Search results
Found 30971 matches for
TCPNet: A Novel Tumor Contour Prediction Network Using MRIs
Brain cancer ranks among the top ten causes of death globally and stands as the second leading cause of cancer-related deaths among adolescents. Magnetic Resonance Imaging (MRI) is widely used as one of the most important diagnostic imaging techniques for early detection of tumors in brain cancer. In general, the contour is manually identified by experts from the MRIs, causing errors due to subjective assessments. In order to address this issue, image segmentation techniques can be used to automate the process of identifying and delineating the tumor region from the MR images. In this work, we propose a novel deep-learning framework named TCPNet, which is developed in the spirit of the U-Net model. The proposed architecture ensures that the model segments the tumor contours and explicitly estimates data and model uncertainties in the predictions, which is essential for tumor contour detection. Our extensive study on two open-source brain MRI datasets shows that TCPNet performs better than U-Net and other state-of-the-arts in terms of common evaluation metrics. Additionally, the proposed model presents uncertainties in model predictions, demonstrating confidence in segmented tissues or advising for expert intervention if necessary.
Uncertainty Quantification in Deep Learning Framework for Mallampati Classification
Mallampati classification is an indication to predict whether a patient might have crowded airways. According to the scale, there are four classes with increasing severity of airway crowding, which may indicate obstructive sleep apnea, as reported in multiple studies. Conventionally, the Mallampati scale is manually identified by an expert in the clinic, but the same can be done by assessing the image of a person's oral cavity. In this regard, this study aims to develop a deep learning framework to perform Mallampati classification using the oral cavity images of individuals. The proposed framework for Mallampati classification develops a loss function by combining the aleatoric and epistemic uncertainty principles to improve the reliability of predictions of the ConvNeXt model for image classification. The experimental analysis was performed on a dataset of 262 subjects acquired from the sleep lab at All India Institute of Medical Sciences Bhopal in India, demonstrating that the proposed framework performs better than the state-of-the-art in terms of common evaluation metrics. Notably, experimental results reveal that the proposed uncertainty principle performs well as the experiments were conducted on all states of the arts with and without using this principle.
Knowledge, attitudes, and behaviours about gender equality: a cross-sectional survey of adolescents from rural India.
Background: Gender equality is a fundamental human right and vital to accelerate global progress towards several Sustainable Development Goals (SDGs). Adolescents' involvement is essential to achieve such equality and SDGs to develop peaceful sustainable societies. However, there are limited data especially from developing countries such as India to plan gender equality related programmes targeted at adolescents. Methods: We conducted a cross-sectional survey to assess gender equality related knowledge, attitudes, and behaviours among 16 to 19 year-old adolescents from sixty villages of the Maharashtra state of India. Results: Data from 1306 respondents (667 females and 639 males) showed a mean score of 30 out of 44, suggesting an overall moderate gender equality score in rural adolescents. The majority of girls (68.3%) were in the high scoring group, whereas the majority of boys were in the moderate group (60.3%). Regression analysis showed that responses from boys were associated with lower scores compared to responses from girls by five points (adjusted β-coefficient: -4.99, 95%CI: -5.85 to -4.12, p<0.001). Conclusions: Our findings suggest that there is a need to involve adolescents with a major focus on boys to improve gender equality in rural areas of Maharashtra. This will help introduce concepts of equality from an early age to educate boys, empower girls, and address gender-based discrimination and violence against girls and women.
Diagnostic utility of electrocardiogram for screening of cardiac injury on cardiac magnetic resonance in post-hospitalised COVID-19 patients: a prospective multicenter study.
BACKGROUND: The role of ECG in ruling out myocardial complications on cardiac magnetic resonance (CMR) is unclear. We examined the clinical utility of ECG in screening for cardiac abnormalities on CMR among post-hospitalised COVID-19 patients. METHODS: Post-hospitalised patients (n = 212) and age, sex and comorbidity-matched controls (n = 38) underwent CMR and 12‑lead ECG in a prospective multicenter follow-up study. Participants were screened for routinely reported ECG abnormalities, including arrhythmia, conduction and R wave abnormalities and ST-T changes (excluding repolarisation intervals). Quantitative repolarisation analyses included corrected QT (QTc), corrected QT dispersion (QTc disp), corrected JT (JTc) and corrected T peak-end (cTPe) intervals. RESULTS: At a median of 5.6 months, patients had a higher burden of ECG abnormalities (72.2% vs controls 42.1%, p = 0.001) and lower LVEF but a comparable cumulative burden of CMR abnormalities than controls. Patients with CMR abnormalities had more ECG abnormalities and longer repolarisation intervals than those with normal CMR and controls (82% vs 69% vs 42%, p
Selective advantage of mutant stem cells in human clonal hematopoiesis is associated with attenuated response to inflammation and aging.
