Image Feature Analysis Research Articles

Imaging of Pancreatic Ductal Adenocarcinoma: An Update on Recent Advances.

Pancreatic ductal carcinoma (PDAC) is one of the leading causes of cancer-related death worldwide. Computed tomography (CT) remains the primary imaging modality for diagnosis of PDAC. However, CT has limitations for early pancreatic tumor detection and tumor characterization so that it is currently challenged by magnetic resonance imaging. More recently, a particular attention has been given to radiomics for the characterization of pancreatic lesions using extraction and analysis of quantitative imaging features. In addition, radiomics has currently many applications that are developed in conjunction with artificial intelligence (AI) with the aim of better characterizing pancreatic lesions and providing a more precise assessment of tumor burden. This review article sums up recent advances in imaging of PDAC in the field of image/data acquisition, tumor detection, tumor characterization, treatment response evaluation, and preoperative planning. In addition, current applications of radiomics and AI in the field of PDAC are discussed.

Single-kernel classification of deoxynivalenol and zearalenone contaminated maize based on visible light imaging under ultraviolet light excitation combined with polarized light imaging

Deoxynivalenol and zearalenone contamination in maize poses a threat to food safety. Thus, the development of a cost-effective and non-destructive method to classify deoxynivalenol and zearalenone contaminated maize is an important research issue. Natural deoxynivalenol and zearalenone contaminated maize were randomly selected for this experiment. Visible light imaging under ultraviolet light excitation combined with polarized light imaging was used to extract texture and color features from the endosperm and germ sides. And six classification models were established for the single-kernel classification of 1035 contaminated maize. The results indicated that the optimal detection surface was the endosperm side, and the optimal model was SVM with a classification accuracy of 88.1%. Then, local feature analysis of images and mycotoxin concentrations verification experiment were used to reduce secondary contamination and further evaluate the performance of the model, respectively. This method can be verified to be effective in identifying three deoxynivalenol and zearalenone contamination levels of maize, and can also reduce the costs of test stands.

Applying artificial intelligence technology to assist with breast cancer diagnosis and prognosis prediction.

Breast cancer remains the most diagnosed cancer in women. Advances in medical imaging modalities and technologies have greatly aided in the early detection of breast cancer and the decline of patient mortality rates. However, reading and interpreting breast images remains difficult due to the high heterogeneity of breast tumors and fibro-glandular tissue, which results in lower cancer detection sensitivity and specificity and large inter-reader variability. In order to help overcome these clinical challenges, researchers have made great efforts to develop computer-aided detection and/or diagnosis (CAD) schemes of breast images to provide radiologists with decision-making support tools. Recent rapid advances in high throughput data analysis methods and artificial intelligence (AI) technologies, particularly radiomics and deep learning techniques, have led to an exponential increase in the development of new AI-based models of breast images that cover a broad range of application topics. In this review paper, we focus on reviewing recent advances in better understanding the association between radiomics features and tumor microenvironment and the progress in developing new AI-based quantitative image feature analysis models in three realms of breast cancer: predicting breast cancer risk, the likelihood of tumor malignancy, and tumor response to treatment. The outlook and three major challenges of applying new AI-based models of breast images to clinical practice are also discussed. Through this review we conclude that although developing new AI-based models of breast images has achieved significant progress and promising results, several obstacles to applying these new AI-based models to clinical practice remain. Therefore, more research effort is needed in future studies.

Clinical Image Feature Analysis and Diagnostic Efficacy Evaluation of Pulmonary Ultrasound in the Diagnosis of Congenital Pulmonary Airway Malformations in Children: Based on a Single Center, Retrospective Cohort Study.

