scholarly journals Prediction of Incident Cancers in the Lifelines Population-Based Cohort

Cancers ◽  
2021 ◽  
Vol 13 (9) ◽  
pp. 2133
Author(s):  
Francisco O. Cortés-Ibañez ◽  
Sunil Belur Nagaraj ◽  
Ludo Cornelissen ◽  
Gerjan J. Navis ◽  
Bert van der Vegt ◽  
...  
Keyword(s):  
Cancer Incidence ◽  
Binary Classification ◽  
Predictive Performance ◽  
Population Based ◽  
Support Vector ◽  
Clinical Variables ◽  
Incident Cancer ◽  
History Of ◽  
Diagnosis Of Cancer ◽  
Auc Value

Cancer incidence is rising, and accurate prediction of incident cancers could be relevant to understanding and reducing cancer incidence. The aim of this study was to develop machine learning (ML) models that could predict an incident diagnosis of cancer. Participants without any history of cancer within the Lifelines population-based cohort were followed for a median of 7 years. Data were available for 116,188 cancer-free participants and 4232 incident cancer cases. At baseline, socioeconomic, lifestyle, and clinical variables were assessed. The main outcome was an incident cancer during follow-up (excluding skin cancer), based on linkage with the national pathology registry. The performance of three ML algorithms was evaluated using supervised binary classification to identify incident cancers among participants. Elastic net regularization and Gini index were used for variables selection. An overall area under the receiver operator curve (AUC) <0.75 was obtained, the highest AUC value was for prostate cancer (random forest AUC = 0.82 (95% CI 0.77–0.87), logistic regression AUC = 0.81 (95% CI 0.76–0.86), and support vector machines AUC = 0.83 (95% CI 0.78–0.88), respectively); age was the most important predictor in these models. Linear and non-linear ML algorithms including socioeconomic, lifestyle, and clinical variables produced a moderate predictive performance of incident cancers in the Lifelines cohort.

scholarly journals Machine learning approaches revealed metabolic signatures of incident chronic kidney disease in persons with pre-and type 2 diabetes

2020 ◽  
Author(s):  
Ada Admin ◽  
Jialing Huang ◽  
Cornelia Huth ◽  
Marcela Covic ◽  
Martina Troll ◽  
...  
Keyword(s):  
Machine Learning ◽  
Type 2 Diabetes ◽  
Chronic Kidney Disease ◽  
Kidney Disease ◽  
Characteristic Curve ◽  
Predictive Performance ◽  
Population Based ◽  
Learning Approaches ◽  
Clinical Variables

Early and precise identification of individuals with pre-diabetes and type 2 diabetes (T2D) at risk of progressing to chronic kidney disease (CKD) is essential to prevent complications of diabetes. Here, we identify and evaluate prospective metabolite biomarkers and the best set of predictors of CKD in the longitudinal, population-based Cooperative Health Research in the Region of Augsburg (KORA) cohort by targeted metabolomics and machine learning approaches. Out of 125 targeted metabolites, sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0 were identified as candidate metabolite biomarkers of incident CKD specifically in hyperglycemic individuals followed during 6.5 years. Sets of predictors for incident CKD developed from 125 metabolites and 14 clinical variables showed highly stable performances in all three machine learning approaches and outperformed the currently established clinical algorithm for CKD. The two metabolites in combination with five clinical variables were identified as the best set of predictors and their predictive performance yielded a mean area value under the receiver operating characteristic curve of 0.857. The inclusion of metabolite variables in the clinical prediction of future CKD may thus improve the risk prediction in persons with pre- and T2D. The metabolite link with hyperglycemia-related early kidney dysfunction warrants further investigation.

The Genes and Genetics of Malignant Melanoma

2002 ◽  
Vol 6 (3) ◽  
pp. 229-235 ◽  
Author(s):  
Peter Gibbs ◽  
Benjamin M. R. Brady ◽  
William A. Robinson
Keyword(s):  
High Risk ◽  
Sun Exposure ◽  
Population Based ◽  
Uv Exposure ◽  
Molecular Defect ◽  
Clinical Variables ◽  
Red Hair ◽  
Increased Risk ◽  
History Of ◽  
Self Examination

