scholarly journals Improving five-year survival prediction via multitask learning across HPV-related cancers

PLoS ONE ◽  
2020 ◽  
Vol 15 (11) ◽  
pp. e0241225
Author(s):  
Andre Goncalves ◽  
Braden Soper ◽  
Mari Nygård ◽  
Jan F. Nygård ◽  
Priyadip Ray ◽  
...  
Keyword(s):  
Relevant Information ◽  
The United States ◽  
Multitask Learning ◽  
Survival Models ◽  
Survival Prediction ◽  
Study Cohort ◽  
Single Task ◽  
Task Learning ◽  
Cancer Types ◽  
Underlying Processes

Oncology is a highly siloed field of research in which sub-disciplinary specialization has limited the amount of information shared between researchers of distinct cancer types. This can be attributed to legitimate differences in the physiology and carcinogenesis of cancers affecting distinct anatomical sites. However, underlying processes that are shared across seemingly disparate cancers probably affect prognosis. The objective of the current study is to investigate whether multitask learning improves 5-year survival cancer patient survival prediction by leveraging information across anatomically distinct HPV related cancers. Data were obtained from the Surveillance, Epidemiology, and End Results (SEER) program database. The study cohort consisted of 29,768 primary cancer cases diagnosed in the United States between 2004 and 2015. Ten different cancer diagnoses were selected, all with a known association with HPV risk. In the analysis, the cancer diagnoses were categorized into three distinct topography groups of varying specificity. The most specific topography grouping consisted of 10 original cancer diagnoses differentiated by the first two digits of the ICD-O-3 topography code. The second topography grouping consisted of cancer diagnoses categorized into six distinct organ groups. Finally, the third topography grouping consisted of just two groups, head-neck cancers and ano-genital cancers. The tasks were to predict 5-year survival for patients within the different topography groups using 14 predictive features which were selected among descriptive variables available in the SEER database. The information from the predictive features was shared between tasks in three different ways, resulting in three distinct predictive models: 1) Information was not shared between patients assigned to different tasks (single task learning); 2) Information was shared between all patients, regardless of task (pooled model); 3) Only relevant information was shared between patients grouped to different tasks (multitask learning). Prediction performance was evaluated with Brier scores. All three models were evaluated against one another on each of the three distinct topography-defined tasks. The results showed that multitask classifiers achieved relative improvement for the majority of the scenarios studied compared to single task learning and pooled baseline methods. In this study, we have demonstrated that sharing information among anatomically distinct cancer types can lead to improved predictive survival models.

scholarly journals Outcome Prediction in Patients with Severe COVID-19 Requiring Extracorporeal Membrane Oxygenation—A Retrospective International Multicenter Study

Membranes ◽  
2021 ◽  
Vol 11 (3) ◽  
pp. 170
Author(s):  
Alexander Supady ◽  
Jeff DellaVolpe ◽  
Fabio Silvio Taccone ◽  
Dominik Scharpf ◽  
Matthias Ulmer ◽  
...  
Keyword(s):  
Extracorporeal Membrane Oxygenation ◽  
The United States ◽  
Apache Ii ◽  
Apache Ii Score ◽  
Survival Prediction ◽  
Study Cohort ◽  
Additional Tool ◽  
Saps Ii ◽  
Membrane Oxygenation ◽  
International Multicenter

The role of veno-venous extracorporeal membrane oxygenation therapy (V-V ECMO) in severe COVID-19 acute respiratory distress syndrome (ARDS) is still under debate and conclusive data from large cohorts are scarce. Furthermore, criteria for the selection of patients that benefit most from this highly invasive and resource-demanding therapy are yet to be defined. In this study, we assess survival in an international multicenter cohort of COVID-19 patients treated with V-V ECMO and evaluate the performance of several clinical scores to predict 30-day survival. Methods: This is an investigator-initiated retrospective non-interventional international multicenter registry study (NCT04405973, first registered 28 May 2020). In 127 patients treated with V-V ECMO at 15 centers in Germany, Switzerland, Italy, Belgium, and the United States, we calculated the Sequential Organ Failure Assessment (SOFA) Score, Simplified Acute Physiology Score II (SAPS II), Acute Physiology And Chronic Health Evaluation II (APACHE II) Score, Respiratory Extracorporeal Membrane Oxygenation Survival Prediction (RESP) Score, Predicting Death for Severe ARDS on V‑V ECMO (PRESERVE) Score, and 30-day survival. Results: In our study cohort which enrolled 127 patients, overall 30-day survival was 54%. Median SOFA, SAPS II, APACHE II, RESP, and PRESERVE were 9, 36, 17, 1, and 4, respectively. The prognostic accuracy for all these scores (area under the receiver operating characteristic—AUROC) ranged between 0.548 and 0.605. Conclusions: The use of scores for the prediction of mortality cannot be recommended for treatment decisions in severe COVID-19 ARDS undergoing V-V ECMO; nevertheless, scoring results below or above a specific cut-off value may be considered as an additional tool in the evaluation of prognosis. Survival rates in this cohort of COVID-19 patients treated with V‑V ECMO were slightly lower than those reported in non-COVID-19 ARDS patients treated with V-V ECMO.

