COVID-19 Hierarchical Classification Using a Deep Learning Multi-Modal.

Coronavirus disease 2019 (COVID-19), originating in China, has rapidly spread worldwide. Physicians must examine infected patients and make timely decisions to isolate them. However, completing these processes is difficult due to limited time and availability of expert radiologists, as well as limitations of the reverse-transcription polymerase chain reaction (RT-PCR) method. Deep learning, a sophisticated machine learning technique, leverages radiological imaging modalities for disease diagnosis and image classification tasks. Previous research on COVID-19 classification has encountered several limitations, including binary classification methods, single-feature modalities, small public datasets, and reliance on CT diagnostic processes. Additionally, studies have often utilized a flat structure, disregarding the hierarchical structure of pneumonia classification. This study aims to overcome these limitations by identifying pneumonia caused by COVID-19, distinguishing it from other types of pneumonia and healthy lungs using chest X-ray (CXR) images and related tabular medical data, and demonstrate the value of incorporating tabular medical data in achieving more accurate diagnoses. Resnet-based and VGG-based pre-trained convolutional neural network (CNN) models were employed to extract features, which were then combined using early fusion for the classification of eight distinct classes. We leveraged the hierarchal structure of pneumonia classification within our approach to achieve improved classification outcomes. Since an imbalanced dataset is common in this field, a variety of versions of generative adversarial networks (GANs) were used to generate synthetic data. The proposed approach tested in our private datasets of 4523 patients achieved a macro-avg F1-score of 95.9% and an F1-score of 87.5% for COVID-19 identification using a Resnet-based structure. In conclusion, in this study, we were able to create an accurate deep learning multi-modal to diagnose COVID-19 and differentiate it from other kinds of pneumonia and normal lungs, which will enhance the radiological diagnostic process.

Harnessing Machine Learning in Early COVID-19 Detection and Prognosis: A Comprehensive Systematic Review.

During the early phase of the COVID-19 pandemic, reverse transcriptase-polymerase chain reaction (RT-PCR) testing faced limitations, prompting the exploration of machine learning (ML) alternatives for diagnosis and prognosis. Providing a comprehensive appraisal of such decision support systems and their use in COVID-19 management can aid the medical community in making informed decisions during the risk assessment of their patients, especially in low-resource settings. Therefore, the objective of this study was to systematically review the studies that predicted the diagnosis of COVID-19 or the severity of the disease using ML. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA), we conducted a literature search of MEDLINE (OVID), Scopus, EMBASE, and IEEE Xplore from January 1 to June 31, 2020. The outcomes were COVID-19 diagnosis or prognostic measures such as death, need for mechanical ventilation, admission, and acute respiratory distress syndrome. We included peer-reviewed observational studies, clinical trials, research letters, case series, and reports. We extracted data about the study's country, setting, sample size, data source, dataset, diagnostic or prognostic outcomes, prediction measures, type of ML model, and measures of diagnostic accuracy. Bias was assessed using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). This study was registered in the International Prospective Register of Systematic Reviews (PROSPERO), with the number CRD42020197109. The final records included for data extraction were 66. Forty-three (64%) studies used secondary data. The majority of studies were from Chinese authors (30%). Most of the literature (79%) relied on chest imaging for prediction, while the remainder used various laboratory indicators, including hematological, biochemical, and immunological markers. Thirteen studies explored predicting COVID-19 severity, while the rest predicted diagnosis. Seventy percent of the articles used deep learning models, while 30% used traditional ML algorithms. Most studies reported high sensitivity, specificity, and accuracy for the ML models (exceeding 90%). The overall concern about the risk of bias was "unclear" in 56% of the studies. This was mainly due to concerns about selection bias. ML may help identify COVID-19 patients in the early phase of the pandemic, particularly in the context of chest imaging. Although these studies reflect that these ML models exhibit high accuracy, the novelty of these models and the biases in dataset selection make using them as a replacement for the clinicians' cognitive decision-making questionable. Continued research is needed to enhance the robustness and reliability of ML systems in COVID-19 diagnosis and prognosis.

Intracranial Volume in Relation to Parent Satisfaction and Expert Opinion: A Novel Look at Surgical Outcomes.