Objective A single center, retrospective cohort study was conducted to analyze the clinical image features and diagnostic efficiency of pulmonary ultrasound in the diagnosis of congenital pulmonary airway malformations (CPAMs) in children. Methods The starting and ending time of this study is from May 2019 to December 2021. This study included 200 children with CPAM diagnosed by prenatal ultrasound and postpartum CT imaging (aged from 1 hour to 3 years), including 103 males and 97 females. All of them were diagnosed by fetal ultrasound and were examined by chest X-ray (CXR), chest CT, and lung ultrasound (LUS). The clinical image characteristics and diagnostic efficiency of CXR, chest CT, and LUS in the diagnosis of CPAM in children were analyzed. Results 200 lesions were limited to single lung, and the most common were right lower lobe, right lower lobe in 80 cases (40.0%), left lower lobe in 60 cases (30.0%), right upper lobe in 30 cases (15.0%), left upper lobe in 20 cases (10.0%), and right middle lobe in 10 cases (5.0%). Among the 200 cases of preoperative CT examination, 196 cases (98.00%) showed lesions and confirmed diagnosis, and 4 cases were missed. Chest X-ray showed multiple focal circular low-density shadow in the right lung, and the heart shadow and mediastinum moved slightly to the left. CXR showed multiple cystic transparent shadows in the left lower lung and slightly to the right of the mediastinum and heart. CXR showed multiple balloon cavities of different sizes in the right lung field, and the mediastinum and heart shadow shifted to the left. The direct signs of LUS (including single or multiple cystic lesions) were not significantly different from those of CXR, but the indirect signs were significantly higher than those of CXR. Conclusion The most common CT findings of CPAM in children are cystic lesions, especially polycystic lesions, while LUS images of CPAM in children are various. LUS is a noninvasive and nonradiological examination method, which is easy to operate and repeat. LUS can be used for preliminary qualitative screening of CPAM in children, and the diagnostic value of indirect signs of LUS is better than that of CXR.

Peripheral vascular lesions in adults referred to MRI/MRA: Multivariable analysis of imaging features to help differentiate benign vascular anomalies from malignancies

PurposeSymptomatic peripheral vascular lesions in adults are often clinically diagnosed as benign vascular anomalies and may receive MRI/MRA for pre-treatment vascular mapping. Malignant neoplasms are difficult to distinguish from benign vascular anomalies on MRI/MRA. This study was performed to determine if there are imaging signs that can distinguish malignancies from benign vascular anomalies in adults imaged with MRI/MRA. Materials and methodsA radiology database was retrospectively searched for ISSVA classification terms in MRI/MRA reports from 1/1/2002–1/1/2019. Adult patients (n = 50, 52 corresponding lesions) with contrast-enhanced MRI/MRA, peripheral soft tissue based lesion (s), and available pathology or long-term (>1 year) imaging follow-up were included. MRI/MRA images were reviewed by 3 readers for the following lesional characteristics: morphology (marginal lobulation, internal septations, distinct soft tissue mass), peri-articular location, T2-weighted characteristics (hyperintensity, heterogeneity, perilesional edema, and adjacent triangular T2-peaks), bulk fat, hemorrhage, enhancement pattern (peripheral, diffuse, or absent), neovascularity, low-flow venous malformation type enhancement, arterial enhancement within 6 s, enhancement curve (progressive, plateau, or washout), measured size, and multifocality. The MRI/MRA features' associated sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. To identify factors predictive of malignancy, a two-stage multivariable analysis was performed. Results23% (12/52) of the lesions, corresponding to 22% (11/50) of the patients, were malignant neoplasms. No single imaging feature reliably predicted malignancy (PPV ≤ 60%). Absence of distinct soft tissue mass excluded malignancy (NPV 100%). Multivariate analysis derived a summary score based on the five strongest predictors of malignancy: adjacent T2 peaks, age ≥ 70 years, distinct soft tissue mass, lesion size ≥ 5 cm, and absence of septations. A score ≥ 3 resulted in sensitivity of 92% and specificity of 85%. ConclusionExtremity MRI/MRA rarely differentiates malignant from benign soft-tissue vascular tumors in adults. However, MRI/MRA can suggest malignancy when patient age and multiple imaging features are considered. Periodic clinical follow-up after the planned endovascular or operative procedure should be performed to avoid missing a malignancy.