scholarly journals Assessing the effect of data integration on predictive ability of cancer survival models

2019 ◽  
Vol 26 (1) ◽  
pp. 8-20 ◽  
Author(s):  
Yi Guo ◽  
Jiang Bian ◽  
Francois Modave ◽  
Qian Li ◽  
Thomas J George ◽  
...  
Keyword(s):  
Data Integration ◽  
Cancer Survival ◽  
Prediction Models ◽  
Survival Rates ◽  
Model Performance ◽  
Predictive Ability ◽  
The United States ◽  
Cancer Prognosis ◽  
Survival Models ◽  
Survival Prediction

Cancer is the second leading cause of death in the United States. To improve cancer prognosis and survival rates, a better understanding of multi-level contributory factors associated with cancer survival is needed. However, prior research on cancer survival has primarily focused on factors from the individual level due to limited availability of integrated datasets. In this study, we sought to examine how data integration impacts the performance of cancer survival prediction models. We linked data from four different sources and evaluated the performance of Cox proportional hazard models for breast, lung, and colorectal cancers under three common data integration scenarios. We showed that adding additional contextual-level predictors to survival models through linking multiple datasets improved model fit and performance. We also showed that different representations of the same variable or concept have differential impacts on model performance. When building statistical models for cancer outcomes, it is important to consider cross-level predictor interactions.

scholarly journals Exploring Pathway-Based Group Lasso for Cancer Survival Analysis: A Special Case of Multi-Task Learning

2021 ◽  
Vol 12 ◽  
Author(s):  
Gabriela Malenová ◽  
Daniel Rowson ◽  
Valentina Boeva
Keyword(s):  
Cancer Survival ◽  
Group Lasso ◽  
Small Sample ◽  
Survival Models ◽  
Biological Knowledge ◽  
Cancer Data ◽  
Task Learning ◽  
Lasso Penalty ◽  
Cancer Types ◽  
Latent Group

Motivation: The Cox proportional hazard models are widely used in the study of cancer survival. However, these models often meet challenges such as the large number of features and small sample sizes of cancer data sets. While this issue can be partially solved by applying regularization techniques such as lasso, the models still suffer from unsatisfactory predictive power and low stability.Methods: Here, we investigated two methods to improve survival models. Firstly, we leveraged the biological knowledge that groups of genes act together in pathways and regularized both at the group and gene level using latent group lasso penalty term. Secondly, we designed and applied a multi-task learning penalty that allowed us leveraging the relationship between survival models for different cancers.Results: We observed modest improvements over the simple lasso model with the inclusion of latent group lasso penalty for six of the 16 cancer types tested. The addition of a multi-task penalty, which penalized coefficients in pairs of cancers from diverging too greatly, significantly improved accuracy for a single cancer, lung squamous cell carcinoma, while having minimal effect on other cancer types.Conclusion: While the use of pathway information and multi-tasking shows some promise, these methods do not provide a substantial improvement when compared with standard methods.

Integration of gene interaction information into a reweighted Lasso-Cox model for accurate survival prediction

2020 ◽  
Author(s):  
Wei Wang ◽  
Wei Liu
Keyword(s):  
Cox Model ◽  
Gene Interaction ◽  
Supplementary Information ◽  
Survival Prediction ◽  
Gene Interaction Network ◽  
Clinical Cancer Research ◽  
Prognostic Performance ◽  
Interaction Information ◽  
Cancer Types ◽  
Risk Of Cancer

