Single cell resolution of SARS-CoV-2 tropism, antiviral responses, and susceptibility to therapies in primary human airway epithelium.

The human airway epithelium is the initial site of SARS-CoV-2 infection. We used flow cytometry and single cell RNA-sequencing to understand how the heterogeneity of this diverse cell population contributes to elements of viral tropism and pathogenesis, antiviral immunity, and treatment response to remdesivir. We found that, while a variety of epithelial cell types are susceptible to infection, ciliated cells are the predominant cell target of SARS-CoV-2. The host protease TMPRSS2 was required for infection of these cells. Importantly, remdesivir treatment effectively inhibited viral replication across cell types, and blunted hyperinflammatory responses. Induction of interferon responses within infected cells was rare and there was significant heterogeneity in the antiviral gene signatures, varying with the burden of infection in each cell. We also found that heavily infected secretory cells expressed abundant IL-6, a potential mediator of COVID-19 pathogenesis.

Subtraction multiphase CT angiography: A new technique for faster detection of intracranial arterial occlusions.

OBJECTIVE: To describe and evaluate a novel technical development to improve detection of intracranial vessel occlusions using multiphase CT angiography (MPCTA). MATERIALS AND METHODS: The institutional ethics committee approved the study. Fifty patients (30 consecutive distal (M2 or smaller) anterior circulation occlusions, ten M1 occlusions, ten cases without occlusion) presenting with suspected AIS who underwent MPCTA were included. Post-processing of MPCTA studies created "subtraction" and "delayed enhancement" (DE) datasets. Initially, non-contrast CT and MPCTA studies for each patient were evaluated. Readers' confidence, speed and sensitivity of detection of intracranial vessel occlusions were recorded. After an interval of at least 4 weeks, readers were provided with post-processed images and studies were re-evaluated. RESULTS: While the sensitivity of detection of intracranial vessel occlusions was equal for both conventional MPCTA and subMPCTA, the mean time taken to identify a vessel occlusion decreased by 64 % using subMPCTA (16 s vs. 45 s with conventional MPCTA) (p<0.001). In addition, confidence in interpretation improved (from 4.4 to 4.9) using subMPCTA (p<0.001). CONCLUSION: SubMPCTA is a novel technique that aids in identifying small intracranial vessel occlusions in the suspected AIS patient. SubMPCTA increases confidence in interpretation and reduces the time taken to detect intracranial vessel occlusions. KEY POINTS: • SubMPCTA processes MPCTA data to better demonstrate intracranial arterial occlusions. • SubMPCTA increases confidence and speed of interpretation of MPCTA studies. • SubMPCTA may aid in rapidly differentiating acute ischaemic stroke from stroke mimics.

Autonomic alterations as a clinical manifestation of encephalopathy associated with autoimmune thyroid disease.

Encephalopathy associated with autoimmune thyroid disease (EAATD), also known as Hashimoto's encephalopathy, is a rare neurological condition that may occur in patients with clinical or sub-clinical autoimmune thyroid disease. The pathogenesis of EAATD has been not clearly elucidated yet. The diagnostic criteria include neurological or psychiatric symptoms, high levels of anti-thyroid antibodies, and exclusion of other possible causes of encephalopathy. In the large majority of cases, EAATD patients respond to immunosuppressant therapies, in particular to corticosteroids. We report the case of a patient with Hashimoto's thyroiditis and recurrent manifestations of encephalopathy over the previous few years responding to corticosteroid treatment. The patient presented with language and cognitive impairment, ataxia, and neurovegetative/autonomic symptoms. She was euthyroid with mildly raised anti-thyroid peroxidase antibodies. An extensive diagnostic work-up, including electroencephalogram, brain magnetic resonance, hormonal assessment, and an exhaustive panel of antibodies possibly associated with autoimmune encephalopathy, was carried out and excluded other possible etiologies of encephalopathy. The diagnosis of EAATD possibly affecting the hypothalamus and/or the neurovegetative regulatory centers was made and treatment with prednisolone was timely commenced with a dramatic and rapid improvement with progressive normalization of the symptoms. To the best of our knowledge, this is the first report of neurovegetative/autonomic alterations in the setting of EAATD.