Feature analysis and automatic classification of B-mode ultrasound images of fatty liver

Fatty liver refers to the pathological changes caused by excessive fat accumulation in hepatocytes with various causes. The medical analysis of B-mode ultrasound images(BMUI) is an important method for the identification of the degree of fatty liver. In this paper, a transfer learning method based on Densely Connected Convolutional Networks (DenseNet) and Light Gradient Boosting Machine (LightGBM) was proposed to automatically recognize the four classifications of BMUI of fatty liver. In comparison with the traditional gray and texture feature analysis, this method has the advantages of avoiding the manual extraction the region of interest of BMUI required by the traditional machine learning method for the analysis of feature performance. The complete BMUI were used to train the pathological feature classification model of fatty liver, and this method not only provides the comparison of ultrasonic echo intensity in the near field and far field of BMUI, but also retains the internal texture features and gray information of fatty liver. On the experimental data set, the classification model combined with DenseNet201 network and LightGBM proposed in this paper improved the recognition accuracy by 17.5% on average compared with the support vector machine and migration models based on ResNet101 network. In comparison with the two models above, the size was reduced by 99% on average, indicating remarkably improvement.

Artificial Intelligence-Based Feature Analysis of Ultrasound Images of Liver Fibrosis.

Liver fibrosis is a common liver disease that seriously endangers human health. Liver biopsy is the gold standard for diagnosing liver fibrosis, but its clinical use is limited due to its invasive nature. Ultrasound image examination is a widely used liver fibrosis examination method. Clinicians can diagnose the severity of liver fibrosis according to their own experience by observing the roughness of the texture of the ultrasound image, and this method is highly subjective. Under the premise that artificial intelligence technology is widely used in medical image analysis, this paper uses convolutional neural network analysis to extract the characteristics of ultrasound images of liver fibrosis and then classify the degree of liver fibrosis. Using neural network for image classification can avoid the subjectivity of manual classification and improve the accuracy of judging the degree of liver fibrosis, so as to complete the prevention and treatment of liver fibrosis. Therefore, the following work is done in this paper: (1) the research background, research significance, research status at home and abroad, and the impact of the development of medical imaging on the diagnosis of liver fibrosis are introduced; (2) the related technologies of deep learning and deep convolutional network are introduced, and the indicators of liver fibrosis degree assessment are constructed by using ultrasonic image extraction features; (3) using the collected liver fibrosis dataset to conduct model evaluation experiments, four classic CNN models are selected to compare and analyze the recognition rate. The experiments show that the GoogLeNet model has the best classification and recognition effect.

Monofractal Dimension in Quantifying the Image of Neurons in the Plane: Analysis of Image Features of Multipolar Neurons from the Principal Olivary Nucleus in Humans with Age

The existing study examines four features of 2D images from the principal olivary nucleus of the adult human. The main goal of the research is to investigate the relationship between monofractal and computational parameters that quantify three features of neuronal images. An additional goal of the research is to examine the change in the four features of the image with age. The samples belonged to the histological collection from the Department of Anatomy of the University of Novi Sad. From the pool of binary images, a sample of medium-sized neurons was selected and further processed. A public computer program (Image J with FracLac plugin) with standard commands that calculate computational and monofractal parameters analyzed all images. The relationship between parameters or between parameters with age was performed by statistical evaluation of Pearson’s correlation coefficient. The monofractal dimension of corresponding images can qualitatively represent image properties and some features decrease with age, while some do not. This study confirms the previous conclusions according to which the monofractal dimension of neuronal image reliably quantifies three features of the image. On the other hand, the study shows the distribution of four features with age.

Biomedical Microscopic Imaging in Computational Intelligence Using Deep Learning Ensemble Convolution Learning-Based Feature Extraction and Classification.

Microscopy image analysis gives quantitative support for enhancing the characterizations of various diseases, including breast cancer, lung cancer, and brain tumors. As a result, it is crucial in computer-assisted diagnosis and prognosis. Understanding the biological principles underlying these dynamic image sequences often necessitates precise analysis and statistical quantification, a major discipline issue. Deep learning methods are increasingly used in bioimage processing as they grow rapidly. This research proposes novel biomedical microscopic image analysis techniques using deep learning architectures based on feature extraction and classification. Here, the input image has been taken as microscopic image, and it has been processed and analyzed for noise removal, edge smoothening, and normalization. The processed image has been extracted based on their features in microscopic image analysis using ConVol_NN architecture with AlexNet model. Then, the features have been classified using ensemble of Inception-ResNet and VGG-16 (EN_InResNet_VGG-16) architectures. The experimental results show various dataset analyses in terms of accuracy of 98%, precision of 90%, computational time of 79%, SNR of 89%, and MSE of 62%.