Abstract Motivation Accurately predicting the risk of cancer patients is a central challenge for clinical cancer research. For high-dimensional gene expression data, Cox proportional hazard model with the least absolute shrinkage and selection operator for variable selection (Lasso-Cox) is one of the most popular feature selection and risk prediction algorithms. However, the Lasso-Cox model treats all genes equally, ignoring the biological characteristics of the genes themselves. This often encounters the problem of poor prognostic performance on independent datasets. Results Here, we propose a Reweighted Lasso-Cox (RLasso-Cox) model to ameliorate this problem by integrating gene interaction information. It is based on the hypothesis that topologically important genes in the gene interaction network tend to have stable expression changes. We used random walk to evaluate the topological weight of genes, and then highlighted topologically important genes to improve the generalization ability of the RLasso-Cox model. Experiments on datasets of three cancer types showed that the RLasso-Cox model improves the prognostic accuracy and robustness compared with the Lasso-Cox model and several existing network-based methods. More importantly, the RLasso-Cox model has the advantage of identifying small gene sets with high prognostic performance on independent datasets, which may play an important role in identifying robust survival biomarkers for various cancer types. Availability and implementation http://bioconductor.org/packages/devel/bioc/html/RLassoCox.html Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals Learning mutational signatures and their multidimensional genomic properties with TensorSignatures

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Harald Vöhringer ◽  
Arne Van Hoeck ◽  
Edwin Cuppen ◽  
Moritz Gerstung
Keyword(s):  
Somatic Hypermutation ◽  
Excision Repair ◽  
Metastatic Cancer ◽  
Strand Break ◽  
Lymphoid Leukaemia ◽  
Mutational Signatures ◽  
Transcription Start Sites ◽  
Nucleotide Excision ◽  
Cancer Types ◽  
Underlying Processes

AbstractWe present TensorSignatures, an algorithm to learn mutational signatures jointly across different variant categories and their genomic localisation and properties. The analysis of 2778 primary and 3824 metastatic cancer genomes of the PCAWG consortium and the HMF cohort shows that all signatures operate dynamically in response to genomic states. The analysis pins differential spectra of UV mutagenesis found in active and inactive chromatin to global genome nucleotide excision repair. TensorSignatures accurately characterises transcription-associated mutagenesis in 7 different cancer types. The algorithm also extracts distinct signatures of replication- and double strand break repair-driven mutagenesis by APOBEC3A and 3B with differential numbers and length of mutation clusters. Finally, TensorSignatures reproduces a signature of somatic hypermutation generating highly clustered variants at transcription start sites of active genes in lymphoid leukaemia, distinct from a general and less clustered signature of Polη-driven translesion synthesis found in a broad range of cancer types. In summary, TensorSignatures elucidates complex mutational footprints by characterising their underlying processes with respect to a multitude of genomic variables.

scholarly journals Analyzing Predictors of Control Measures and Psychosocial Problems Associated with COVID-19 Pandemic: Evidence from Eight Countries

2021 ◽  
Vol 11 (8) ◽  
pp. 106
Author(s):  
Sheikh Saifur Rahman Jony ◽  
Ubydul Haque ◽  
Nathaniel J. Webb ◽  
Emily Spence ◽  
Md. Siddikur Rahman ◽  
...  
Keyword(s):  
Psychological Health ◽  
Age Groups ◽  
Psychosocial Problems ◽  
Relevant Information ◽  
The United States ◽  
Control Measures ◽  
Cross Sectional Survey ◽  
Online Questionnaire ◽  
Cross Sectional ◽  
Primary Determinant

COVID-19 has harshly impacted communities globally. This study provides relevant information for creating equitable policy interventions to combat the spread of COVID-19. This study aims to predict the knowledge, attitude, and practice (KAP) of the COVID-19 pandemic at a global level to determine control measures and psychosocial problems. A cross-sectional survey was conducted from July to October 2020 using an online questionnaire. Questionnaires were initially distributed to academicians worldwide. These participants distributed the survey among their social, professional, and personal groups. Responses were collected and analyzed from 67 countries, with a sample size of 3031. Finally, based on the number of respondents, eight countries, including Bangladesh, China, Japan, Malaysia, Mexico, Pakistan, the United States, and Zambia were rigorously analyzed. Specifically, questionnaire responses related to COVID-19 accessibility, behavior, knowledge, opinion, psychological health, and susceptibility were collected and analyzed. As per our analysis, age groups were found to be a primary determinant of behavior, knowledge, opinion, psychological health, and susceptibility scores. Gender was the second most influential determinant for all metrics except information about COVID-19 accessibility, for which education was the second most important determinant. Respondent profession was the third most important metric for all scores. Our findings suggest that greater encouragement from government health authorities and the promotion of health education and policies are essential in the dissemination of COVID-19-awareness and increased control of the spread of COVID-19.