Background: Population-based studies have identified several clinical variables associated with an increased risk of developing cutaneous melanoma that include phenotype, amount of and response to sun exposure, and family history. However, these observations are of limited relevance to clinical practice as the risk associated with each factor is individually modest and the characteristics of these variables lack precision when applied to a particular individual. Objective: To review the literature regarding recent advances made in the understanding of the genes and genetics of clinical variables associated with an increased risk of melanoma. Conclusion: Variants of the MC1R (melanocortin-1 receptor) have been identified as major determinants of high-risk phenotypes, such as red hair and pale skin, and the ability to tan in response to UV exposure. Several studies also suggest that such variants may increase melanoma risk independent of their contribution to phenotype. A strong genetic basis for both nevus density and size has been demonstrated and the link between nevi and the development of MM has become better defined. Finally, germline defects in several genes involved in cell cycle regulation, namely, p16 and CDK4, have been demonstrated in many familial melanoma kindreds. This progress has introduced the prospect of genetic testing as a means of identifying a limited number of high-risk individuals who can be targeted with regular screening and education regarding UV exposure and skin self-examination. Ultimately, through rational genetic therapy targeted to correcting the underlying molecular defect, altering the natural history of melanoma development may be possible.

scholarly journals Machine learning approaches revealed metabolic signatures of incident chronic kidney disease in persons with pre-and type 2 diabetes

2020 ◽  
Author(s):  
Ada Admin ◽  
Jialing Huang ◽  
Cornelia Huth ◽  
Marcela Covic ◽  
Martina Troll ◽  
...  
Keyword(s):  
Machine Learning ◽  
Type 2 Diabetes ◽  
Chronic Kidney Disease ◽  
Kidney Disease ◽  
Characteristic Curve ◽  
Predictive Performance ◽  
Population Based ◽  
Learning Approaches ◽  
Clinical Variables

Early and precise identification of individuals with pre-diabetes and type 2 diabetes (T2D) at risk of progressing to chronic kidney disease (CKD) is essential to prevent complications of diabetes. Here, we identify and evaluate prospective metabolite biomarkers and the best set of predictors of CKD in the longitudinal, population-based Cooperative Health Research in the Region of Augsburg (KORA) cohort by targeted metabolomics and machine learning approaches. Out of 125 targeted metabolites, sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0 were identified as candidate metabolite biomarkers of incident CKD specifically in hyperglycemic individuals followed during 6.5 years. Sets of predictors for incident CKD developed from 125 metabolites and 14 clinical variables showed highly stable performances in all three machine learning approaches and outperformed the currently established clinical algorithm for CKD. The two metabolites in combination with five clinical variables were identified as the best set of predictors and their predictive performance yielded a mean area value under the receiver operating characteristic curve of 0.857. The inclusion of metabolite variables in the clinical prediction of future CKD may thus improve the risk prediction in persons with pre- and T2D. The metabolite link with hyperglycemia-related early kidney dysfunction warrants further investigation.

scholarly journals Personal history of psoriasis and risk of incident cancer among women: a population-based cohort study

2016 ◽  
Vol 174 (5) ◽  
pp. 1108-1111 ◽  
Author(s):  
W-Q. Li ◽  
J. Han ◽  
E. Cho ◽  
S. Wu ◽  
H. Dai ◽  
...  
Keyword(s):  
Cohort Study ◽  
Population Based ◽  
Personal History ◽  
Incident Cancer ◽  
History Of

Supervised Learning for Binary Classification on US Adult Income

2021 ◽  
Vol 13 (2) ◽  
pp. 80-91
Author(s):  
Li-Pang Chen
Keyword(s):  
Missing Values ◽  
Binary Classification ◽  
Roc Curves ◽  
Income Level ◽  
Support Vector ◽  
Classification Methods ◽  
Gender Education ◽  
The Us ◽  
Auc Value ◽  
R Functions

In this project, various binary classification methods have been used to make predictions about US adult income level in relation to social factors including age, gender, education, and marital status. We first explore descriptive statistics for the dataset and deal with missing values. After that, we examine some widely used classification methods, including logistic regression, discriminant analysis, support vector machine, random forest, and boosting. Meanwhile, we also provide suitable R functions to demonstrate applications. Various metrics such as ROC curves, accuracy, recall and F-measure are calculated to compare the performance of these models. We find the boosting is the best method in our data analysis due to its highest AUC value and the highest prediction accuracy. In addition, among all predictor variables, we also find three variables that have the largest impact on the US adult income level.

Plasma adiponectin to predict incident cancer in a large multiethnic population-based cohort study.