Clonal hematopoiesis (CH) arises when hematopoietic stem cells (HSCs) acquire mutations, most frequently in the DNMT3A and TET2 genes, conferring a competitive advantage through mechanisms that remain unclear. To gain insight into how CH mutations enable gradual clonal expansion, we used single-cell multi-omics with high-fidelity genotyping on human CH bone marrow (BM) samples. Most of the selective advantage of mutant cells occurs within HSCs. DNMT3A- and TET2-mutant clones expand further in early progenitors, while TET2 mutations accelerate myeloid maturation in a dose-dependent manner. Unexpectedly, both mutant and non-mutant HSCs from CH samples are enriched for inflammatory and aging transcriptomic signatures, compared with HSCs from non-CH samples, revealing a non-cell-autonomous effect. However, DNMT3A- and TET2-mutant HSCs have an attenuated inflammatory response relative to wild-type HSCs within the same sample. Our data support a model whereby CH clones are gradually selected because they are resistant to the deleterious impact of inflammation and aging.
Ketogenic diet but not free-sugar restriction alters glucose tolerance, lipid metabolism, peripheral tissue phenotype, and gut microbiome: RCT.
Restricted sugar and ketogenic diets can alter energy balance/metabolism, but decreased energy intake may be compensated by reduced expenditure. In healthy adults, randomization to restricting free sugars or overall carbohydrates (ketogenic diet) for 12 weeks reduces fat mass without changing energy expenditure versus control. Free-sugar restriction minimally affects metabolism or gut microbiome but decreases low-density lipoprotein cholesterol (LDL-C). In contrast, a ketogenic diet decreases glucose tolerance, increases skeletal muscle PDK4, and reduces AMPK and GLUT4 levels. By week 4, the ketogenic diet reduces fasting glucose and increases apolipoprotein B, C-reactive protein, and postprandial glycerol concentrations. However, despite sustained ketosis, these effects are no longer apparent by week 12, when gut microbial beta diversity is altered, possibly reflective of longer-term adjustments to the ketogenic diet and/or energy balance. These data demonstrate that restricting free sugars or overall carbohydrates reduces energy intake without altering physical activity, but with divergent effects on glucose tolerance, lipoprotein profiles, and gut microbiome.
Bidirectional Mendelian randomization highlights causal relationships between circulating INHBC and multiple cardiometabolic diseases and traits.
Human genetic and transgenic mouse studies have highlighted a potential liver-adipose tissue endocrine axis, involving activin C (Act-C) and/or Act-E and ALK7, influencing fat distribution and systemic metabolism. We investigated the bidirectional effects between circulating INHBC, which homodimerizes into Act-C, and adiposity traits, insulin resistance, inflammation, and cardiometabolic disease risk. Additionally, we examined if Act-C is an ALK7 ligand in human adipocytes. We used Mendelian randomization and in vitro studies in immortalized human abdominal and gluteal adipocytes. Circulating INHBC was causally linked to reduced lower-body fat, dyslipidaemia, and increased risks of coronary artery disease (CAD) and non-alcoholic fatty liver disease (NAFLD). Conversely, upper-body fat distribution, obesity, hypertriglyceridemia, subclinical inflammation, and type 2 diabetes positively impacted plasma INHBC levels. Mechanistically, an atherogenic lipid profile may partly explain the INHBC-CAD link, while inflammation and hypertriglyceridemia may partly explain how adiposity traits affect circulating INHBC. Phenome-wide Mendelian randomization showed weak causal relationships between higher plasma INHBC and impaired kidney function and higher gout risk. In human adipocytes, recombinant Act-C activated SMAD2/3 signaling via ALK7 and suppressed lipolysis. In summary, INHBC influences systemic metabolism by activating ALK7 in adipose tissue and may serve as a drug target for atherogenic dyslipidemia, CAD, and NAFLD.
Familial severe skeletal Class II malocclusion with gingival hyperplasia caused by a complex structural rearrangement at the KCNJ2-KCNJ16 locus.
The aim of this work was to identify the underlying genetic cause in a four-generation family segregating an unusual phenotype comprising a severe form of skeletal Class II malocclusion with gingival hyperplasia. SNP-array identified a copy number gain on chr1, however this chromosomal region did not segregate correctly in the extended family. Exome sequencing also failed to identify a candidate causative variant, but highlighted co-segregating genetic markers on chr17 and chr19. Short- and long-read genome sequencing allowed us to pinpoint and characterize at nucleotide-level resolution a chromothripsis-like complex rearrangement (CR) inserted into the chr17 co-segregating region at the KCNJ2-SOX9 locus. The CR involved the gain of five different regions from chr1 that are shuffled, chained and inserted as a single block (∼828 kb) at chr17q24.3. The inserted sequences contain craniofacial enhancers that are predicted to interact with KCNJ2/KCNJ16 through neo-topologically associating domain (TAD) formation to induce ectopic activation. Our findings suggest that the CR inserted at chr17q24.3 is the cause of the severe skeletal Class II malocclusion with gingival hyperplasia in this family and expands the panoply of phenotypes linked to variation at the KCNJ2-SOX9 locus. In addition, we highlight a previously overlooked potential role for misregulation of the KCNJ2/KCNJ16 genes in the pathomechanism of gingival hyperplasia associated with deletions and other rearrangements of the 17q24.2-q24.3 region (MIM 135400).
