection with improved clinician diagnostic performance, high acceptance, and practical implementation in real-world clinical cases. FUNDING: National Natural Science Foundation of China. TRANSLATION: For the Chinese translation of the abstract see Supplementary Materials section.

Extreme environments and human health: From the immune microenvironments to immune cells.

Exposure to extreme environments causes specific acute and chronic physiological responses in humans. The adaptation and the physiological processes under extreme environments predominantly affect multiple functional systems of the organism, in particular, the immune system. Dysfunction of the immune system affected by several extreme environments (including hyperbaric environment, hypoxia, blast shock, microgravity, hypergravity, radiation exposure, and magnetic environment) has been observed from clinical macroscopic symptoms to intracorporal immune microenvironments. Therefore, simulated extreme conditions are engineered for verifying the main influenced characteristics and factors in the immune microenvironments. This review summarizes the responses of immune microenvironments to these extreme environments during in vivo or in vitro exposure, and the approaches of engineering simulated extreme environments in recent decades. The related microenvironment engineering, signaling pathways, molecular mechanisms, clinical therapy, and prevention strategies are also discussed.

Reliable Mutual Distillation for Medical Image Segmentation Under Imperfect Annotations.

Convolutional neural networks (CNNs) have made enormous progress in medical image segmentation. The learning of CNNs is dependent on a large amount of training data with fine annotations. The workload of data labeling can be significantly relieved via collecting imperfect annotations which only match the underlying ground truths coarsely. However, label noises which are systematically introduced by the annotation protocols, severely hinders the learning of CNN-based segmentation models. Hence, we devise a novel collaborative learning framework in which two segmentation models cooperate to combat label noises in coarse annotations. First, the complementary knowledge of two models is explored by making one model clean training data for the other model. Secondly, to further alleviate the negative impact of label noises and make sufficient usage of the training data, the specific reliable knowledge of each model is distilled into the other model with augmentation-based consistency constraints. A reliability-aware sample selection strategy is incorporated for guaranteeing the quality of the distilled knowledge. Moreover, we employ joint data and model augmentations to expand the usage of reliable knowledge. Extensive experiments on two benchmarks showcase the superiority of our proposed method against existing methods under annotations with different noise levels. For example, our approach can improve existing methods by nearly 3% DSC on the lung lesion segmentation dataset LIDC-IDRI under annotations with 80% noise ratio. Code is available at: https://github.com/Amber-Believe/ReliableMutualDistillation.

Graph Convolution Based Cross-Network Multiscale Feature Fusion for Deep Vessel Segmentation.

Vessel segmentation is widely used to help with vascular disease diagnosis. Vessels reconstructed using existing methods are often not sufficiently accurate to meet clinical use standards. This is because 3D vessel structures are highly complicated and exhibit unique characteristics, including sparsity and anisotropy. In this paper, we propose a novel hybrid deep neural network for vessel segmentation. Our network consists of two cascaded subnetworks performing initial and refined segmentation respectively. The second subnetwork further has two tightly coupled components, a traditional CNN-based U-Net and a graph U-Net. Cross-network multi-scale feature fusion is performed between these two U-shaped networks to effectively support high-quality vessel segmentation. The entire cascaded network can be trained from end to end. The graph in the second subnetwork is constructed according to a vessel probability map as well as appearance and semantic similarities in the original CT volume. To tackle the challenges caused by the sparsity and anisotropy of vessels, a higher percentage of graph nodes are distributed in areas that potentially contain vessels while a higher percentage of edges follow the orientation of potential nearby vessels. Extensive experiments demonstrate our deep network achieves state-of-the-art 3D vessel segmentation performance on multiple public and in-house datasets.

GREN: Graph-Regularized Embedding Network for Weakly-Supervised Disease Localization in X-Ray Images.

Locating diseases in chest X-ray images with few careful annotations saves large human effort. Recent works approached this task with innovative weakly-supervised algorithms such as multi-instance learning (MIL) and class activation maps (CAM), however, these methods often yield inaccurate or incomplete regions. One of the reasons is the neglection of the pathological implications hidden in the relationship across anatomical regions within each image and the relationship across images. In this paper, we argue that the cross-region and cross-image relationship, as contextual and compensating information, is vital to obtain more consistent and integral regions. To model the relationship, we propose the Graph Regularized Embedding Network (GREN), which leverages the intra-image and inter-image information to locate diseases on chest X-ray images. GREN uses a pre-trained U-Net to segment the lung lobes, and then models the intra-image relationship between the lung lobes using an intra-image graph to compare different regions. Meanwhile, the relationship between in-batch images is modeled by an inter-image graph to compare multiple images. This process mimics the training and decision-making process of a radiologist: comparing multiple regions and images for diagnosis. In order for the deep embedding layers of the neural network to retain structural information (important in the localization task), we use the Hash coding and Hamming distance to compute the graphs, which are used as regularizers to facilitate training. By means of this, our approach achieves the state-of-the-art result on NIH chest X-ray dataset for weakly-supervised disease localization. Our codes are accessible online.

MFI-Net: Multiscale Feature Interaction Network for Retinal Vessel Segmentation.

