Machine learning of clinical variables and coronary artery calcium scoring for the prediction of obstructive coronary artery disease on coronary computed tomography angiography: analysis from the CONFIRM registry

Al’Aref, Subhi J.; Maliakal, Gabriel; Singh, Gurpreet; Rosendael, Alexander R. van; Ma, Xiaoyue; Xu, Zhuoran; Lee, Benjamin; Pandey, Mohit; Achenbach, Stephan; Al‐Mallah, Mouaz H.; Andreini, Daniele; Bax, Jeroen J.; Berman, Daniel S.; Budoff, Matthew J.; Cademartiri, Filippo; Callister, Tracy Q.; Chang, Hyuk Jae; Chinnaiyan, Kavitha M.; Chow, Benjamin J.W.; Cury, Ricardo C.; Delago, Augustin; Feuchtner, Gudrun; Hadamitzky, Martin; Hausleiter, Jöerg; Kaufmann, Philipp A.; Kim, Yong Jin; Leipsic, Jonathon; Maffei, Erica; Marques, Hugo; Gonçalves, Pedro de Araújo; Pontone, Gianluca; Raff, Gilbert; Rubinshtein, Ronen; Villines, Todd C.; Gransar, Heidi; Lu, Yao; Jones, Edward; Peña, Jessica M.; Lin, Fay Y.; Min, James K.; Shaw, Leslee J.

doi:10.1093/eurheartj/ehz565

Cited by 145 publications

(103 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…XGBoost is a novel, state-of-the-art machine learning algorithm that has been shown to outperform other more traditional algorithms in its accuracy and efficiency [ 12 ]. It can also take both continuous and discrete inputs and handle sparse data, in addition to having highly optimizable hyper-parameters [ 15 ].…”

Section: Resultsmentioning

confidence: 99%

Using machine learning of clinical data to diagnose COVID-19: a systematic review and meta-analysis

Shende

et al. 2020

BMC Med Inform Decis Mak

128

View full text Add to dashboard Cite

Background The recent Coronavirus Disease 2019 (COVID-19) pandemic has placed severe stress on healthcare systems worldwide, which is amplified by the critical shortage of COVID-19 tests. Methods In this study, we propose to generate a more accurate diagnosis model of COVID-19 based on patient symptoms and routine test results by applying machine learning to reanalyzing COVID-19 data from 151 published studies. We aim to investigate correlations between clinical variables, cluster COVID-19 patients into subtypes, and generate a computational classification model for discriminating between COVID-19 patients and influenza patients based on clinical variables alone. Results We discovered several novel associations between clinical variables, including correlations between being male and having higher levels of serum lymphocytes and neutrophils. We found that COVID-19 patients could be clustered into subtypes based on serum levels of immune cells, gender, and reported symptoms. Finally, we trained an XGBoost model to achieve a sensitivity of 92.5% and a specificity of 97.9% in discriminating COVID-19 patients from influenza patients. Conclusions We demonstrated that computational methods trained on large clinical datasets could yield ever more accurate COVID-19 diagnostic models to mitigate the impact of lack of testing. We also presented previously unknown COVID-19 clinical variable correlations and clinical subgroups.

show abstract

Section: Resultsmentioning

confidence: 99%

Using machine learning of clinical data to diagnose COVID-19: a systematic review and meta-analysis

Shende

et al. 2020

BMC Med Inform Decis Mak

128

View full text Add to dashboard Cite

show abstract

“…With the progress of science and technology, arti cial intelligence methods have been widely used in the eld of medicine [29][30][31][32]. There is considerable research on machine learning methods in trauma [33][34][35].…”

Section: Discussionmentioning

confidence: 99%

Intelligent prediction of RBC demand in trauma patients using decision tree methods

