1
|
Teshale AB, Htun HL, Vered M, Owen AJ, Freak-Poli R. A Systematic Review of Artificial Intelligence Models for Time-to-Event Outcome Applied in Cardiovascular Disease Risk Prediction. J Med Syst 2024; 48:68. [PMID: 39028429 PMCID: PMC11271333 DOI: 10.1007/s10916-024-02087-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/28/2024] [Accepted: 07/09/2024] [Indexed: 07/20/2024]
Abstract
Artificial intelligence (AI) based predictive models for early detection of cardiovascular disease (CVD) risk are increasingly being utilised. However, AI based risk prediction models that account for right-censored data have been overlooked. This systematic review (PROSPERO protocol CRD42023492655) includes 33 studies that utilised machine learning (ML) and deep learning (DL) models for survival outcome in CVD prediction. We provided details on the employed ML and DL models, eXplainable AI (XAI) techniques, and type of included variables, with a focus on social determinants of health (SDoH) and gender-stratification. Approximately half of the studies were published in 2023 with the majority from the United States. Random Survival Forest (RSF), Survival Gradient Boosting models, and Penalised Cox models were the most frequently employed ML models. DeepSurv was the most frequently employed DL model. DL models were better at predicting CVD outcomes than ML models. Permutation-based feature importance and Shapley values were the most utilised XAI methods for explaining AI models. Moreover, only one in five studies performed gender-stratification analysis and very few incorporate the wide range of SDoH factors in their prediction model. In conclusion, the evidence indicates that RSF and DeepSurv models are currently the optimal models for predicting CVD outcomes. This study also highlights the better predictive ability of DL survival models, compared to ML models. Future research should ensure the appropriate interpretation of AI models, accounting for SDoH, and gender stratification, as gender plays a significant role in CVD occurrence.
Collapse
Affiliation(s)
- Achamyeleh Birhanu Teshale
- School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC, Australia
- Department of Epidemiology and Biostatistics, Institute of Public Health, College of Medicine and Health Sciences, University of Gondar, Gondar, Ethiopia
| | - Htet Lin Htun
- School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC, Australia
| | - Mor Vered
- Department of Data Science and AI, Faculty of Information Technology, Monash University, Clayton, VIC, Australia
| | - Alice J Owen
- School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC, Australia
| | - Rosanne Freak-Poli
- School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC, Australia.
- Stroke and Ageing Research, Department of Medicine, School of Clinical Sciences at Monash Health, Monash University, Melbourne, VIC, Australia.
| |
Collapse
|
2
|
Stabellini N, Cullen J, Bittencourt MS, Moore JX, Sutton A, Nain P, Hamerschlak N, Weintraub NL, Dent S, Tsai M, Banerjee A, Ghosh AK, Sadler D, Coughlin SS, Barac A, Shanahan J, Montero AJ, Guha A. Allostatic Load/Chronic Stress and Cardiovascular Outcomes in Patients Diagnosed With Breast, Lung, or Colorectal Cancer. J Am Heart Assoc 2024; 13:e033295. [PMID: 38979791 PMCID: PMC11292743 DOI: 10.1161/jaha.123.033295] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 01/04/2024] [Accepted: 06/06/2024] [Indexed: 07/10/2024]
Abstract
BACKGROUND Cardiovascular disease and cancer share a common risk factor: chronic stress/allostatic load (AL). A 1-point increase in AL is linked to up to a 30% higher risk of major cardiac events (MACE) in patients with prostate cancer. However, AL's role in MACE in breast cancer, lung cancer, or colorectal cancer remains unknown. METHODS AND RESULTS Patients ≥18 years of age diagnosed with the mentioned 3 cancers of interest (2010-2019) and followed up at a large, hybrid academic-community practice were included in this retrospective cohort study. AL was modeled as an ordinal measure (0-11). Adjusted Fine-Gray competing risks regressions estimated the impact of AL precancer diagnosis on 2-year MACE (a composite of heart failure, ischemic stroke, acute coronary syndrome, and atrial fibrillation). The effect of AL changes over time on MACE was calculated via piecewise Cox regression (before, and 2 months, 6 months, and 1 year after cancer diagnosis). Among 16 467 patients, 50.5% had breast cancer, 27.9% had lung cancer, and 21.4% had colorectal cancer. A 1-point elevation in AL before breast cancer diagnosis corresponded to a 10% heightened associated risk of MACE (adjusted hazard ratio, 1.10 [95% CI, 1.06-1.13]). Similar findings were noted in lung cancer (adjusted hazard ratio, 1.16 [95% CI, 1.12-1.20]) and colorectal cancer (adjusted hazard ratio, 1.13 [95% CI, 1.08-1.19]). When considering AL as a time-varying exposure, the peak associated MACE risk occurred with a 1-point AL rise between 6 and 12 months post- breast cancer, lung cancer, and colorectal cancer diagnosis. CONCLUSIONS AL warrants investigation as a potential marker in these patients to identify those at elevated cardiovascular risk and intervene accordingly.
