The degradation in image resolution harms the performance of medical image diagnosis. By inferring high-frequency details from low-resolution (LR) images, super-resolution (SR) techniques can ...introduce additional knowledge and assist high-level tasks. In this paper, we propose a SR enhanced diagnosis framework, consisting of an efficient SR network and a diagnosis network. Specifically, a Multi-scale Refined Context Network (MRC-Net) with Refined Context Fusion (RCF) is devised to leverage global and local features for SR tasks. Instead of learning from scratch, we first develop a recursive MRC-Net with temporal context, and then propose a recursion distillation scheme to enhance the performance of MRC-Net from the knowledge of the recursive one and reduce the computational cost. The diagnosis network jointly utilizes the reliable original images and more informative SR images by two branches, with the proposed Sample Affinity Interaction (SAI) blocks at different stages to effectively extract and integrate discriminative features towards diagnosis. Moreover, two novel constraints, sample affinity consistency and sample affinity regularization, are devised to refine the features and achieve the mutual promotion of these two branches. Extensive experiments of synthetic and real LR cases are conducted on wireless capsule endoscopy and histopathology images, verifying that our proposed method is significantly effective for medical image diagnosis.
•We present an Instance Importance-aware Graph Convolutional Network (I2GCN) to achieve 3D medical diagnosis with merely patient-level supervision, which enables the multi-instance learning (MIL) ...framework to exploit the relationship among instances comprehensively.•We evaluate the diagnostic importance of each instance by revisiting the channelwise contribution of embeddings, which can be utilized to perform the refined diagnosis. To the best of our knowledge, this work represents the first attempt to estimate the instance importance from the supervised knowledge in MIL.•In the refined diagnosis branch, we propose the Instance Importance-aware Graph Convolutional Layer (I2GCLayer) to explore the complementary information with the importance-based and feature-based topologies under the orthogonal constraint. Besides, the importance-based Sub-Graph Augmentation (SGA) is devised to enhance the graph-based learning compatible with MIL.•We perform extensive experiments of 3D medical diagnosis on publicly available lung CT and prostate MRI datasets to validate the effectiveness of the proposed I2GCN and instance importance calculation thoroughly. The proposed I2GCN outperforms state-of-the-art methods by a large margin.
Display omitted
Automatic diagnosis of 3D medical data is a significant goal of intelligent healthcare. By exploiting the abundant pathological information of 3D data, human experts and algorithms can provide accurate predictions for patients. Considering the high cost of collecting exhaustive annotations for 3D data, a sustainable alternative is to develop diagnosis algorithms with merely patient-level labels. Motivated by the fact that 2D slices of 3D data hold explicit diagnostic efficacy, we propose the Instance Importance-aware Graph Convolutional Network (I2GCN) under the multi-instance learning (MIL). Specifically, we first calculate the instance importance of each slice towards diagnosis using a preliminary MIL classifier, which is further utilized to promote the refined diagnosis branch. In the refined diagnosis branch, we devise the Instance Importance-aware Graph Convolutional Layer (I2GCLayer) to exploit complementary features in both importance-based and feature-based topologies. Moreover, to alleviate the deficient supervision of 3D dataset, we propose the importance-based Sub-Graph Augmentation (SGA) to effectively regularize the framework training. Extensive experiments confirm the effectiveness of our method with different organs and modals on the CC-CCII and PROSTATEx datasets, which outperforms state-of-the-art methods by a large margin. The source code is available at https://github.com/CityU-AIM-Group/I2GCN.
We conducted a comprehensive evaluation of features associated with stroke records.
We screened the dietary nutrients, blood biomarkers, and clinical information from the National Health and ...Nutrition Examination Survey (NHANES) 2015–16 database to assess a self-reported history of all strokes (136 strokes, n = 4381). We computed feature importance, built machine learning (ML) models, developed a nomogram, and validated the nomogram on NHANES 2007–08, 2017–18, and the baseline UK Biobank. We calculated the odds ratios with/without adjusting sampling weights (OR/ORw).