An Integrated Analysis of Clinical, Genomic, and Imaging Features Reveals Predictors of Neurocognitive Outcomes in a Longitudinal Cohort of Pediatric Cancer Survivors, Enriched with CNS Tumors (Rad ART Pro).

BackgroundNeurocognitive deficits in pediatric cancer survivors occur frequently; however, individual outcomes are unpredictable. We investigate clinical, genetic, and imaging predictors of neurocognition in pediatric cancer survivors, with a focus on survivors of central nervous system (CNS) tumors exposed to radiation.MethodsOne hundred eighteen patients with benign or malignant cancers (median diagnosis age: 7; 32% embryonal CNS tumors) were selected from an existing multi-institutional cohort (RadART Pro) if they had: 1) neurocognitive evaluation; 2) available DNA; 3) standard imaging. Utilizing RadART Pro, we collected clinical history, genomic sequencing, CNS imaging, and neurocognitive outcomes. We performed single nucleotide polymorphism (SNP) genotyping for candidate genes associated with neurocognition: COMT, BDNF, KIBRA, APOE, KLOTHO. Longitudinal neurocognitive testing were performed using validated computer-based CogState batteries. The imaging cohort was made of patients with available iron-sensitive (n = 28) and/or T2 FLAIR (n = 41) sequences. Cerebral microbleeds (CMB) were identified using a semi-automated algorithm. Volume of T2 FLAIR white matter lesions (WML) was measured using an automated method based on a convolutional neural network. Summary statistics were performed for patient characteristics, neurocognitive assessments, and imaging. Linear mixed effects and hierarchical models assessed patient characteristics and SNP relationship with neurocognition over time. Nested case-control analysis was performed to compare candidate gene carriers to non-carriers.ResultsCMB presence at baseline correlated with worse performance in 3 of 7 domains, including executive function. Higher baseline WML volumes correlated with worse performance in executive function and verbal learning. No candidate gene reliably predicted neurocognitive outcomes; however, APOE ϵ4 carriers trended toward worse neurocognitive function over time compared to other candidate genes and carried the highest odds of low neurocognitive performance across all domains (odds ratio 2.85, P=0.002). Hydrocephalus and seizures at diagnosis were the clinical characteristics most frequently associated with worse performance in neurocognitive domains (5 of 7 domains). Overall, executive function and verbal learning were the most frequently negatively impacted neurocognitive domains.ConclusionPresence of CMB, APOE ϵ4 carrier status, hydrocephalus, and seizures correlate with worse neurocognitive outcomes in pediatric cancer survivors, enriched with CNS tumors exposed to radiation. Ongoing research is underway to verify trends in larger cohorts.

Value of 4D CT Angiography Combined with Whole Brain CT Perfusion Imaging Feature Analysis under Deep Learning in Imaging Examination of Acute Ischemic Stroke.

This study was aimed at investigating the application of deep learning 4D computed tomography angiography (CTA) combined with whole brain CT perfusion (CTP) imaging in acute ischemic stroke (AIS). A total of 46 patients with ischemic stroke were selected from the hospital as the research objects. Image quality was analyzed after the 4D CTA images were obtained by perfusion imaging. The results showed that whole brain perfusion imaging based on FCN can achieve automatic segmentation. FCN segmentation results took a short time, an average of 2-3 seconds, and the Dice similarity coefficient (DSC) and mean absolute distance (MAD) were lower than those of other algorithms. FCN segmentation distance was 17.87. The parameters of the central area, the peripheral area, and the mirror area of the perfusion map were compared, and the mean transit time (MTT) and time to peak (TTP) of the lesion were prolonged compared with the mirror area. Moreover, the peripheral CBV was increased, and the differences between the parameters were significant (P < 0.05). In conclusion, using the deep learning FCN network, 4D CTA combined with whole brain CTP imaging technology can effectively analyze the perfusion state and achieve clinically personalized treatment.