Challenges and Potential of Artificial Intelligence in Neuroradiology.

PURPOSE: Artificial intelligence (AI) has emerged as a transformative force in medical research and is garnering increased attention in the public consciousness. This represents a critical time period in which medical researchers, healthcare providers, insurers, regulatory agencies, and patients are all developing and shaping their beliefs and policies regarding the use of AI in the healthcare sector. The successful deployment of AI will require support from all these groups. This commentary proposes that widespread support for medical AI must be driven by clear and transparent scientific reporting, beginning at the earliest stages of scientific research. METHODS: A review of relevant guidelines and literature describing how scientific reporting plays a central role at key stages in the life cycle of an AI software product was conducted. To contextualize this principle within a specific medical domain, we discuss the current state of predictive tissue outcome modeling in acute ischemic stroke and the unique challenges presented therein. RESULTS AND CONCLUSION: Translating AI methods from the research to the clinical domain is complicated by challenges related to model design and validation studies, medical product regulations, and healthcare providers' reservations regarding AI's efficacy and affordability. However, each of these limitations is also an opportunity for high-impact research that will help to accelerate the clinical adoption of state-of-the-art medical AI. In all cases, establishing and adhering to appropriate reporting standards is an important responsibility that is shared by all of the parties involved in the life cycle of a prospective AI software product.

A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm.

Distributed learning is a promising alternative to central learning for machine learning (ML) model training, overcoming data-sharing problems in healthcare. Previous studies exploring federated learning (FL) or the traveling model (TM) setup for medical image-based disease classification often relied on large databases with a limited number of centers or simulated artificial centers, raising doubts about real-world applicability. This study develops and evaluates a convolution neural network (CNN) for Parkinson's disease classification using data acquired by 83 diverse real centers around the world, mostly contributing small training samples. Our approach specifically makes use of the TM setup, which has proven effective in scenarios with limited data availability but has never been used for image-based disease classification. Our findings reveal that TM is effective for training CNN models, even in complex real-world scenarios with variable data distributions. After sufficient training cycles, the TM-trained CNN matches or slightly surpasses the performance of the centrally trained counterpart (AUROC of 83% vs. 80%). Our study highlights, for the first time, the effectiveness of TM in 3D medical image classification, especially in scenarios with limited training samples and heterogeneous distributed data. These insights are relevant for situations where ML models are supposed to be trained using data from small or remote medical centers, and rare diseases with sparse cases. The simplicity of this approach enables a broad application to many deep learning tasks, enhancing its clinical utility across various contexts and medical facilities.

Identifying Biases in a Multicenter MRI Database for Parkinson's Disease Classification: Is the Disease Classifier a Secret Site Classifier?

Sharing multicenter imaging datasets can be advantageous to increase data diversity and size but may lead to spurious correlations between site-related biological and non-biological image features and target labels, which machine learning (ML) models may exploit as shortcuts. To date, studies analyzing how and if deep learning models may use such effects as a shortcut are scarce. Thus, the aim of this work was to investigate if site-related effects are encoded in the feature space of an established deep learning model designed for Parkinson's disease (PD) classification based on T1-weighted MRI datasets. Therefore, all layers of the PD classifier were frozen, except for the last layer of the network, which was replaced by a linear layer that was exclusively re-trained to predict three potential bias types (biological sex, scanner type, and originating site). Our findings based on a large database consisting of 1880 MRI scans collected across 41 centers show that the feature space of the established PD model (74% accuracy) can be used to classify sex (75% accuracy), scanner type (79% accuracy), and site location (71% accuracy) with high accuracies despite this information never being explicitly provided to the PD model during original training. Overall, the results of this study suggest that trained image-based classifiers may use unwanted shortcuts that are not meaningful for the actual clinical task at hand. This finding may explain why many image-based deep learning models do not perform well when applied to data from centers not contributing to the training set.

Towards objective and systematic evaluation of bias in artificial intelligence for medical imaging.