The clinical features have the best predictive power compared to dietary nutrients and blood biomarkers, with 22.8% increased average area under the receiver operating characteristic curves (AUROC) in ML models. We further modeled with ten most important clinical features without compromising the predictive performance. The key features positively associated with stroke include age, cigarette smoking, tobacco smoking, Caucasian or African American race, hypertension, diabetes mellitus, asthma history; the negatively associated feature is the family income. The nomogram based on these key features achieved good performances (AUROC between 0.753 and 0.822) on the test set, the NHANES 2007–08, 2017–18, and the UK Biobank. Key features from the nomogram model include age (OR = 1.05, ORw = 1.06), Caucasian/African American (OR = 2.68, ORw = 2.67), diabetes mellitus (OR = 2.30, ORw = 1.99), asthma (OR = 2.10, ORw = 2.41), hypertension (OR = 1.86, ORw = 2.10), and income (OR = 0.83, ORw = 0.81).
We identified clinical key features and built predictive models for assessing stroke records with high performance. A nomogram consisting of questionnaire-based variables would help identify stroke survivors and evaluate the potential risk of stroke.
•Questionnaire-based features are more informative than the diet and blood biomarkers associated with stroke records.•Top questionnaire-based features are robust, characterizing the stroke survivors.•Nomogram is developed to estimate stroke records without the deployment of complex modeling.
Camels carry Middle East respiratory syndrome coronavirus, but little is known about infection age or prevalence. We studied >800 dromedaries of all ages and 15 mother-calf pairs. This syndrome ...constitutes an acute, epidemic, and time-limited infection in camels <4 years of age, particularly calves. Delayed social separation of calves might reduce human infection risk.
Background
Meningo-cerebral adhesions are frequently encountered during recurrent high-grade glioma resections. Adhesiolysis not only lengthens operation times, but can also induce focal cortical ...tissue injury that could affect overall survival.
Methods
Immediately after the primary resection of a high-grade glioma, a polyesterurethane interpositional graft was implanted in the subdural space covering the entire exposed cortex as well as beneath the dural suture line. No postoperative complications were documented. All patients received adjuvant radiotherapy. Upon repeat resection for focal tumor recurrence, the graft was shown to effectively reduce meningo-cerebral adhesion development.
Conclusion
The implantation of a synthetic subdural graft is a safe and effective method for preventing meningo-cerebral adhesions.
Abstract
BACKGROUND
Gliomas are often in close proximity to functional regions of the brain; therefore, electrocortical stimulation (ECS) mapping is a common technique utilized during glioma ...resection to identify functional areas. Stimulation-induced seizure (SIS) remains the most common reason for aborted procedures. Few studies have focused on oncological factors impacting cortical stimulation thresholds.
OBJECTIVE
To examine oncological factors thought to impact stimulation threshold in order to understand whether a linear relationship exists between stimulation current and number of functional cortical sites identified.
METHODS
We retrospectively reviewed single-institution prospectively collected brain mapping data of patients with dominant hemisphere gliomas. Comparisons of stimulation threshold were made using t-tests and ANOVAs. Associations between oncologic factors and stimulation threshold were made using multivariate regressions. The association between stimulation current and number of positive sites was made using a Poisson model.
RESULTS
Of the 586 patients included in the study, SIS occurred in 3.92% and the rate of SIS events differed by cortical location (frontal 8.5%, insular 1.6%, parietal 1.3%, and temporal 2.8%; P = .009). Stimulation current was lower when mapping frontal cortex (P = .002). Stimulation current was not associated with tumor plus peritumor edema volume, world health organization) (WHO grade, histology, or isocitrate dehydrogenase (IDH) mutation status but was associated with tumor volume within the frontal lobe (P = .018). Stimulation current was not associated with number of positive sites identified during ECS mapping (P = .118).
CONCLUSION
SISs are rare but serious events during ECS mapping. SISs are most common when mapping the frontal lobe. Greater stimulation current is not associated with the identification of more cortical functional sites during glioma surgery.
Sodium valproate (VPA) is a commonly prescribed antiepileptic drug (AED) in daily neurosurgical practice. However, the incidence of VPA-associated hyperammonemia (VAH) and its life-threatening ...consequence, VPA-induced hyperammonemic encephalopathy (VHE), in neurosurgical patients is unknown. We determined the incidence, clinical presentation, and risk factors for VAH.