Tuberculum Sella Meningioma: Surgical Management and Results with Emphasis on Visual Outcome.

Background Tuberculum sella meningioma form a distinct surgical entity with significant morbidity for the patient due to early and profound visual involvement. Surgical treatment of tuberculum sella meningioma is challenging in view of the proximity of vital neuro-vasculature. These lesions may be approached via a high-route (trans-cranial) or a low-route (trans-sphenoidal). Materials and Methods The authors present a consecutive series of 36 patients, operated by the first author from 1996 onward, with an analysis of clinical presentation (emphasizing visual deterioration), imaging features, operative approaches, and outcome with postoperative visual status. Results There were 36 patients (72 eyes) in the series, 75% females and 25% males. The mean age of patients was 48.05 ± 10.02 years. In all, 88.89% had visual diminution, 50% had headache, 5.56% had behavioral changes, and 2.78% had seizures. The mean tumor size was 4.01 ± 1.01 cm. The frontotemporal approach was employed in 66.67% of patients, and all patients underwent a right-sided approach. There was no significant difference between improvement in the vision between the right eye and left eye, despite the fact that right-sided approach was employed in nearly cases. Conclusion While tuberculum sella meningioma have profound morbidity in terms of visual deterioration, timely surgical intervention may help patients gain serviceable vision postoperatively. We advocate the trans-cranial approach and find no difference in visual outcomes on employing right-sided approach, regardless of the side with worse vision.

NFB-03. Neurological manifestations in children and adolescents with Neurofibromatosis type 1 - Implications for management and surveillance

INTRODUCTION: We aimed to (1) characterize the spectrum of clinical phenotypes of NF1 in a random pediatric population, (2) correlate genotype with phenotypic expression for those with a genetic diagnosis, and (3) explore radiological features of NF1 in the central nervous system (CNS) by radiomics analyses to predict clinical course. METHODS: We performed a database search in the hospital information system of the University Children′s Hospital between January 2017 and December 2020 for patients with NF1 and evaluated the clinical phenotype by retrospective chart review. RESULTS: 75 children/adolescents were identified with suspicion/clinical diagnosis of NF1 (median age 10.0 years (range, 1.1-22.6); 35 female), confirmatory revised “diagnostic criteria” were met in 57 patients at the last follow-up. Per number of documented items, major signs were detected as 73/75 café-au-lait macules, 31/63 freckling, 38/71 neurofibromas (thereof 21 plexiform neurofibromas), 18/43 optic pathway glioma, 5/66 Lisch nodules, and two patients with sphenoid dysplasia. Genetic analysis (31/75) identified pathogenic NF1 variants in 27 patients. In 20/66 cases a parent met diagnostic criteria. Cognitive symptoms included developmental delay (28/68), learning deficits (12/48), attention-deficit hyperactivity disorder (3/53), and behavior anomalies (7/63). Classical unidentified bright objects were seen in 29/43, other intracranial tumors in 7/43, and cerebrovascular abnormalities in 5/43. Analysis of imaging features of the CNS in these patients will involve lesion segmentation and radiomics features. Symptomatic/progressive low-grade glioma necessitated neurosurgical resection (4/25) and/or chemotherapy (12/25). In 10/25 neuropsychological functions were assessed by the German neuropsychological basic diagnostic instrument. Until June 30th, 2021, one patient died of progressive plexiform neurofibroma. CONCLUSIONS: A wide range of neurological manifestations, including neuropsychological deficits, should raise the suspicion of NF1 in an unselected pediatric population. We expect imaging features of the CNS to better predict the clinical course and enhance decision-making.