OBJECTIVE: Artificial intelligence (AI) models trained using medical images for clinical tasks often exhibit bias in the form of subgroup performance disparities. However, since not all sources of bias in real-world medical imaging data are easily identifiable, it is challenging to comprehensively assess their impacts. In this article, we introduce an analysis framework for systematically and objectively investigating the impact of biases in medical images on AI models. MATERIALS AND METHODS: Our framework utilizes synthetic neuroimages with known disease effects and sources of bias. We evaluated the impact of bias effects and the efficacy of 3 bias mitigation strategies in counterfactual data scenarios on a convolutional neural network (CNN) classifier. RESULTS: The analysis revealed that training a CNN model on the datasets containing bias effects resulted in expected subgroup performance disparities. Moreover, reweighing was the most successful bias mitigation strategy for this setup. Finally, we demonstrated that explainable AI methods can aid in investigating the manifestation of bias in the model using this framework. DISCUSSION: The value of this framework is showcased in our findings on the impact of bias scenarios and efficacy of bias mitigation in a deep learning model pipeline. This systematic analysis can be easily expanded to conduct further controlled in silico trials in other investigations of bias in medical imaging AI. CONCLUSION: Our novel methodology for objectively studying bias in medical imaging AI can help support the development of clinical decision-support tools that are robust and responsible.

Detection of pathogenic bacteria using a homogeneous immunoassay based on shear alignment of virus particles and linear dichroism.

Biomolecular detection has for a long time depended on a relatively small number of established methodologies. Many of these depend on the detection of a ligand-antibody complex using some kind of optical technique, e.g., chemiluminescence. Before this measurement can be made, the ligand-antibody complex generally has to be separated from bulk contaminants. This process involves the implementation of a heterogeneous assay format involving immobilization of the antibody onto a solid support to enable washing of the unbound ligand. This approach has a number of inherent issues including being slow and complex and requiring the use of expensive reagents. In this paper, we demonstrate how we have harnessed a biologically inspired nanoparticle to provide the basis for a homogeneous assay which requires no immobilization. The method relies on using fluid shear flow to align a fiber-like nanoparticle formed from a filamentous virus, M13, combined with a ligand-specific antibody. This renders the particle visible to linear dichroic spectroscopy, which provides an easily interpretable signal. The addition of the target ligand (in this case Escherichia coli O157) leads to the formation of a nanoparticle-ligand particle that is unable to align, leading to the perturbation of the linear dichroism signal.

Fairness-related performance and explainability effects in deep learning models for brain image analysis.

Purpose: Explainability and fairness are two key factors for the effective and ethical clinical implementation of deep learning-based machine learning models in healthcare settings. However, there has been limited work on investigating how unfair performance manifests in explainable artificial intelligence (XAI) methods, and how XAI can be used to investigate potential reasons for unfairness. Thus, the aim of this work was to analyze the effects of previously established sociodemographic-related confounders on classifier performance and explainability methods. Approach: A convolutional neural network (CNN) was trained to predict biological sex from T1-weighted brain MRI datasets of 4547 9- to 10-year-old adolescents from the Adolescent Brain Cognitive Development study. Performance disparities of the trained CNN between White and Black subjects were analyzed and saliency maps were generated for each subgroup at the intersection of sex and race. Results: The classification model demonstrated a significant difference in the percentage of correctly classified White male ( 90.3 % ± 1.7 % ) and Black male ( 81.1 % ± 4.5 % ) children. Conversely, slightly higher performance was found for Black female ( 89.3 % ± 4.8 % ) compared with White female ( 86.5 % ± 2.0 % ) children. Saliency maps showed subgroup-specific differences, corresponding to brain regions previously associated with pubertal development. In line with this finding, average pubertal development scores of subjects used in this study were significantly different between Black and White females ( p < 0.001 ) and males ( p < 0.001 ). Conclusions: We demonstrate that a CNN with significantly different sex classification performance between Black and White adolescents can identify different important brain regions when comparing subgroup saliency maps. Importance scores vary substantially between subgroups within brain structures associated with pubertal development, a race-associated confounder for predicting sex. We illustrate that unfair models can produce different XAI results between subgroups and that these results may explain potential reasons for biased performance.

Study of robustness of filamentous bacteriophages for industrial applications.