Collapse
Affiliation(s)
- Nickolas Stabellini
- Case Western Reserve University School of Medicine, Case Western Reserve UniversityClevelandOHUSA
- Department of Hematology‐OncologyUniversity Hospitals Seidman Cancer CenterClevelandOHUSA
- Faculdade Israelita de Ciências da Saúde Albert EinsteinHospital Israelita Albert EinsteinSão PauloSPBrazil
- Department of MedicineMedical College of Georgia at Augusta UniversityAugustaGAUSA
| | - Jennifer Cullen
- Case Western Reserve University School of Medicine, Case Western Reserve UniversityClevelandOHUSA
- Case Comprehensive Cancer CenterCase Western Reserve UniversityClevelandOHUSA
| | - Marcio S. Bittencourt
- Division of Cardiology, Department of MedicineUniversity of PittsburghPittsburghPAUSA
| | - Justin X. Moore
- Center for Health Equity Transformation, Department of Behavioral Science, Department of Internal Medicine, Markey Cancer CenterUniversity of Kentucky College of MedicineLexingtonKYUSA
| | - Arnethea Sutton
- Department of Kinesiology and Health SciencesCollege of Humanities and Sciences, Virginia Commonwealth UniversityRichmondVAUSA
| | - Priyanshu Nain
- Department of MedicineMedical College of Georgia at Augusta UniversityAugustaGAUSA
| | - Nelson Hamerschlak
- Oncohematology DepartmentHospital Israelita Albert EinsteinSão PauloSPBrazil
| | - Neal L. Weintraub
- Department of MedicineMedical College of Georgia at Augusta UniversityAugustaGAUSA
- Vascular Biology CenterMedical College of Georgia at Augusta UniversityAugustaGAUSA
| | - Susan Dent
- Duke Cancer Institute, Department of MedicineDuke UniversityDurhamNCUSA
| | - Meng‐Han Tsai
- Cancer Prevention, Control, & Population Health Program, Department of MedicineMedical College of Georgia at Augusta UniversityAugustaGAUSA
- Georgia Prevention Institution, Augusta UniversityAugustaGAUSA
| | - Amitava Banerjee
- Institute of Health Informatics, University College LondonLondonUK
| | - Arjun K Ghosh
- Cardio‐Oncology ServiceHatter Cardiovascular Institute, University College London HospitalLondonUK
| | - Diego Sadler
- Department of Cardiovascular MedicineCleveland Clinic FloridaWestonFLUSA
| | - Steven S. Coughlin
- Department of MedicineMedical College of Georgia at Augusta UniversityAugustaGAUSA
- Department of Population Health SciencesMedical College of Georgia at Augusta UniversityAugustaGAUSA
| | - Ana Barac
- Cardio‐Oncology ProgramInova Schar Cancer Institute, Inova Heart and Vascular InstituteFairfaxVAUSA
| | - John Shanahan
- Cancer InformaticsSeidman Cancer Center at University HospitalsClevelandOHUSA
| | - Alberto J. Montero
- Department of Hematology‐OncologyUniversity Hospitals Seidman Cancer CenterClevelandOHUSA
| | - Avirup Guha
- Department of MedicineMedical College of Georgia at Augusta UniversityAugustaGAUSA
- Cardio‐Oncology Program, Department of Medicine, Cardiology DivisionMedical College of Georgia at Augusta UniversityAugustaGAUSA
| |
Collapse
|
3
|
Tan MC, Stabellini N, Tan JY, Thong JY, Hedrick C, Moore JX, Cullen J, Hines A, Sutton A, Sheppard V, Agarwal N, Guha A. Reducing racial and ethnic disparities in cardiovascular outcomes among cancer survivors. Curr Oncol Rep 2024:10.1007/s11912-024-01578-7. [PMID: 39002054 DOI: 10.1007/s11912-024-01578-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 07/02/2024] [Indexed: 07/15/2024]
Abstract
PURPOSE OF REVIEW Analyze current evidence on racial/ethnic disparities in cardiovascular outcomes among cancer survivors, identifying factors and proposing measures to address health inequities. RECENT FINDINGS Existing literature indicates that the Black population experiences worse cardiovascular outcomes following the diagnosis of both initial primary cancer and second primary cancer, with a notably higher prevalence of cardio-toxic events, particularly among breast cancer survivors. Contributing socioeconomic factors to these disparities include unfavorable social determinants of health, inadequate insurance coverage, and structural racism within the healthcare system. Additionally, proinflammatory epigenetic modification is hypothesized to be a contributing genetic variation factor. Addressing these disparities requires a multiperspective approach, encompassing efforts to address racial disparities and social determinants of health within the healthcare system, refine healthcare policies and access, and integrate historically stigmatized racial groups into clinical research. Racial and ethnic disparities persist in cardiovascular outcomes among cancer survivors, driven by multifactorial causes, predominantly associated with social determinants of health. Addressing these healthcare inequities is imperative, and timely efforts must be implemented to narrow the existing gap effectively.