Segmentation of retinal vessels on fundus images plays a critical role in the diagnosis of micro-vascular and ophthalmological diseases. Although being extensively studied, this task remains challenging due to many factors including the highly variable vessel width and poor vessel-background contrast. In this paper, we propose a multiscale feature interaction network (MFI-Net) for retinal vessel segmentation, which is a U-shaped convolutional neural network equipped with the pyramid squeeze-and-excitation (PSE) module, coarse-to-fine (C2F) module, deep supervision, and feature fusion. We extend the SE operator to multiscale features, resulting in the PSE module, which uses the channel attention learned at multiple scales to enhance multiscale features and enables the network to handle the vessels with variable width. We further design the C2F module to generate and re-process the residual feature maps, aiming to preserve more vessel details during the decoding process. The proposed MFI-Net has been evaluated against several public models on the DRIVE, STARE, CHASE_DB1, and HRF datasets. Our results suggest that both PSE and C2F modules are effective in improving the accuracy of MFI-Net, and also indicate that our model has superior segmentation performance and generalization ability over existing models on four public datasets.

Single-cell RNA sequencing to track novel perspectives in HSC heterogeneity.

As the importance of cell heterogeneity has begun to be emphasized, single-cell sequencing approaches are rapidly adopted to study cell heterogeneity and cellular evolutionary relationships of various cells, including stem cell populations. The hematopoietic stem and progenitor cell (HSPC) compartment contains HSC hematopoietic stem cells (HSCs) and distinct hematopoietic cells with different abilities to self-renew. These cells perform their own functions to maintain different hematopoietic lineages. Undeniably, single-cell sequencing approaches, including single-cell RNA sequencing (scRNA-seq) technologies, empower more opportunities to study the heterogeneity of normal and pathological HSCs. In this review, we discuss how these scRNA-seq technologies contribute to tracing origin and lineage commitment of HSCs, profiling the bone marrow microenvironment and providing high-resolution dissection of malignant hematopoiesis, leading to exciting new findings in HSC biology.

Comparison of automated and manual DWI-ASPECTS in acute ischemic stroke: total and region-specific assessment.

OBJECTIVE: To compare the DWI-Alberta Stroke Program Early Computed Tomography Score calculated by a deep learning-based automatic software tool (eDWI-ASPECTS) with the neuroradiologists' evaluation for the acute stroke, with emphasis on its performance on 10 individual ASPECTS regions, and to determine the reasons for inconsistencies between eDWI-ASPECTS and neuroradiologists' evaluation. METHODS: This retrospective study included patients with middle cerebral artery stroke who underwent MRI from 2010 to 2019. All scans were evaluated by eDWI-ASPECTS and two independent neuroradiologists (with 15 and 5 years of experience in stroke study). Inter-rater agreement and agreement between manual vs. automated methods for total and each region were evaluated by calculating Kendall's tau-b, intraclass correlation coefficient (ICC), and kappa coefficient. RESULTS: In total, 309 patients met our study criteria. For total ASPECTS, eDWI-ASPECTS and manual raters had a strong positive correlation (Kendall's tau-b = 0.827 for junior raters vs. eDWI-ASPECTS; Kendall's tau-b = 0.870 for inter-raters; Kendall's tau-b = 0.848 for senior raters vs. eDWI-ASPECTS) and excellent agreement (ICC = 0.923 for junior raters and automated scores; ICC = 0.954 for inter-raters; ICC = 0.939 for senior raters and automated scores). Agreement was different for individual ASPECTS regions. All regions except for M5 region (κ = 0.216 for junior raters and automated scores), internal capsule (κ = 0.525 for junior raters and automated scores), and caudate (κ = 0.586 for senior raters and automated scores) showed good to excellent concordance. CONCLUSION: The eDWI-ASPECTS performed equally well as senior neuroradiologists' evaluation, although interference by uncertain scoring rules and midline shift resulted in poor to moderate consistency in the M5, internal capsule, and caudate nucleus regions. KEY POINTS: • The eDWI-ASPECTS based on deep learning perform equally well as senior neuroradiologists' evaluations. • Among the individual ASPECTS regions, the M5, internal capsule, and caudate regions mainly affected the overall consistency. • Uncertain scoring rules and midline shift are the main reasons for regional inconsistency.

Advances in Deep Learning-Based Medical Image Analysis.

Importance. With the booming growth of artificial intelligence (AI), especially the recent advancements of deep learning, utilizing advanced deep learning-based methods for medical image analysis has become an active research area both in medical industry and academia. This paper reviewed the recent progress of deep learning research in medical image analysis and clinical applications. It also discussed the existing problems in the field and provided possible solutions and future directions.Highlights. This paper reviewed the advancement of convolutional neural network-based techniques in clinical applications. More specifically, state-of-the-art clinical applications include four major human body systems: the nervous system, the cardiovascular system, the digestive system, and the skeletal system. Overall, according to the best available evidence, deep learning models performed well in medical image analysis, but what cannot be ignored are the algorithms derived from small-scale medical datasets impeding the clinical applicability. Future direction could include federated learning, benchmark dataset collection, and utilizing domain subject knowledge as priors.Conclusion. Recent advanced deep learning technologies have achieved great success in medical image analysis with high accuracy, efficiency, stability, and scalability. Technological advancements that can alleviate the high demands on high-quality large-scale datasets could be one of the future developments in this area.