This prospective cohort study was performed on adult neurosurgical patients prescribed VPA for at least a week over a 22-month period. Blood tests for ammonia, VPA, and liver function were performed at the time of recruitment. The primary end point was VAH. Secondary end points were VHE and liver dysfunction.
In total, 252 patients were recruited. The commonest disease etiology was brain tumors (27%, 69), followed by aneurysmal subarachnoid hemorrhage (SAH; 26%, 65). VPA was prescribed for primary seizure prophylaxis in 110 patients (44%). The mean daily dose was 1148 mg for a mean duration of 48 months. The mean serum VPA level was 417 μmol/L. In total, 92 patients (37%) were prescribed an additional AED, the most common being phenytoin (65%, 60/92). The mean serum ammonia level was 47 μmol/L. In total, 28% (71/252) of patients had VAH and only 0.7% had VHE. Independent factors were aneurysmal SAH (adjusted odds ratio aOR 2.1; 95% confidence interval CI 1.1–4.2), concomitant phenytoin (aOR 1.9; 95% CI 1.0–3.5), and phenobarbital (aOR 4.6; 95% CI 1.1–20.0). No associations with VPA dose, duration, serum levels, and liver function were observed.
Although VAH is common among neurosurgical patients, VHE is rare. Patients with aneurysmal SAH or on concomitant enzyme-inducing AEDs are at risk. Clinicians should be vigilant for VHE symptoms in these patients.
Cases of iatrogenic cerebral amyloid angiopathy (CAA) have been increasingly reported recently, particularly those associated with neurosurgery. Preclinical studies have shown taxifolin to be ...promising for treating CAA. We describe a young 42-year-old man with a history of childhood traumatic brain injury that required a craniotomy for hematoma evacuation. He later presented with recurrent lobar intracerebral hemorrhage (ICH) decades later, which was histologically confirmed to be CAA. Serial
C-Pittsburgh compound B positron emission tomography (
C-PiB-PET) imaging showed a 24% decrease in global standardized uptake value ratio (SUVR) at 10 months after taxifolin use. During this period, the patient experienced clinical improvement with improved consciousness and reduced recurrent ICH frequency, which may be partly attributable to the potential amyloid-β (Aβ) clearing the effect of taxifolin. However, this effect seemed to have diminished at 15 months, CAA should be considered in young patients presenting with recurrent lobar ICH with a history of childhood neurosurgery, and serial
C-PiB-PET scans warrant further validation as a strategy for monitoring treatment response in CAA for candidate Aβ-clearing therapeutic agents such as taxifolin.
Cycling is associated with a greater risk of traumatic brain injury (TBI) than other recreational activities. This study aimed to investigate the epidemiology of sports-related TBI in Hong Kong and ...to examine predictors for recreational cycling-induced intracranial haemorrhage.
This retrospective multicentre study included patients diagnosed with sports-related TBI in public hospitals in Hong Kong from 2015 to 2019. Computed tomography scans were reviewed by an independent assessor. The primary endpoint was traumatic intracranial haemorrhage. The secondary endpoint was an unfavourable Glasgow Outcome Scale (GOS) score at discharge from hospital.
In total, 720 patients were hospitalised with sports-related TBI. The most common sport was cycling (59.2%). The crude incidence of cycling-related TBI was 1.1 per 100 000 population. Cyclists were more likely to exhibit intracranial haemorrhage and an unfavourable GOS score, compared with patients who had TBI because of other sports. Although 47% of cyclists had intracranial haemorrhage, only 15% wore a helmet. In multivariate analysis, significant predictors for intracranial haemorrhage were age ≥60 years, antiplatelet medication, moderate or severe TBI, and skull fracture. Among 426 cyclists, 375 (88%) had mild TBI, and helmet wearing was protective against intracranial haemorrhage, regardless of age, antiplatelet medication intake, and mechanism of injury. Of 426 cyclists, 31 (7.3%) had unfavourable outcomes on discharge from hospital.
The incidence of sports-related TBI is low in Hong Kong. Although cycling-related head injuries carried greater risks of intracranial haemorrhage and unfavourable outcomes compared with other sports, most cyclists experienced good recovery. Helmet wearing among recreational cyclists with mild TBI was protective against intracranial haemorrhage and skull fracture.