Factors like parent satisfaction and expert opinion have been proposed as outcome measures related to craniosynostosis (CS) surgery. However, there is no real tangible score for CS surgery outcomes. In our study, we aimed to explore different factors considered as a tangible outcome measure of CS surgery. Methods: A retrospective cohort study of 23 patients with CS who were operated on in a tertiary care university hospital. Parents were interviewed to assess their satisfaction of aesthetic outcome. This was correlated to two expert opinions and to the amount of skull expansion. Results: The mean follow-up duration was 2.24 ± 1.12 years. Twelve of the 23 fathers were satisfied, whereas 11 of the 23 mothers were satisfied. The overall combined satisfaction rate of both parents was on the higher side with no difference in between. There was a significant association between expansion rate of 7.65 ± 4.99% and the overall parent's satisfaction (P = 0.002). Additionally, there was a good correlation between both experts with statistically significant association (P = 0.004). No correlation was found between the parents' satisfaction and the experts' opinions. Conclusions: The study is valuable, as it investigates the relationship between the expansion rate, parents' satisfaction, and expert opinion as predicted values of craniosynostosis surgery. The overall satisfaction correlated significantly well with the expansion rate. However, such numerical assessment is not a real guide for assessing clinical outcomes' as no association was found between expansion rate, satisfaction rate, and expert opinion.

Effects of compressive lesions on intraoperative human spinal cord elasticity.

OBJECTIVE: Although evaluating tissue elasticity has various clinical applications, spinal cord elasticity (SCE) in humans has never been well documented. In this study, the authors aimed to evaluate the impact of compression on human SCE in vivo. METHODS: The authors prospectively assessed SCE using intraoperative shear wave elastography (SWE). All consecutive patients undergoing spinal cord (SC) decompression (laminectomy or corpectomy) between June 2018 and June 2019 were included. After intraoperative exposure of the patient's dura mater, at least three SWE measurements of the SC and its coverings were performed. Intraoperative neurological monitoring in the form of motor and somatosensory evoked potentials was utilized. Cases were divided into two groups based on the state of SC compression following bone removal (laminectomy or corpectomy): patients with adequate decompression (the decompressed SC group [DCG]) following bone removal and patients with remining compression, e.g., compressing tumor or instability (the compressed SC group [COG]). RESULTS: A total of 25 patients were included (8 females and 17 males) with a mean age of 48.28 ± 21.47 years. Most cases were degenerative diseases (10 cases) followed by tumors (6 cases), and the compression was observed at cervical (n = 14), thoracic (n = 9), and conus medullaris (n = 2) levels. The COG (6 cases) expressed significantly higher elasticity values, i.e., greater stiffness (median 93.84, IQR 75.27-121.75 kPa) than the decompressed SC in DCG (median 9.35, IQR 6.95-11.22 kPa, p < 0.001). Similarly, the compressed dura mater in the COG was significantly stiffer (mean ± SD 121.83 ± 70.63 kPa) than that in the DCG (29.78 ± 18.31 kPa, p = 0.042). Following SC decompression in COG, SCE values were significantly reduced (p = 0.006; adjusted for multiple comparisons). Intraoperative monitoring demonstrated no worsening from the baseline. CONCLUSIONS: The current study is to the authors' knowledge the first to quantitatively demonstrate increased stiffness (i.e., elasticity value) of the human SC and dura mater in response to external compression in vivo. It appears that SCE is a dynamic phenomenon and is reduced following decompression. Moreover, the evaluation of human SCE using the SWE technique is feasible and safe. Information from future studies aiming to further define SCE could be valuable in the early and accurate diagnosis of the compressed SC.

Association between craniovertebral junction abnormalities and syringomyelia in patients with chiari malformation type-1.

OBJECTIVE: To assess the correlation between craniovertebral junction (CVJ) abnormalities and syringomyelia in patients with Chiari malformation type-1 (CM1). METHODS: This was a retrospective study including patients with CM1. Identification of cases was done by searching a radiology database at a university hospital from 2012 to 2017. Patients were divided into 2 groups based on whether CVJ abnormalities were present (CVJ+) or absent (CVJ-). The patients` demographic and clinical data were reviewed. All magnetic resonance imaging studies were examined by a certified neuroradiologist. RESULTS: Sixty-four consecutive patients with CM1 were included. The mean age was 24+/-17 years; 59% were females. The CVJ+ group had more female patients (p=0.012). The most frequent CVJ abnormality was platybasia (71%), followed by short clivus (44%) and cervical kyphosis (33%). The CVJ abnormalities were more in Syringomyelia cases (p=0.045). However, the results were not significant when hydrocephalus cases were excluded. CONCLUSION: Among CM1 patients, CVJ abnormalities were found more in patients with syringomyelia. Future studies with larger sample size are required to further study the correlation between CVJ abnormalities and both syringomyelia and hydrocephalus in CM1 patients.