Adhesive Abrasive Detection for Diamond Images based on Improved Watershed Algorithm

Diamond images have characteristics of adhesive abrasives besides much background noise, irregular abrasive shapes and different abrasive size, which brings big challenges to accurate diamond abrasive detection. Therefore, an improved watershed algorithm is put forward to detect adhesive abrasive in this work. Firstly, the diamond abrasive image is filtered by Gaussian to suppress the background noise, and then the pre-processed diamond abrasive image is reconstructed by morphology. Through the distance transform of the reconstructed image, marking features for subsequent abrasive detection can be obtained. Finally, the watershed algorithm and extended-minima transform are used to finely detect diamond abrasives and separate adhesive abrasives, respectively, so as to realize the accurate detection of adhesive diamond abrasive image. According to the results of detection experiments, the proposed method based on improved watershed algorithm and extended-minima transform can accurately detect adhesive diamond abrasive images, the recall rate of abrasive is 94.8%, which indicates good recognition accuracy and robustness of the proposed method. The accurate detection results can be further used for subsequent image analysis and abrasive feature parameter extraction.

Deep Learning Methods to Reveal Important X-ray Features in COVID-19 Detection: Investigation of Explainability and Feature Reproducibility

X-ray technology has been recently employed for the detection of the lethal human coronavirus disease 2019 (COVID-19) as a timely, cheap, and helpful ancillary method for diagnosis. The scientific community evaluated deep learning methods to aid in the automatic detection of the disease, utilizing publicly available small samples of X-ray images. In the majority of cases, the results demonstrate the effectiveness of deep learning and suggest valid detection of the disease from X-ray scans. However, little has been investigated regarding the actual findings of deep learning through the image process. In the present study, a large-scale dataset of pulmonary diseases, including COVID-19, was utilized for experiments, aiming to shed light on this issue. For the detection task, MobileNet (v2) was employed, which has been proven very effective in our previous works. Through analytical experiments utilizing feature visualization techniques and altering the input dataset classes, it was suggested that MobileNet (v2) discovers important image findings and not only features. It was demonstrated that MobileNet (v2) is an effective, accurate, and low-computational-cost solution for distinguishing COVID-19 from 12 various other pulmonary abnormalities and normal subjects. This study offers an analysis of image features extracted from MobileNet (v2), aiming to investigate the validity of those features and their medical importance. The pipeline can detect abnormal X-rays with an accuracy of 95.45 ± 1.54% and can distinguish COVID-19 with an accuracy of 89.88 ± 3.66%. The visualized results of the Grad-CAM algorithm provide evidence that the methodology identifies meaningful areas on the images. Finally, the detected image features were reproducible in 98% of the times after repeating the experiment for three times.

Ultrasonic Image Feature Analysis under Deep Learning Algorithm to Evaluate the Efficacy of Drug-Coated Balloon for Treatment of Arteriosclerotic Occlusion.

In order to explore the adoption of ultrasonic images under deep learning (DL) algorithm to evaluate the efficacy of drug-coated balloon (DCB) for treatment of arteriosclerotic occlusion, 56 patients who underwent DCB surgery of lower limb artery were selected and all the patients received the examinations of algorithmic ultrasound and digital subtraction angiography (DSA) before surgery. According to the examination methods, they were classified into algorithmic ultrasound group and DSA group. One to two months after DCB surgery, ultrasound examination was performed with the region-based faster convolutional neural network (faster R-CNN) target detection algorithm to check the therapeutic effect. The results showed that the image effect processed by the target detection algorithm based on DL was signally better than that of traditional ultrasonic processing algorithm in Dice, precision (Pre), and sensitivity, with significant difference (P < 0.05). Compared with DSA, algorithmic ultrasound showed better consistency between the two groups in the diagnosis of common femoral artery, superficial femoral artery, and popliteal artery stenosis, with statistical significance (P < 0.05). However, for the diagnosis of anterior tibial artery stenosis, the consistency between algorithmic ultrasound and DSA was general. The residual stenosis of each artery segment decreased obviously in postoperative review compared with that before surgery, and the difference was statistically significant (P < 0.05). Besides, both the pulsatility index (PI) and the blood flow velocity of the dorsalis pedis artery increased after surgery, compared with those before surgery, with significant differences (P < 0.05). To sum up, ultrasound based on DL target detection algorithm had good imaging effect and good consistency with DSA, which was of the clinical reference value. Additionally, DCB surgery was helpful to treat arteriosclerosis occlusion and improve limb blood supply, which had clinical adoption value.