A clinically applicable deep-learning model for detecting intracranial aneurysm in computed tomography angiography images.

Intracranial aneurysm is a common life-threatening disease. Computed tomography angiography is recommended as the standard diagnosis tool; yet, interpretation can be time-consuming and challenging. We present a specific deep-learning-based model trained on 1,177 digital subtraction angiography verified bone-removal computed tomography angiography cases. The model has good tolerance to image quality and is tested with different manufacturers. Simulated real-world studies are conducted in consecutive internal and external cohorts, in which it achieves an improved patient-level sensitivity and lesion-level sensitivity compared to that of radiologists and expert neurosurgeons. A specific cohort of suspected acute ischemic stroke is employed and it is found that 99.0% predicted-negative cases can be trusted with high confidence, leading to a potential reduction in human workload. A prospective study is warranted to determine whether the algorithm could improve patients' care in comparison to clinicians' assessment.

Reliability and Reproducibility of the Diameters Method in Rapid Determination of Acute Infarct Volume in Magnetic Resonance Diffusion-Weighted Imaging.

BACKGROUND: In acute ischemic stroke, diffusion-weighted imaging (DWI) volume is an independent predictive factor of poor outcome and an exclusion criterion for thrombolytic treatment. A simplified diameters method (ABC/2, orthogonal diameter [OD], and the maximum diameter [MD]) was proposed to replace the conventional measuring method and overcome the tedious and time-consuming defects, but its accuracy remains to be determined. OBJECTIVE: The objective of this study is to clarify the reliability and reproducibility of the diameter-based estimations in the infarct volume in DWI (Vol-DWI) measured by automated software. METHODS: Data of 316 patients with acute ischemic stroke who underwent MRI within 72 h at Jinling Hospital were retrospectively reviewed. Subgroup analysis by the location (cortex, white matter and deep gray nuclei, and combined) and volume (<70 and >70 mL) of cerebral infarction was evaluated. Relationship and consistency between the diameters methods and Vol-DWI were determined using Spearman rank correlation, Wilcoxon signed-rank test, and Bland-Altman plots. The OD and MD thresholds indicating infarct size >15, 70, and 100 mL were determined by generating receiver-operating characteristic (ROC) curves. Interobserver reliability was established using intraclass correlation coefficient and Bland-Altman plot. RESULTS: There was a strong positive correlation between the diameters and the Vol-DWI (ABC/2: r = 0.992, OD: r = 0.984, MD: r = 0.970, p < 0.001). Infarct volumes measured using the ABC/2 formula were significantly lower than those measured with Vol-DWI (Wilcoxon signed-rank test, z = 6.476, p < 0.001). Bland-Altman plot showed that the agreement of the volume <70 mL group, and white matter and deep gray nuclei groups was better than that of the other subgroups. For infarct volumes >15, 70, and 100 mL, the cutoff value for the MD was identified at 5, 6.9, and 8.4 cm, and the OD was identified at 12.47, 26.4, and 36.4 cm2, respectively, with a sensitivity and specificity >90%. CONCLUSIONS: The MD method was the best for achieving a rapid and excellent interobserver reliability for estimating infarct volume. Both OD and MD methods can quickly screen patients suitable for recanalization treatment and predict poor prognosis through threshold evaluation.

Probiotic culture supernatant improves metabolic function through FGF21-adiponectin pathway in mice.

High-fat/high-fructose diet plus intermittent hypoxia exposure (HFDIH) causes metabolic disorders such as insulin resistance, obesity, nonalcoholic fatty liver disease (NAFLD) and type 2 diabetes. The purpose of this study is to examine the effects and understand the mechanism of action of Lactobacillus rhamnosus GG culture supernatant (LGGs) on HFDIH-induced metabolic dysfunction. Mice were fed high-fat:high-fructose diet for 15 weeks. After 3 weeks of feeding, the mice were exposed to chronic intermittent hypoxia for the next 12 weeks (HFDIH), and LGGs was supplemented over the entire experiment. HFDIH exposure significantly led to metabolic disorders. LGGs treatment showed significant improvements in indices of metabolic disorders including fat mass, energy expenditure, glucose intolerance, insulin resistance, increased hepatic steatosis and liver injury. HFDIH mice markedly increased adipose inflammation and adipocyte size, and reduced circulating adiponectin, which was restored by LGGs treatment. LGGs treatment increased hepatic FGF21 mRNA expression and circulating FGF21 protein levels, which were associated with increased hepatic PPARα expression and fecal butyrate concentration. In addition, HFDIH-induced hepatic fat accumulation and apoptosis were significantly reduced by LGGs supplementation. In summary, LGGs treatment increased energy expenditure and insulin sensitivity and prevented metabolic abnormalities in HFDIH mice, and this is associated with the FGF21-adiponectin signaling pathway. LGGs may be a potential prevention/treatment strategy in subjects with the metabolic syndrome.