Automatic detection of the meningioma tumor firmness in MRI images.

Meningioma is among the most common primary tumors of the brain. The firmness of Meningioma is a critical factor that influences operative strategy and patient counseling. Conventional methods to predict the tumor firmness rely on the correlation between the consistency of Meningioma and their preoperative MRI findings such as the signal intensity ratio between the tumor and the normal grey matter of the brain. Machine learning techniques have not been investigated yet to address the Meningioma firmness detection problem. The main purpose of this research is to couple supervised learning algorithms with typical descriptors for developing a computer-aided detection (CAD) of the Meningioma tumor firmness in MRI images. Specifically, Local Binary Patterns (LBP), Gray Level Co-occurrence Matrix (GLCM) and Discrete Wavelet Transform (DWT) are extracted from real labeled MRI-T2 weighted images and fed into classifiers, namely support vector machine (SVM) and k-nearest neighbor (KNN) algorithm to learn association between the visual properties of the region of interest and the pre-defined firm and soft classes. The learned model is then used to classify unlabeled MRI-T2 weighted images. This paper represents a baseline comparison of different features used in CAD system that intends to accurately recognize the Meningioma tumor firmness. The proposed system was implemented and assessed using a clinical dataset. Using LBP feature yielded the best performance with 95% of F-score, 87% of balanced accuracy and 0.87 of the area under ROC curve (AUC) when coupled with KNN classifier, respectively.

In vivo assessment of spinal cord elasticity using shear wave ultrasound in dogs.

OBJECTIVE: Evaluation of living tissue elasticity has wide applications in disease characterization and prognosis prediction. Few previous ex vivo attempts have been made to characterize spinal cord elasticity (SCE). Recently, tissue elasticity assessment has been clinically feasible using ultrasound shear wave elastography (SWE). The current study aims to characterize SCE in healthy dogs, in vivo, utilizing SWE, and to address SCE changes during compression. METHODS: Ten Greyhound dogs (mean age 14 months; mean weight 14.3 kg) were anesthetized and tracheally intubated, with hemodynamic and neurological monitoring. A 3-level, midcervical laminectomy was performed. SCE was assessed at baseline. Next, 8- and 13-mm balloon compressions were sequentially applied ventral to the spinal cord. RESULTS: The mean SCE was 18.5 ± 7 kPa. Elasticity of the central canal, pia mater, and dura mater were 21.7 ± 9.6 kPa, 26.1 ± 14.8 kPa, and 63.2 ± 11.5 kPa, respectively. As expected, the spinal cord demonstrated less elasticity than the dura mater (p < 0.0001) and pia mater (trend toward significance p = 0.08). Notably, the 13-mm balloon compression resulted in a stiffer spinal cord than at baseline (233 ± 73 kPa versus 18.5 ± 7 kPa, p < 0.0001) and 8-mm balloon compression (233 ± 73 kPa versus 185 ± 68 kPa, p < 0.048). CONCLUSIONS: In vivo SCE evaluation using SWE is feasible and comparable to earlier reports, as demonstrated by physical sectioning of the spinal cord. The compressed spinal cord is stiffer than a free spinal cord, with a linear increase in SCE with increasing mechanical compression. Knowledge of the biomechanical properties of the spinal cord including SCE has potential implications for disease management and prognosis.

Mesenteric panniculitis. Imaging of a rare cause of chronic abdominal pain.

We report a case of mesenteric panniculitis. This rare and poorly-known disease is characterized by a nonspecific inflammatory process involving the adipose tissue of the mesentery. This case illustrates its computerized tomographic and magnetic resonance imaging features and the value of imaging in differentiating it from other mesenteric diseases and thus, avoiding unnecessary surgery.