The development of a whole new class of industrial agents, such as biologically based nanomaterials and viral vectors, has raised many challenges for their large-scale manufacture, principally due to the lack of essential physical data and bioprocessing knowledge. A new example is the promise of filamentous bacteriophages and their derivatives. As a result, there is now an increasing need for the establishment of strong biochemical engineering foundations to serve as a guide for future manufacture. This article investigates the effect of high-energy fluid flow on filamentous bacteriophage M13 to determine its robustness for large-scale processing. By the application of well-understood ultra scale-down predictive techniques, the viability of bacteriophage M13 was studied as a measure of its robustness and as a function of energy dissipation rate and fluid conditions. These experiments suggested that despite being perceived as a relatively fragile molecule in the literature, bacteriophage M13 should tolerate processing conditions in existing large-scale equipment designs. No loss of viability was noted up to a maximum energy dissipation rate of 2.9 × 10(6) W kg(-1) . Furthermore, significant losses above this threshold only occurred over periods well in excess of the exposure times expected in a bioprocess environment. Filamentous bacteriophages may therefore be regarded as a viable process material for industrial applications.

Readability of patient education materials related to radiation safety: What are the implications for patient-centred radiology care?

BACKGROUND: Increasing numbers of patients and carers rely on online resources for healthcare information. Radiation safety can be misunderstood by patients and clinicians and lead to patient anxiety. We aimed to assess the readability of online patient educational materials (PEMs) related to radiation safety. METHODS: A total of 84 articles pertaining to radiation safety from 14 well-known online resources were identified. PEMs were then analysed using Readability Studio Professional Edition Version 2019. Readability was assessed using eight different instruments: the Flesch-Kincaid Reading Grade Level, Raygor Estimate, SMOG, Coleman-Liau, Fry, FORCAST, Gunning Fog, and Flesch Reading Ease Score formula. The mean reading grade level (RGL) of each article was compared to the 6th and 8th grade reading level using 1-sample t-tests. RESULTS: The cumulative mean RGL for all 84 articles was 13.3 (range = 8.6-17.4), and none were written at or below the 6th or 8th grade level. The cumulative mean RGL exceeded the 6th grade reading level by an average of 7.3 levels (95% CI, 6.8-7.8; p < 0.001) and the 8th grade level by an average of 5.3 grade levels (95% CI, 4.8-5.8; p < 0.001). The mean Flesch Reading Ease Score was 39/100 ('difficult'). CONCLUSION: Currently available online PEMs related to radiation safety are still written at higher than recommended reading levels. Radiation safety is a topic in which the specialist training of radiologists is crucial in providing guidance to patients. Addressing the readability of online PEMs can improve radiology-patient communication and support the shift to a patient-centred model of practice.

The needle and the damage done: musculoskeletal and vascular complications associated with injected drug use.

Injected drug use is associated with a wide range of medical complications which are predominantly musculoskeletal and vascular in nature. Illicit drug use is increasing worldwide. Patients with complications of injected drug use often present in a non-specific manner without a reliable clinical history. Musculoskeletal complications are typically infective in aetiology and may vary widely in severity from mild to life-threatening. A multimodal imaging approach is often required for both diagnostic imaging and image-guided sampling. Plain radiographs are often an important initial test, for example in identifying retained needles from injection. Ultrasound and CT play important roles in the assessment of complex soft tissue complications and MRI is the imaging modality of choice for bone and joint disorders. Vascular complications may be venous or arterial in nature and usually occur locally at the injection site. These complications may be related to direct injury to the vessel wall by a needle, or secondary to local infection and inflammation. A multimodal imaging strategy is also often required in the assessment of these vascular complications, typically involving a combination of ultrasound and CT. Familiarity with the multimodal imaging features of the complications related to injected drug use is crucially important as they may be rapidly progressive and life-threatening and require timely diagnosis.

Single cell resolution of SARS-CoV-2 tropism, antiviral responses, and susceptibility to therapies in primary human airway epithelium.