Feng

Liu

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: The vital signs of trauma patients are complex and changeable, and the prediction of blood transfusion demand mainly depends on doctors' experience and trauma scoring system; therefore, it cannot be accurately predicted. In this study, a machine learning decision tree algorithm (classification and regression tree, CRT and eXtreme gradient boosting, XGBoost) was proposed for the demand prediction of traumatic blood transfusion to provide technical support for doctors. Methods: A total of 1,371 trauma patients who were diverted to the emergency department from January 2014 to January 2018 were collected from an emergency trauma database. The vital signs, laboratory examination parameters and blood transfusion volume were used as variables, and the non-invasive parameters and all (non-invasive + invasive) parameters were used to construct an intelligent prediction model for RBC demand by logistic regression (LR), CRT and XGBoost. The prediction accuracy of the model was compared with the area under the curve (AUC).Results: The studies we performed showed that non-invasive parameters were used to predict blood transfusion. The LR method was the best, with an AUC of 0.72 (95% confidence interval [CI] 0.657-0.775), which was higher than the CRT AUC of 0.69 (95% CI 0.633-0.751) and the XGBoost AUC of 0.71 (95% CI 0.654-0.756) (P<0.05). The trauma site and shock index are important prediction parameters. For all the prediction parameters, XGBoost was the best, with an AUC of 0.94 (95% CI 0.893-0.981), which was higher than the LR AUC of 0.80 (95% CI 0.744-0.850) and the CRT AUC of 0.82 (95% CI 0.779-0.853) (P<0.05). Haematocrit/Haemoglobin is an important prediction parameter. Conclusions: The classification performance of the intelligent prediction model constructed by the decision tree algorithm is not inferior to that of the traditional LR method. With the increase in the data quantity, the accuracy of the model improved in the iteration process, and the prediction performance continuously improved, which is conducive to clinical application and wide promotion.

show abstract

“…Machine and deep learning approaches involve using raw imaging data to extrapolate imaging features and, together with clinical information and big data, are processed into algorithms to build disease models and neural networks in order to improve diagnosis, make clinical decisions, and predict outcomes towards a personalized medical approach 81) . Recent studies on machine learning approaches exploiting CTCA showed high prognostic accuracy in risk stratification and prediction of mortality in CHD patients [82][83][84] . Furthermore, in a recent study, a novel algorithm for T1 mapping and extracellular volume fraction in CMR was tested, providing a better computational efficiency than other well-consolidated methods 85) .…”

Section: Personalized Therapy For Primary Prevention and Secondary Prmentioning

confidence: 99%

Network Medicine: A Clinical Approach for Precision Medicine and Personalized Therapy in Coronary Heart Disease

Infante

Viscovo

Rimini

et al. 2020

JAT

View full text Add to dashboard Cite

The official journal of the Japan Atherosclerosis Society and the Asian Pacific Society of Atherosclerosis and Vascular Diseases Review Early identification of coronary atherosclerotic pathogenic mechanisms is useful for predicting the risk of coronary heart disease (CHD) and future cardiac events. Epigenome changes may clarify a significant fraction of this "missing hereditability", thus offering novel potential biomarkers for prevention and care of CHD. The rapidly growing disciplines of systems biology and network science are now poised to meet the fields of precision medicine and personalized therapy. Network medicine integrates standard clinical recording and non-invasive, advanced cardiac imaging tools with epigenetics into deep learning for in-depth CHD molecular phenotyping. This approach could potentially explore developing novel drugs from natural compounds (i.e. polyphenols, folic acid) and repurposing current drugs, such as statins and metformin. Several clinical trials have exploited epigenetic tags and epigenetic sensitive drugs both in primary and secondary prevention. Due to their stability in plasma and easiness of detection, many ongoing clinical trials are focused on the evaluation of circulating miRNAs (e.g. miR-8059 and miR-320a) in blood, in association with imaging parameters such as coronary calcifications and stenosis degree detected by coronary computed tomography angiography (CCTA), or functional parameters provided by FFR/CT and PET/CT. Although epigenetic modifications have also been prioritized through network based approaches, the whole set of molecular interactions (interactome) in CHD is still under investigation for primary prevention strategies. events 10, 11). In the era of next generation sequencing, the development of a plethora of high-throughput platforms has provided significant impact concerning the knowledge of CHD molecular characterization generating big "omics" data sets 12-14). Furthermore, several non-invasive imaging techniques have been developed over the last few years to detect CHD with the aim to guide optimal patient management, such as coronary computed tomography angiography (CCTA), cardiac magnetic resonance (CMR), and nuclear medicine technique positron emission tomography (PET) 15). These advances have raised challenges Copyright©2020 Japan Atherosclerosis Society This article is distributed under the terms of the latest version of CC BY-NC-SA defined by the Creative Commons Attribution License.

show abstract

Machine learning of clinical variables and coronary artery calcium scoring for the prediction of obstructive coronary artery disease on coronary computed tomography angiography: analysis from the CONFIRM registry

Cited by 145 publications

References 36 publications

Using machine learning of clinical data to diagnose COVID-19: a systematic review and meta-analysis

Using machine learning of clinical data to diagnose COVID-19: a systematic review and meta-analysis

Intelligent prediction of RBC demand in trauma patients using decision tree methods

Network Medicine: A Clinical Approach for Precision Medicine and Personalized Therapy in Coronary Heart Disease

Contact Info

Product

Resources

About