scholarly journals Analysis of treatment adherence and cost among patients with epilepsy: a four‐year retrospective cohort study in Pakistan

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Muhammad Arif Asghar ◽  
Ahad Abdul Rehman ◽  
Muhammad Liaquat Raza ◽  
Yousra Shafiq ◽  
Muhammad Asif Asghar
Keyword(s):  
Associated Factors ◽  
Cox Regression ◽  
Tertiary Care ◽  
The United States ◽  
Initial Therapy ◽  
Study Cohort ◽  
Entire Study ◽  
Cox Regression Analysis ◽  
Recommended Treatment ◽  
Patients With Epilepsy

Abstract Background The adherence pattern of antiepileptic drugs (AEDs) among patients with epilepsy is relatively lower in the United States and different European countries. However, adherence and cost analysis of AEDs in Asian countries have not been thoroughly studied. Therefore, the present study aimed to analyze the cost and adherence of AEDs and its associated factors in patients followed in Pakistan. Methods Data from prescriptions collected from patients with epilepsy who have visited the Outpatient Department (OPD) of different tertiary care hospitals at the cosmopolitan city of Karachi, Pakistan from December 2015 to November 2019. The mean follow-up period for each participant was about 22 months. Pairwise comparisons from Cox regression/hazard ratios were used to assess the predictors of adherence. Direct costs of AEDs were calculated and presented as the annual cost of drugs. Results A total of 11,490 patients were included in this study, 51.2 % were male and 48.8 % were female with a mean age of 45.2 ± 15.8 y. Levetiracetam was found as the most prescribing AED in all study participants (32.9 %). Of them, 49.1 % of patients continued their initial recommended treatment. However, 31.3 % of patients have discontinued the therapy, while, 19.6 % were switched to other AED. Adherence with initial treatment was more profound in male (57.4 %) patients, compared to female with a mean age of 44.2 years. Lamotrigine users (60.6 %) showed a higher tendency to retain on initially prescribed drugs. The total cost of epilepsy treatment in the entire study cohort was 153280.5 PKR ($941.9). By applying the Cox regression analysis, it can be observed that the patients with increasing age (OR, 2.04), migraine (OR, 2.21), psychiatric disorders (OR, 4.28), other comorbidities (OR, 1.52) and users of other than top five prescribing AEDs (2.35) were at higher risk of treatment discontinuation. However, levetiracetam (OR, 0.69), valproic acid (OR, 0.52), carbamazepine (OR, 0.81), lamotrigine (OR, 0.80) or lacosamide (OR, 0.65) users have more chances to continue their initial therapy. Conclusions Similar to western countries, the majority of patients with epilepsy exhibited low adherence with AEDs. Various associated factors for improving adherence were identified in this study.

scholarly journals Spectrum of Neonatal Surgical Problems in the Newborn: Incidence, Clinical Course, Immediate Outcome, Follow up Study in a Tertiary Care NICU

2019 ◽  
pp. 1-9
Author(s):  
K. G. Sachin ◽  
K. R. Sachin ◽  
H. Ramesh ◽  
Guru Prasad ◽  
Harsha Bullapur
Keyword(s):  
Congenital Anomalies ◽  
Congenital Malformations ◽  
Population Control ◽  
Tertiary Care ◽  
Relevant Information ◽  
The United States ◽  
Physical Structure ◽  
Neonatal Deaths ◽  
Medical College

Background: A congenital anomaly may be defined in terms of physical structure as a malformation, an abnormality of physical structure or form usually found at birth or during the first few weeks of life. Congenital anomalies affect approximately 1 in 33 infants and result in approximately 3.2 million birth defect-related disabilities every year. Congenital anomalies or birth defects are relatively common, affecting 3% to 5% of live births in the United States (US) and 2.1% in Europe. Congenital anomalies account for 8% to 15% of perinatal deaths and 13% to 16% of neonatal deaths in India. Objectives: To provide an insight on the burden and types of surgical problems encountered in our NICU of Bapuji Child Health Institute & Research Center, JJM Medical College, Davangere, Karnataka, India and to study the incidence, clinical profile and outcome of surgical condition. Methodology: A total of 3820 babies were examined over a period of 2 years. The relevant information was documented on a semi-structured proforma and analysed. Results: Overall incidence of congenital malformations at birth was 24.8 per 1000 births. The GIT system (51.58%) was most commonly involved followed by respiratory system (26.32%). The incidence of congenital malformation was more in male babies than female babies. Increased frequency was seen in babies born to mothers between 26–30 years & primigravida. The factors which significantly increased the rate of congenital malformations were consanguinity in parents & bad obstetric history. Out of 95 cases, 72% got discharged normally, 18% died in NICU and 10% got discharged against medical advise. Conclusion: With emphasis on “small family” norms and population control it is necessary to identify malformations so that interventional programmes can be planned. Systematic clinical examination of newborns for early detection of anomalies that may warrant medical or surgical intervention. Accurate antenatal anomaly scan need to be done to identify major malformations and terminate the pregnancy.