The human airway epithelium is the initial site of SARS-CoV-2 infection. We used flow cytometry and single cell RNA-sequencing to understand how the heterogeneity of this diverse cell population contributes to elements of viral tropism and pathogenesis, antiviral immunity, and treatment response to remdesivir. We found that, while a variety of epithelial cell types are susceptible to infection, ciliated cells are the predominant cell target of SARS-CoV-2. The host protease TMPRSS2 was required for infection of these cells. Importantly, remdesivir treatment effectively inhibited viral replication across cell types, and blunted hyperinflammatory responses. Induction of interferon responses within infected cells was rare and there was significant heterogeneity in the antiviral gene signatures, varying with the burden of infection in each cell. We also found that heavily infected secretory cells expressed abundant IL-6, a potential mediator of COVID-19 pathogenesis.

Quantitative evaluation of the expression and activity of five major sulfotransferases (SULTs) in human tissues: the SULT "pie".

Expression levels of the major human sulfotransferases (SULTs) involved in xenobiotic detoxification in a range of human tissues (i.e., SULT "pies") are not available in a form allowing comparison between tissues and individuals. Here we have determined, by quantitative immunoblotting, expression levels for the five principal human SULTs-SULT1A1, SULT1A3/4, SULT1B1, SULT1E1, and SULT2A1-and determined the kinetic properties toward probe substrates, where available, for these enzymes in cytosol samples from a bank of adult human liver, small intestine, kidney, and lung. We produced new isoform-selective antibodies against SULT1B1 and SULT2A1, which were used alongside antibodies against SULT1A3 and SULT1A1 previously produced in our laboratory or available commercially (SULT1E1). Expression levels were derived using purified recombinant enzymes to construct standard curves for each individual isoform and immunoblot. Substantial intertissue and interindividual differences in expression were observed. SULT1A1 was the major enzyme (>50% of total, range 420-4900 ng/mg cytosol protein) in the liver, followed by SULT2A1, SULT1B1, and SULT1E1. SULT1A3 was completely absent from this tissue. In contrast, the small intestine contained the largest overall amount of SULT of any of the tissues, with SULT1B1 the major enzyme (36%), closely followed by SULT1A3 (31%), and SULT1A1, SULT1E1, and SULT2A1 more minor forms (19, 8, and 6% of total, respectively). The kidney and lung contained low levels of SULT. We provide a unique data set that will add value to the study of the role and contribution of sulfation to drug and xenobiotic metabolism in humans.

Relationship between cobalamin and folate deficiencies and anemia in dogs.

BACKGROUND: Megaloblastic, nonregenerative anemia is a well-known consequence of cobalamin or folate deficiencies in humans but is not recognized in hypocobalaminemic or hypofolatemic dogs. Establishment of relationships between hypocobalaminemia or hypofolatemia and hematologic disease would encourage vitamin B testing, and potentially supplementation, in anemic dogs. OBJECTIVES: To determine the prevalence of anemia in hypocobalaminemic or hypofolatemic dogs and to report the prevalence of hypocobalaminemia and hypofolatemia and nonregenerative anemia, macrocytosis, and anisocytosis in anemic dogs. ANIMALS: One hundred and fourteen client-owned dogs with known serum cobalamin and folate concentrations and CBCs and 42 client-owned anemic dogs. METHODS: Retrospective comparison of anemia prevalence in hypocobalaminemic or hypofolatemic and normocobalaminemic or normofolatemic dogs was performed. Prospective measurement of erythrocyte variables and cobalamin and folate concentrations in anemic dogs was carried out; relationships among hypocobalaminemia and regenerative status, mean corpuscular volume, and red cell distribution width were evaluated. RESULTS: Significant differences in prevalence of anemia between hypocobalaminemic (36%) and normocobalaminemic dogs (26%; P = .23) or between hypofolatemic (31%) and normofolatemic dogs (30%; P = .99) were not detected. Between hypocobalaminemic and normocobalaminemic dogs, no significant differences in prevalence of nonregenerative anemia (69% vs 63%; P = .65), macrocytosis (17% vs 0%; P = .53), or anisocytosis (28% vs 0%; P = .14) were detected. Anemic dogs had high prevalence of vitamin B deficiencies (nonregenerative: 64% hypocobalaminemic, 18% hypofolatemic; regenerative: 57% hypocobalaminemic, 21% hypofolatemic). CONCLUSIONS AND CLINICAL IMPORTANCE: The association between cobalamin and folate deficiencies and macrocytic, nonregenerative anemia established in humans is not routinely present in dogs.