2013 ◽  
Vol 31 (15_suppl) ◽  
pp. 1560-1560
Author(s):  
Muhammad Shaalan Beg ◽  
Sadia Saleem ◽  
Aslan Turer ◽  
Colby Ayers ◽  
Song Zhang ◽  
...  
Keyword(s):  
Insulin Resistance ◽  
Cohort Study ◽  
Cancer Incidence ◽  
Population Based ◽  
Minority Population ◽  
Cancer Group ◽  
Incident Cancer ◽  
Diabetes Status ◽  
Biomarker Analysis ◽  
Multiethnic Population

1560 Background: Previous retrospective studies have suggested an inverse relationship between cancer and markers of insulin resistance, including adiponectin. Whether circulating levels of adiponectin are associated with future cancer incidence has not been well demonstrated. As such, we examined the association of adiponectin on the risk of future cancer development in a large population based cohort study. Methods: The Dallas Heart Study (DHS) is a multiethnic population based study of 3,531 Dallas County residents with an intentional oversampling of minorities which are underrepresented in other population studies. Incident and prevalent cancer cases were identified in the DHS by the Texas Cancer Registry. Cases were defined as ‘incident’ if they were diagnosed 24 months after enrollment into DHS. Fasting blood samples were obtained from DHS participants between 2000 and 2002, and used for biomarker analysis. Univariate and multivariate analysis was performed to test the association of incident cancer and adiponectin after adjusting for age, diabetes status, gender, ethnicity and body mass index (BMI). Results: Of 3,531 individuals, 150 incident cancer cases were diagnosed. The study population comprised 55.8% females, 51.5% were black and 39.6% had a BMI above 30 kg/m2. Mean adiponectin was 8.12 μg/ml in incident cancer group and 7.46 μg/ml in non-cancer group (p=0.129). In multivariable analysis, adiponectin was not an independent predictor of cancer incidence after adjusting for age, diabetes status, gender, ethnicity and BMI (p=0.167) Conclusions: Previous population based studies have demonstrated conflicting results on the association of markers insulin resistance, including adiponectin, and cancer incidence. In this study of a predominant ethnic minority population, an association between adiponectin and cancer incidence was not seen. Further prospective studies are needed to understand the association between markers of insulin resistance and risk of future cancer risk.

scholarly journals Machine learning approaches revealed metabolic signatures of incident chronic kidney disease in persons with pre-and type 2 diabetes

2020 ◽  
Author(s):  
Ada Admin ◽  
Jialing Huang ◽  
Cornelia Huth ◽  
Marcela Covic ◽  
Martina Troll ◽  
...  
Keyword(s):  
Machine Learning ◽  
Type 2 Diabetes ◽  
Chronic Kidney Disease ◽  
Kidney Disease ◽  
Characteristic Curve ◽  
Predictive Performance ◽  
Population Based ◽  
Learning Approaches ◽  
Clinical Variables

Early and precise identification of individuals with pre-diabetes and type 2 diabetes (T2D) at risk of progressing to chronic kidney disease (CKD) is essential to prevent complications of diabetes. Here, we identify and evaluate prospective metabolite biomarkers and the best set of predictors of CKD in the longitudinal, population-based Cooperative Health Research in the Region of Augsburg (KORA) cohort by targeted metabolomics and machine learning approaches. Out of 125 targeted metabolites, sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0 were identified as candidate metabolite biomarkers of incident CKD specifically in hyperglycemic individuals followed during 6.5 years. Sets of predictors for incident CKD developed from 125 metabolites and 14 clinical variables showed highly stable performances in all three machine learning approaches and outperformed the currently established clinical algorithm for CKD. The two metabolites in combination with five clinical variables were identified as the best set of predictors and their predictive performance yielded a mean area value under the receiver operating characteristic curve of 0.857. The inclusion of metabolite variables in the clinical prediction of future CKD may thus improve the risk prediction in persons with pre- and T2D. The metabolite link with hyperglycemia-related early kidney dysfunction warrants further investigation.

scholarly journals Differential risk of incident cancer in patients with heart failure: A nationwide population-based cohort study

2020 ◽  
Author(s):  
Soongu Kwak ◽  
Soonil Kwon ◽  
Seo-Young Lee ◽  
Seokhun Yang ◽  
Hyun-Jung Lee ◽  
...  
Keyword(s):  
Heart Failure ◽  
Hematologic Malignancy ◽  
Population Based ◽  
Cancer Development ◽  
Site Specific ◽  
Incident Cancer ◽  
Increased Risk ◽  
History Of ◽  
Cox Analysis ◽  
History Of Cancer

ABSTRACTBackgroundHeart failure (HF) and cancer are currently two leading causes of mortality, and sometimes coexist. However, the relationship between them is not completely elucidated. We aimed to investigate whether patients with HF are predisposed to cancer development using the large Korean National Health Insurance claims database.Methods and findingsThis study included 128,441 HF patients without a history of cancer and 642,205 age- and sex-matched individuals with no history of cancer and HF between 1 January 2010 and 31 December 2015. During a median follow-up of 4.06 years, 11,808 patients from the HF group and 40,805 participants from the control were newly diagnosed with cancer (cumulative incidence, 9.2% vs. 6.4%, p<0.0001). Patients with HF presented a higher risk for cancer development compared to controls in multivariable Cox analysis (hazard ratio [HR] 1.64, 95% confidence interval [CI] 1.61 - 1.68). The increased risk was consistent for all site-specific cancers. To minimize potential surveillance bias, additional analysis was performed by eliminating participants who developed cancer within the initial 2 years of HF diagnosis (i.e. 2-year lag analysis). In the 2-year lag analysis, the higher risk of overall cancer remained significant in patients with HF (HR 1.09, 95% CI 1.05 - 1.13), although the association was weaker. Among the site-specific cancers, three types of cancer (lung, liver/biliary/pancreas, and hematologic malignancy) were consistently at higher risk in patients with HF.ConclusionsCancer incidence is higher in patients with HF than in the general population. Active surveillance of coexisting malignancy needs to be considered in these patients.