scholarly journals Explainable Transformer-Based Neural Network forthe Prediction of Survival Outcomes in Non-SmallCell Lung Cancer (NSCLC)

2021 ◽  
Author(s):  
Gustavo Arango ◽  
Elly Kipkogei ◽  
Etai Jacob ◽  
Ioannis Kagiampakis ◽  
Arijit Patra
Keyword(s):  
Immune System ◽  
Precision Medicine ◽  
Cancer Patients ◽  
Patient Survival ◽  
Response To Therapy ◽  
Survival Models ◽  
Attractive Alternative ◽  
Survival Prediction ◽  
Molecular Features ◽  
Clinical Measurements

In this paper, we introduce the Clinical Transformer - a recasting of the widely used transformer architecture as a method for precision medicine to model relations between molecular and clinical measurements, and the survival of cancer patients. Although the emergence of immunotherapy offers a new hope for cancer patients with dramatic and durable responses having been reported, only a subset of patients demonstrate benefit. Such treatments do not directly target the tumor but recruit the patient immune system to fight the disease. Therefore, the response to therapy is more complicated to understand as it is affected by the patients physical condition, immune system fitness and the tumor. As in text, where the semantics of a word is dependent on the context of the sentence it belongs to, in immuno-therapy a biomarker may have limited meaning if measured independent of other clinical or molecular features. Hence, we hypothesize that the transformer-inspired model may potentially enable effective modelling of the semantics of different biomarkers with respect to patient survival time. Herein, we demonstrate that this approach can offer an attractive alternative to the survival models utilized incurrent practices as follows: (1) We formulate an embedding strategy applied to molecular and clinical data obtained from the patients. (2) We propose a customized objective function to predict patient survival. (3) We show the applicability of our proposed method to bioinformatics and precision medicine. Applying the clinical transformer to several immuno-oncology clinical studies, we demonstrate how the clinical transformer outperforms other linear and non-linear methods used in current practice for survival prediction. We also show that when initializing the weights of a domain-specific transformer by the weights of a cross-domain transformer, we further improve the predictions. Lastly, we show how the attention mechanism successfully captures some of the known biology behind these therapies

scholarly journals C-mix: A high-dimensional mixture model for censored durations, with applications to genetic data

2018 ◽  
Vol 28 (5) ◽  
pp. 1523-1539
Author(s):  
Simon Bussy ◽  
Agathe Guilloux ◽  
Stéphane Gaïffas ◽  
Anne-Sophie Jannot
Keyword(s):  
Mixture Model ◽  
Proportional Hazards ◽  
Expectation Maximization Algorithm ◽  
Cox Proportional Hazards ◽  
Elastic Net ◽  
Survival Models ◽  
High Dimensional ◽  
Survival Prediction ◽  
Monte Carlo Simulation Study ◽  
Cox Proportional Hazards Models

We introduce a supervised learning mixture model for censored durations (C-mix) to simultaneously detect subgroups of patients with different prognosis and order them based on their risk. Our method is applicable in a high-dimensional setting, i.e. with a large number of biomedical covariates. Indeed, we penalize the negative log-likelihood by the Elastic-Net, which leads to a sparse parameterization of the model and automatically pinpoints the relevant covariates for the survival prediction. Inference is achieved using an efficient Quasi-Newton Expectation Maximization algorithm, for which we provide convergence properties. The statistical performance of the method is examined on an extensive Monte Carlo simulation study and finally illustrated on three publicly available genetic cancer datasets with high-dimensional covariates. We show that our approach outperforms the state-of-the-art survival models in this context, namely both the CURE and Cox proportional hazards models penalized by the Elastic-Net, in terms of C-index, AUC( t) and survival prediction. Thus, we propose a powerful tool for personalized medicine in cancerology.

Sign in / Sign up

Export Citation Format

Share Document