MR imaging of spinal haematoma: a pictorial review.

Spinal haematoma is a rare occurrence, however has the potential to cause significant spinal injury and morbidity. MRI is the gold-standard of investigation, and urgent imaging is required for early diagnosis and treatment to ensure best patient outcomes. We present a pictorial review demonstrating the imaging features of spinal haematoma based on meningeal space assignment; epidural, subdural, subarachnoid, intramedullary and a combination of these locations. In this review, we summarise the literature and imaging findings of spinal haematoma on MRI. Particular imaging features which help to differentiate between haematoma in the different spinal meningeal compartments are discussed below.

Acetabular paralabral cyst causing compression of the sciatic nerve.

Acetabular paralabral cysts are common. They vary in their clinical presentation and may be asymptomatic or cause pain and restriction at the hip joint. In rare instances they may cause symptoms by compressing local neurovascular structures. We report a case of symptomatic compression of the sciatic nerve by a posteriorly displaced acetabular paralabral cyst.

Advanced multimodality imaging of inflammatory bowel disease in 2015: An update.

The diagnosis and effective management of inflammatory bowel disease (IBD) requires a combination clinical, endoscopic, histological, biological, and imaging data. While endoscopy and biopsy remains the gold standard for diagnosis of IBD, imaging plays a central role in the assessment of extra mural disease, in disease surveillance and in the assessment of response to medical treatments, which are often expensive. Imaging is also vital in the detection and diagnosis of disease related complications, both acute and chronic. In this review, we will describe, with illustrative images, the imaging features of IBD in adults, with emphasis on up-to-date imaging techniques focusing predominantly on cross sectional imaging and new magnetic resonance imaging techniques.

Impact of sensory design interventions on image quality, patient anxiety and overall patient experience at MRI.

OBJECTIVE: This study aimed to assess the effect of sensory stimulation on patient MRI experience and to assess whether sensory stimulation has a significant effect on MR image quality. METHODS: A case-control study was conducted over 4 months, involving patients undergoing MRI brain, cervical spine, breast and prostate. The study involved 106 patients, 64 cases and 42 controls. Cases underwent sensory stimulation during the scan in the form of a scented cotton pad placed in the scanner near their head and/or calming bird noises were played over headphones. Post-scan, participants completed a questionnaire regarding their experience of MRI. Scanning radiographers completed a questionnaire regarding patient tolerance of the scan. All studies were evaluated by two radiologists for the presence of movement artefact. RESULTS: 39% of cases and 38% of controls reported anxiety in the days preceding MRI. 6.2% of cases required coaching during image acquisition, while 9.7% of controls required coaching. 4.7% of cases and 4.8% controls required sequence repetition owing to movement artefact. Mean patient experience score (as graded by the patient) for controls was 1.74 ± 0.63 standard deviation (SD) and for cases, it was 1.67 ± 0.60 SD. (Lower assigned scores equated to a better experience). Mean patient experience score based on comments on a 5-point scale as graded by two observers was 2.81 ± 0.70 SD for controls, 2.42 + 0.94 SD for sound intervention and 2.46 ± 1.01 SD for scent intervention. Mean motion artefact score graded by the two radiologists was 1.13 ± 0.53 SD for controls and 1.08 ± 0.36 SD for cases. (A lower score equated to less movement artefact). We demonstrated a trend towards a relaxing experience in those patients undergoing MRI for the first time who underwent sensory intervention. Participant positive ratings of the smell pleasantness were associated with a reduced likelihood of experiencing anxiety (p = 0.13). Results were not statistically significant. CONCLUSION: This study demonstrated a subjectively improved experience of MRI for some patients with the intervention of sound and smell. The study failed to show a significant decrease in patient movement during MRI investigations. This may relate to the small study size and a low level of patient movement in the case group. Advances in knowledge: Sound and olfactory sensory environment interventions at MRI can improve the patient experience. These low-cost interventions are well tolerated, may improve acceptance of MRI in patients with anxiety and offer a competitive advantage to imaging centres.