scholarly journals Cancer Incidence in Egypt: Results of the National Population-Based Cancer Registry Program

2014 ◽  
Vol 2014 ◽  
pp. 1-18 ◽  
Author(s):  
Amal S. Ibrahim ◽  
Hussein M. Khaled ◽  
Nabiel NH Mikhail ◽  
Hoda Baraka ◽  
Hossam Kamel
Keyword(s):  
Cancer Registry ◽  
Cancer Incidence ◽  
Tertiary Care ◽  
National Level ◽  
Data Entry ◽  
Fold Increase ◽  
Population Based ◽  
Incidence Rates ◽  
Cancer Breast ◽  
Incident Cancer

Background. This paper aims to present cancer incidence rates at national and regional level of Egypt, based upon results of National Cancer Registry Program (NCRP).Methods. NCRP stratified Egypt into 3 geographical strata: lower, middle, and upper. One governorate represented each region. Abstractors collected data from medical records of cancer centers, national tertiary care institutions, Health Insurance Organization, Government-Subsidized Treatment Program, and death records. Data entry was online. Incidence rates were calculated at a regional and a national level. Future projection up to 2050 was also calculated.Results. Age-standardized incidence rates per 100,000 were 166.6 (both sexes), 175.9 (males), and 157.0 (females). Commonest sites were liver (23.8%), breast (15.4%), and bladder (6.9%) (both sexes): liver (33.6%) and bladder (10.7%) among men, and breast (32.0%) and liver (13.5%) among women. By 2050, a 3-fold increase in incident cancer relative to 2013 was estimated.Conclusion. These data are the only available cancer rates at national and regional levels of Egypt. The pattern of cancer indicated the increased burden of liver cancer. Breast cancer occupied the second rank. Study of rates of individual sites of cancer might help in giving clues for preventive programs.

Proximity to oil refineries and risk of cancer: A population-based analysis.

2020 ◽  
Vol 38 (15_suppl) ◽  
pp. e13586-e13586
Author(s):  
Stephen B. Williams ◽  
Yong Shan ◽  
Usama Jazzar ◽  
Preston Kerr ◽  
Ikenna Okereke ◽  
...  
Keyword(s):  
Cancer Incidence ◽  
Systemic Disease ◽  
Population Based ◽  
Oil Refinery ◽  
Multiple Cancer ◽  
Oil Refineries ◽  
Incident Cancer ◽  
Increased Risk ◽  
Cancer Types ◽  
Risk Of Cancer

e13586 Background: Proximity to oil refineries and cancer incidence is largely unknown. We sought to compare the incidence of cancer (bladder, breast, colon, lung, lymphoma, and prostate) according to proximity to an oil refinery in the State of Texas. Methods: A total of 6,302,265 persons aged ≥20 years from January 1, 2001 through December 31, 2014 were identified. We used zero-inflated Poisson regression models to examine the association of proximity to an oil refinery with cancer incidence. Results: We observed that proximity to an oil refinery was associated with a significantly increased risk of incident cancer diagnosis across all cancer types. For example, persons residing within 0-10 (Risk Ratio (RR) 1.16, 95% Confidence Interval (CI) 1.13-1.19) and 11-20 (RR 1.08, 95% CI 1.05-1.11) miles were significantly more likely to be diagnosed with lymphoma than individuals who lived within 21-30 miles from an oil refinery. We also observed differences in stage of cancer at diagnosis according to proximity to an oil refinery. We also found persons residing within 0-10 miles were more likely to be diagnosed with distant metastasis and/or systemic disease than people residing 21-30 miles from an oil refinery. The greatest risk of distant disease was observed in patients diagnosed with bladder cancer living within 0-10 vs. 21-30 miles (RR 1.33, 95% CI 1.06-1.68), respectively. Conclusions: Proximity to an oil refinery was associated with an increased risk of multiple cancer types. We also observed significantly increased risk of regional and distant/metastatic disease according to proximity to an oil refinery.

Sign in / Sign up

Export Citation Format

Share Document