Search

The braincase features appear phylogenetically informative and key in assessing the still contentious relationships within Crocodylomorpha. Yet, the neurocrania of many non-crocodylian taxa are imperfectly studied. In the present paper, we describe the braincase osteology and neuroanatomy of a paralligatorid crocodylomorph Kansajsuchus extensus from the Upper Cretaceous Kansai locality in Tadzhikistan based on CT-scanning, segmentation, and 3D modeling. The detailed comparison of K. extensus with other taxa expands our understanding of the neurocranial anatomy and evolution in Neosuchia. The braincase anatomy appears to be relatively consistent across most non-crocodylian neosuchians, but important changes in its structure are traced through the neosuchian–eusuchian transition and especially at the base of Crocodylia. We question the close affinities of Dyrosauridae and Pholidosauridae based on significant differences in their braincases. • Neurocranial anatomy is described for Kansajsuchus extensus from the Upper Cretaceous Kansai locality in Tadzhikistan. • It is a first detailed study of neuroanatomy in Paralligatoridae. • Comparison of neurocranial osteology allowed to trace evolutionary modifications in the braincase of Neosuchia. • Close affinities of Dyrosauridae and Pholidosauridae are questioned based on significant differences in braincase structure. [ABSTRACT FROM AUTHOR]

Opportunistic infections in people living with human immunodeficiency virus (HIV) are readily detected with thoracic computed tomography (CT), but differential diagnosis remains a challenge. The global COVID-19 pandemic further exacerbates the issue, with SARS-CoV-2 having overlapping CT findings with infections common in HIV patients and complicating prior epidemiological data. We present a pictorial review of CT findings associated with COVID-19-mimicking opportunistic infections that can be encountered in HIV patients. PubMed database was searched for the complete list of relevant conditions, and a Venn diagram was constructed to highlight overlapping entities. The diagram showed five major disease groups: viral pneumonia, fungal pneumonia, bacterial pneumonia, sarcoidosis, and lung cancer. As these pathologies possess a wide range of features, the findings were grouped as "typical" and "other" for easier comprehension with provided relevant epidemiological data and discrepancies observed in available literature. The review highlights the importance of a specific approach to differential diagnosis in immunocompromised patients compared to immunocompetent hosts and the utility of follow-up scans. • CT DDx in PLHIV is a challenge • Five major disease groups with overlapping CT features • The review provides radiologists data that may assist the diagnostic decision-making [ABSTRACT FROM AUTHOR]

• Usage of most spatially detailed feature maps boosts classification quality within the proposed multitask framework. • A deep convolutional network for triage of COVID-19 patients based on identification and severity quantification scores. • A carefully selected independent external test set with other lung pathologies. • Publicly released annotated data and code to ensure reproducibility. [Display omitted] The current COVID-19 pandemic overloads healthcare systems, including radiology departments. Though several deep learning approaches were developed to assist in CT analysis, nobody considered study triage directly as a computer science problem. We describe two basic setups: Identification of COVID-19 to prioritize studies of potentially infected patients to isolate them as early as possible; Severity quantification to highlight patients with severe COVID-19, thus direct them to a hospital or provide emergency medical care. We formalize these tasks as binary classification and estimation of affected lung percentage. Though similar problems were well-studied separately, we show that existing methods could provide reasonable quality only for one of these setups. We employ a multitask approach to consolidate both triage approaches and propose a convolutional neural network to leverage all available labels within a single model. In contrast with the related multitask approaches, we show the benefit from applying the classification layers to the most spatially detailed feature map at the upper part of U-Net instead of the less detailed latent representation at the bottom. We train our model on approximately 1500 publicly available CT studies and test it on the holdout dataset that consists of 123 chest CT studies of patients drawn from the same healthcare system, specifically 32 COVID-19 and 30 bacterial pneumonia cases, 30 cases with cancerous nodules, and 31 healthy controls. The proposed multitask model outperforms the other approaches and achieves ROC AUC scores of 0.87 ± 0.01 vs. bacterial pneumonia, 0.93 ± 0.01 vs. cancerous nodules, and 0.97 ± 0.01 vs. healthy controls in Identification of COVID-19, and achieves 0.97 ± 0.01 Spearman Correlation in Severity quantification. We have released our code and shared the annotated lesions masks for 32 CT images of patients with COVID-19 from the test dataset. [ABSTRACT FROM AUTHOR]