Can lymphovascular invasion be predicted by contrast-enhanced CT imaging features in patients with esophageal squamous cell carcinoma? A preliminary retrospective study

BackgroundTo investigate the value of contrast-enhanced CT (CECT)-derived imaging features in predicting lymphovascular invasion (LVI) status in esophageal squamous cell carcinoma (ESCC) patients.MethodsOne hundred and ninety-seven patients with postoperative pathologically confirmed esophageal squamous cell carcinoma treated in our hospital between January 2017 and January 2019 were enrolled in our study, including fifty-nine patients with LVI and one hundred and thirty-eight patients without LVI. The CECT-derived imaging features of all patients were analyzed. The CECT-derived imaging features were divided into quantitative features and qualitative features. The quantitative features consisted of the CT attenuation value of the tumor (CTVTumor), the CT attenuation value of the normal esophageal wall (CTVNormal), the CT attenuation value ratio of the tumor-to-normal esophageal wall (TNR), the CT attenuation value difference between the tumor and normal esophageal wall (ΔTN), the maximum thickness of the tumor measured by CECT (Thickness), the maximum length of the tumor measured by CECT (Length), and the gross tumor volume measured by CECT (GTV). The qualitative features consisted of an enhancement pattern, tumor margin, enlarged blood supply or drainage vessels to the tumor (EVFDT), and tumor necrosis. For the clinicopathological characteristics and CECT-derived imaging feature analysis, the chi-squared test was used for categorical variables, the Mann–Whitney U test was used for continuous variables with a nonnormal distribution, and the independent sample t-test was used for the continuous variables with a normal distribution. The trend test was used for ordinal variables. The association between LVI status and CECT-derived imaging features was analyzed by univariable logistic analysis, followed by multivariable logistic regression and receiver operating characteristic (ROC) curve analysis.ResultsThe CTVTumor, TNR, ΔTN, Thickness, Length, and GTV in the group with LVI were higher than those in the group without LVI (P < 0.05). A higher proportion of patients with heterogeneous enhancement pattern, irregular tumor margin, EVFDT, and tumor necrosis were present in the group with LVI (P < 0.05). As revealed by the univariable logistic analysis, the CECT-derived imaging features, including CTVTumor, TNR, ΔTN and enhancement pattern, Thickness, Length, GTV, tumor margin, EVFDT, and tumor necrosis were associated with LVI status (P < 0.05). Only the TNR (OR 8.655; 95% CI 2.125–37.776), Thickness (OR 6.531; 95% CI 2.410–20.608), and tumor margin (OR 4.384; 95% CI 2.004–9.717) were independent risk factors for LVI in the multivariable logistic regression analysis. The ROC curve analysis incorporating the above three CECT-derived imaging features showed that the area under the curve obtained by the multivariable logistic regression model was 0.820 (95% CI 0.754–0.885).ConclusionThe CECT-derived imaging features, including TNR, Thickness, tumor margin, and their combination, can be used as predictors of LVI status for patients with ESCC.

Navigated Intraoperative 3D Ultrasound in Glioblastoma Surgery: Analysis of Imaging Features and Impact on Extent of Resection.

BackgroundNeuronavigation is routinely used in glioblastoma surgery, but its accuracy decreases during the operative procedure due to brain shift, which can be addressed utilizing intraoperative imaging. Intraoperative ultrasound (iUS) is widely available, offers excellent live imaging, and can be fully integrated into modern navigational systems. Here, we analyze the imaging features of navigated i3D US and its impact on the extent of resection (EOR) in glioblastoma surgery.MethodsDatasets of 31 glioblastoma resection procedures were evaluated. Patient registration was established using intraoperative computed tomography (iCT). Pre-operative MRI (pre-MRI) and pre-resectional ultrasound (pre-US) datasets were compared regarding segmented tumor volume, spatial overlap (Dice coefficient), the Euclidean distance of the geometric center of gravity (CoG), and the Hausdorff distance. Post-resectional ultrasound (post-US) and post-operative MRI (post-MRI) tumor volumes were analyzed and categorized into subtotal resection (STR) or gross total resection (GTR) cases.ResultsThe mean patient age was 59.3 ± 11.9 years. There was no significant difference in pre-resectional segmented tumor volumes (pre-MRI: 24.2 ± 22.3 cm3; pre-US: 24.0 ± 21.8 cm3). The Dice coefficient was 0.71 ± 0.21, the Euclidean distance of the CoG was 3.9 ± 3.0 mm, and the Hausdorff distance was 12.2 ± 6.9 mm. A total of 18 cases were categorized as GTR, 10 cases were concordantly classified as STR on MRI and ultrasound, and 3 cases had to be excluded from post-resectional analysis. In four cases, i3D US triggered further resection.ConclusionNavigated i3D US is reliably adjunct in a multimodal navigational setup for glioblastoma resection. Tumor segmentations revealed similar results in i3D US and MRI, demonstrating the capability of i3D US to delineate tumor boundaries. Additionally, i3D US has a positive influence on the EOR, allows live imaging, and depicts brain shift.