Clinical Significance of Myocardial Injury in Patients Hospitalized for COVID-19: A Prospective, Multicenter, Cohort Study.
BACKGROUND: Hospitalized COVID-19 patients with troponin elevation have a higher prevalence of cardiac abnormalities than control individuals. However, the progression and impact of myocardial injury on COVID-19 survivors remain unclear. OBJECTIVES: This study sought to evaluate myocardial injury in COVID-19 survivors with troponin elevation with baseline and follow-up imaging and to assess medium-term outcomes. METHODS: This was a prospective, longitudinal cohort study in 25 United Kingdom centers (June 2020 to March 2021). Hospitalized COVID-19 patients with myocardial injury underwent cardiac magnetic resonance (CMR) scans within 28 days and 6 months postdischarge. Outcomes were tracked for 12 months, with quality of life surveys (EuroQol-5 Dimension and 36-Item Short Form surveys) taken at discharge and 6 months. RESULTS: Of 342 participants (median age: 61.3 years; 71.1% male) with baseline CMR, 338 had a 12-month follow-up, 235 had a 6-month CMR, and 215 has baseline and follow-up quality of life surveys. Of 338 participants, within 12 months, 1.2% died; 1.8% had new myocardial infarction, acute coronary syndrome, or coronary revascularization; 0.8% had new myopericarditis; and 3.3% had other cardiovascular events requiring hospitalization. At 6 months, there was a minor improvement in left ventricular ejection fraction (1.8% ± 1.0%; P < 0.001), stable right ventricular ejection fraction (0.4% ± 0.8%; P = 0.50), no change in myocardial scar pattern or volume (P = 0.26), and no imaging evidence of continued myocardial inflammation. All pericardial effusions (26 of 26) resolved, and most pneumonitis resolved (95 of 101). EuroQol-5 Dimension scores indicated an overall improvement in quality of life (P < 0.001). CONCLUSIONS: Myocardial injury in severe hospitalized COVID-19 survivors is nonprogressive. Medium-term outcomes show a low incidence of major adverse cardiovascular events and improved quality of life. (COVID-19 Effects on the Heart; ISRCTN58667920).
An automatic deep reinforcement learning bolus calculator for automated insulin delivery systems
AbstractIn hybrid automatic insulin delivery (HAID) systems, meal disturbance is compensated by feedforward control, which requires the announcement of the meal by the patient with type 1 diabetes (DM1) to achieve the desired glycemic control performance. The calculation of insulin bolus in the HAID system is based on the amount of carbohydrates (CHO) in the meal and patient-specific parameters, i.e. carbohydrate-to-insulin ratio (CR) and insulin sensitivity-related correction factor (CF). The estimation of CHO in a meal is prone to errors and is burdensome for patients. This study proposes a fully automatic insulin delivery (FAID) system that eliminates patient intervention by compensating for unannounced meals. This study exploits the deep reinforcement learning (DRL) algorithm to calculate insulin bolus for unannounced meals without utilizing the information on CHO content. The DRL bolus calculator is integrated with a closed-loop controller and a meal detector (both previously developed by our group) to implement the FAID system. An adult cohort of 68 virtual patients based on the modified UVa/Padova simulator was used for in-silico trials. The percentage of the overall duration spent in the target range of 70–180 mg/dL was $$71.2\%$$ 71.2 % and $$76.2\%$$ 76.2 % , $$<70$$ < 70 mg/dL was $$0.9\%$$ 0.9 % and $$0.1\%$$ 0.1 % , and $$>180$$ > 180 mg/dL was $$26.7\%$$ 26.7 % and $$21.1\%$$ 21.1 % , respectively, for the FAID system and HAID system utilizing a standard bolus calculator (SBC) including CHO misestimation. The proposed algorithm can be exploited to realize FAID systems in the future.
Modelling Multi-Phase Cardiac Anatomy Using Point Cloud Variational Autoencoders
The dynamic behaviour of the cardiac anatomy has a considerable impact on cardiac function and disease progression. Geometric deep learning methods have shown promising results for simulating the shape of biventricular anatomy, primarily emphasizing individual phases from the cycle. However, a more comprehensive understanding of cardiovascular disease could be achieved by analysing abnormalities in motion, which can be captured through modelling multiple phases together in the cardiac cycle. In this work, we propose a novel geometric deep learning architecture capable of encoding and predicting 3D shapes of the biventricular anatomy for end-diastolic and end-systolic phases of the cardiac cycle using only a single extreme phase as input. The reconstruction and prediction performance of the proposed architecture are accurate to sub-pixel resolution, with an average Chamfer distance of 1.58 (±0.16) mm. We also investigate the generative properties of the proposed model, finding the model can generate a range of diverse and realistic 3D cardiac shapes and has interpretability in its latent space.