Artificial intelligence applied to magnetic resonance imaging reliably detects the presence, but not the location, of meniscus tears: a systematic review and meta-analysis.

OBJECTIVES: To review and compare the accuracy of convolutional neural networks (CNN) for the diagnosis of meniscal tears in the current literature and analyze the decision-making processes utilized by these CNN algorithms. MATERIALS AND METHODS: PubMed, MEDLINE, EMBASE, and Cochrane databases up to December 2022 were searched in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement. Risk of analysis was used for all identified articles. Predictive performance values, including sensitivity and specificity, were extracted for quantitative analysis. The meta-analysis was divided between AI prediction models identifying the presence of meniscus tears and the location of meniscus tears. RESULTS: Eleven articles were included in the final review, with a total of 13,467 patients and 57,551 images. Heterogeneity was statistically significantly large for the sensitivity of the tear identification analysis (I2 = 79%). A higher level of accuracy was observed in identifying the presence of a meniscal tear over locating tears in specific regions of the meniscus (AUC, 0.939 vs 0.905). Pooled sensitivity and specificity were 0.87 (95% confidence interval (CI) 0.80-0.91) and 0.89 (95% CI 0.83-0.93) for meniscus tear identification and 0.88 (95% CI 0.82-0.91) and 0.84 (95% CI 0.81-0.85) for locating the tears. CONCLUSIONS: AI prediction models achieved favorable performance in the diagnosis, but not location, of meniscus tears. Further studies on the clinical utilities of deep learning should include standardized reporting, external validation, and full reports of the predictive performances of these models, with a view to localizing tears more accurately. CLINICAL RELEVANCE STATEMENT: Meniscus tears are hard to diagnose in the knee magnetic resonance images. AI prediction models may play an important role in improving the diagnostic accuracy of clinicians and radiologists. KEY POINTS: • Artificial intelligence (AI) provides great potential in improving the diagnosis of meniscus tears. • The pooled diagnostic performance for artificial intelligence (AI) in identifying meniscus tears was better (sensitivity 87%, specificity 89%) than locating the tears (sensitivity 88%, specificity 84%). • AI is good at confirming the diagnosis of meniscus tears, but future work is required to guide the management of the disease.

Intrapulmonary shunting is a key contributor to hypoxia in COVID-19: An update on the pathophysiology.

BACKGROUND: The pathophysiology of COVID-19 remains poorly understood. We aimed to estimate the contribution of intrapulmonary shunting and ventilation-to-perfusion (VA/Q) mismatch using a mathematical model to construct oxygen-haemoglobin dissociation curves (ODCs). METHODS: ODCs were constructed using transcutaneous pulse oximetry at two different fractions of inspired oxygen (FiO2). 199 patients were included from two large district general hospitals in the South East of England from 1st to 14th January 2021. The study was supported by the National Institute of Health Research (NIHR) Clinical Research Network. RESULTS: Overall mortality was 29%. Mean age was 68.2 years (SEM 1·2) with 46% female. Median shunt on admission was 17% (IQR 8-24.5); VA/Q was 0.61 (IQR 0.52-0.73). Shunt was 37.5% higher in deaths (median 22%, IQR 9-29) compared to survivors (16%, 8-21; p = 0.0088) and was a predictor of mortality (OR 1.04; 95% CI 1.01-1.07). Admission oxygen saturations were more strongly predictive of mortality (OR 0.91, 95% CI 0.87-0.96). There was no difference in VA/Q mismatch between deaths (0.60; IQR 0.50-0.73) and survivors (0.61; IQR 0.52-0.73; p = 0.63) and it was not predictive of mortality (OR 0.68; 95% CI 0.18-2.52; p = 0.55). Shunt negatively correlated with admission oxygen saturation (R -0.533; p<0.0001) whereas VA/Q was not (R 0.1137; p = 0.12). INTERPRETATION: Shunt, not VA/Q mismatch, was associated with worsening hypoxia, though calculating shunt was not of prognostic value. This study adds to our understanding of the pathophysiology of hypoxaemia in COVID-19. Our inexpensive and reliable technique may provide further insights into the pathophysiology of hypoxia in other respiratory diseases.

A rare cause of acute liver failure due to haemophagocytic lymphohistiocytosis secondary to diffuse large B-cell lymphoma.

Acute liver failure is a life-threatening condition commonly caused by drug-induced hepatotoxicity or viral hepatitides. However, there are a number of rarer causes such as haemophagocytic lymphohistiocytosis. Haemophagocytic lymphohistiocytosis is a syndrome of uncontrolled immune cell activation, triggered by infection or malignancy, which carries a high mortality. Whilst mild to moderate liver injury is commonly seen with haemophagocytic lymphohistiocytosis, acute liver failure has rarely been reported in adults. We present a case of a 74-year-old man with acute liver failure secondary to haemophagocytic lymphohistiocytosis triggered by undiagnosed large B-cell lymphoma. Initially treated for biliary sepsis, there was a delay in the diagnosis of haemophagocytic lymphohistiocytosis and despite initiating chemotherapy, he died soon after. This case highlights the importance of considering haemophagocytic lymphohistiocytosis as a rare cause of acute liver failure, as given the life-threatening potential of haemophagocytic lymphohistiocytosis, a prompt diagnosis may allow early initiation of chemotherapy for any chance of survival.

Plasmon enhanced water splitting mediated by hybrid bimetallic Au-Ag core-shell nanostructures.

In this work, we employed wet chemically synthesized bimetallic Au-Ag core-shell nanostructures (Au-AgNSs) to enhance the photocurrent density of mesoporous TiO2 for water splitting and we compared the results with monometallic Au nanoparticles (AuNPs). While Au-AgNSs incorporated photoanodes give rise to 14× enhancement in incident photon to charge carrier efficiency, AuNPs embedded photoanodes result in 6× enhancement. By varying nanoparticle concentration in the photoanodes, we observed ∼245× less Au-AgNSs are required relative to AuNPs to generate similar photocurrent enhancement for solar fuel conversion. Power-dependent measurements of Au-AgNSs and AuNPs showed a first order dependence to incident light intensity, relative to half-order dependence for TiO2 only photoanodes. This indicated that plasmonic nanostructures enhance charge carriers formed on the surface of the TiO2 which effectively participate in photochemical reactions. Our experiments and simulations suggest the enhanced near-field, far-field, and multipolar resonances of Au-AgNSs facilitating broadband absorption of solar radiation collectively gives rise to their superior performance in water splitting.