Time trends in 30-day mortality in patients with pulmonary embolism

Abstract Funding Acknowledgements Type of funding sources: None. Background There is a lack of evidence about recent mortality rates in patients with acute pulmonary embolism (PE). Purpose To compare 30-day mortality between the current and historical cohort of patients treated for acute PE. Methods We provide a retrospective analysis of clinical and imaging features of 400 consecutive patients who were admitted to our hospital with the diagnosis of acute pulmonary embolism between February 2010 and April 2021. Patients with out-of-hospital cardiac arrest without return of spontaneous circulation (14 patients) were excluded from the study cohort. The 30-day mortality was compared between patients admitted before January 2016 (n =178) and after January 2016 (n = 222). We stratified patients into low-, intermediate-, and high-risk group according to the ‘2019 ESC Guidelines on Acute Pulmonary Embolism’. Results The diagnosis of pulmonary embolism was based on results of computed tomography (86%), pulmonary ventilation/perfusion scan (7%), echocardiography (7%) and autopsy (1%). Most patients were stratified as intermediate-risk (313 patients [78 %]), followed by low-risk (46 [12%]) and high-risk pulmonary embolism (41 [10%]). The risk-stratification profile did not differ between the current and historical cohort (79% vs. 77% for intermediate-risk, 12% vs. 11% for low-risk, and 9% vs. 12% for high-risk patients, all p&amp;gt; 0.05). Seventeen patients (4%) died within 30 days of admission to the hospital: ten patients in the high-risk (24%), seven in the intermediate-risk (2%), and none (0%) in the low-risk group (p &amp;lt;0.0001). There was no significant difference in 30-day mortality between the current and historical cohort (9 [4%] vs. 8 [4%], p = 0.83). In a multivariate regression analysis, only the initial presence of cardiogenic shock was associated with higher 30-day mortality. Conclusion In-hospital mortality of patients with intermediate- and low-risk pulmonary embolism is lower than reported in the past and has not changed over the last decade. High-risk pulmonary embolism is associated with 24% 30-day mortality.

Correlation Analysis of Magnetic Resonance Imaging Characteristics and Prognosis of Invasive Pituitary Adenomas in Neurosurgery Hospitals.

The incidence of pituitary adenoma is second only to glioma and meningioma, and its incidence ranks third among intracranial tumors. Most pituitary adenomas are benign and noninvasive tumors, but invasive pituitary adenomas pose a great threat to human health. In order to explore the risk factors that affect the clinical aggressive behavior of patients with pituitary adenoma, analyze the correlation between different classification methods and clinical aggressive behavior, and lay the foundation for early judgment and individualized treatment of clinical aggressive behavior of patients with pituitary adenoma. We conducted statistical research on patients who were treated for pituitary adenomas in the city's Yangzhou Hongquan Hospital. The results of the study showed that six patients in this study showed aggressiveness in the clinical symptomatic outcome, six patients showed aggressiveness in the serological outcome, and seven patients showed aggressiveness in imaging. In the multimodal classification, the clinical aggressiveness of pituitary adenomas in the invasion + atypical group was significantly higher than that in other groups, and the difference was statistically significant (P < 0.05). The correlation analysis of magnetic resonance imaging features and prognosis of invasive pituitary adenomas were verified to be feasible for the treatment of patients.