Collapse
Affiliation(s)
- Min Choon Tan
- Department of Internal Medicine, New York Medical College at Saint Michael's Medical Center, Newark, NJ, USA
- Department of Cardiovascular Medicine, Mayo Clinic, Phoenix, AZ, USA
| | - Nickolas Stabellini
- Department of Cardiovascular Medicine, Medical College of Georgia at Augusta University, Augusta, GA, USA
- Case Western Reserve University School of Medicine, Case Western Reserve University, Cleveland, OH, USA
- Department of Hematology-Oncology, University Hospitals Seidman Cancer Center, Cleveland, OH, USA
- Faculdade Israelita de Ciências da Saúde Albert Einstein, Hospital Israelita Albert Einstein, São Paulo, SP, Brazil
| | - Jia Yi Tan
- Department of Internal Medicine, New York Medical College at Saint Michael's Medical Center, Newark, NJ, USA
| | - Jia Yean Thong
- Fudan University Shanghai Medical College, Yangpu District, Shanghai, China
| | - Catherine Hedrick
- Department of Cardiovascular Medicine, Medical College of Georgia at Augusta University, Augusta, GA, USA
| | | | | | - Anika Hines
- Virginia Commonwealth University, Richmond, VA, USA
| | | | | | | | - Avirup Guha
- Department of Cardiovascular Medicine, Medical College of Georgia at Augusta University, Augusta, GA, USA.
- Department of Hematology-Oncology, University Hospitals Seidman Cancer Center, Cleveland, OH, USA.
| |
Collapse
|
4
|
Takkavatakarn K, Dai Y, Hsun Wen H, Kauffman J, Charney A, Coca SG, Nadkarni GN, Chan L. Comparison of predicting cardiovascular disease hospitalization using individual, ZIP code-derived, and machine learning model-predicted educational attainment in New York City. PLoS One 2024; 19:e0297919. [PMID: 38329973 PMCID: PMC10852236 DOI: 10.1371/journal.pone.0297919] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/20/2023] [Accepted: 01/15/2024] [Indexed: 02/10/2024] Open
Abstract
BACKGROUND Area-level social determinants of health (SDOH) based on patients' ZIP codes or census tracts have been commonly used in research instead of individual SDOHs. To our knowledge, whether machine learning (ML) could be used to derive individual SDOH measures, specifically individual educational attainment, is unknown. METHODS This is a retrospective study using data from the Mount Sinai BioMe Biobank. We included participants that completed a validated questionnaire on educational attainment and had home addresses in New York City. ZIP code-level education was derived from the American Community Survey matched for the participant's gender and race/ethnicity. We tested several algorithms to predict individual educational attainment from routinely collected clinical and demographic data. To evaluate how using different measures of educational attainment will impact model performance, we developed three distinct models for predicting cardiovascular (CVD) hospitalization. Educational attainment was imputed into models as either survey-derived, ZIP code-derived, or ML-predicted educational attainment. RESULTS A total of 20,805 participants met inclusion criteria. Concordance between survey and ZIP code-derived education was 47%, while the concordance between survey and ML model-predicted education was 67%. A total of 13,715 patients from the cohort were included into our CVD hospitalization prediction models, of which 1,538 (11.2%) had a history of CVD hospitalization. The AUROC of the model predicting CVD hospitalization using survey-derived education was significantly higher than the model using ZIP code-level education (0.77 versus 0.72; p < 0.001) and the model using ML model-predicted education (0.77 versus 0.75; p < 0.001). The AUROC for the model using ML model-predicted education was also significantly higher than that using ZIP code-level education (p = 0.003). CONCLUSION The concordance of survey and ZIP code-level educational attainment in NYC was low. As expected, the model utilizing survey-derived education achieved the highest performance. The model incorporating our ML model-predicted education outperformed the model relying on ZIP code-derived education. Implementing ML techniques can improve the accuracy of SDOH data and consequently increase the predictive performance of outcome models.
Collapse
Affiliation(s)
- Kullaya Takkavatakarn
- Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
- Division of Nephrology, Department of Medicine, King Chulalongkorn Memorial Hospital, Chulalongkorn University, Bangkok, Thailand
| | - Yang Dai
- The Charles Bronfman Institute for Personalized Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| | - Huei Hsun Wen
- Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| | - Justin Kauffman
- The Charles Bronfman Institute for Personalized Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| | - Alexander Charney
- The Charles Bronfman Institute for Personalized Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| | - Steven G. Coca
- Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| | - Girish N. Nadkarni
- Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
- The Charles Bronfman Institute for Personalized Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
- Division of Data Driven and Digital Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| | - Lili Chan
- Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
- The Charles Bronfman Institute for Personalized Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
- Division of Data Driven and Digital Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America
| |
Collapse
|