A machine learning algorithm for the identification elevated Lp(a) in patients with, or high-risk of having, coronary heart disease.
Autor: | Olmo RF; Cardiology Department, Hospital de Jaén, Jaén, Spain., Cortez G; Cardiology Department, Hospital Alto Guadalquivir, Andujar, Spain., Toro MM; Cardiology Department, Hospital Puerto Real, Cádiz, Spain., Sandín M; Cardiology Department, Hospital General Dr. Balmís, Alicante, Spain., Mora J; Cardiology Department, Hospital Regional Universitario de Málaga, Málaga, Spain., Oterino A; Cardiology Department, Complejo Asistencial Universitario de Salamanca, Salamanca, Spain., Bailen MC; Cardiology Department, Hospital de Jaén, Jaén, Spain., Quiles-Granado J; Cardiology Department, Hospital Puerto Real, Cádiz, Spain., Urbiola P; Cardiology Department, Complejo Asistencial Universitario de Salamanca, Salamanca, Spain., Ruz LF; Cardiology Department, Hospital Regional Universitario de Málaga, Málaga, Spain., Cordero A; Cardiology Department, Hospital IMED Elche, Spain; Grupo de Investigación Cardiovascular (GRINCAVA), Universidad Miguel Hernández, Elche, Spain; Centro de Investigación Biomédica en Red de Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain. Electronic address: acorderofort@gmail.com. |
---|---|
Jazyk: | angličtina |
Zdroj: | International journal of cardiology [Int J Cardiol] 2025 Jan 01; Vol. 418, pp. 132612. Date of Electronic Publication: 2024 Oct 02. |
DOI: | 10.1016/j.ijcard.2024.132612 |
Abstrakt: | Background: Decision tree algorithms, obtained by machine learning, provide clusters of patients with similar clinical patterns by the identification of variables that best merge with a given dependent variable. Methods: We performed a multicenter registry, with 7 hospitals form Spain, of patients with, or high-risk of having, coronary heart disease (CHD). Elevated Lp(a) was defined as >50 mg/dl. Machine learning based decision trees were obtained by Chi-square automatic interaction detection. Results: We analyzed 2301 patients. Median Lp(a) was 26.7 (9.3-79.9) mg/dl and 887 (38.6 %) patients had Lp(a) >50 mg/dl. The machine learning algorithm identified 6 clusters based on LDLc, CHD, FH of premature CHD and age (Fig. 1). Clusters 1 (LDLc <100 mg/dl, no CHD and, no FH of CHD) and 3 (LDLc <100 mg/dl, CHD and, no FH and, age < 50 yo) had the lowest Lp(a) values (Fig. 2); patients classified in cluster 5 (LDLc >100 mg/dl, CHD and, FH of CHD) and 6 (LDLc >100 mg/dl) had the highest values. We collapsed clusters in 3 groups: group 1 with clusters 1 and 3; group 2 with clusters 2 and 4; group 3 with clusters 5 and 6. The 3 groups have significantly different (p < 0.001) and progressively higher Lp(a) values. The prevalence of Lp(a) >50 mg/dl was 15.4 % in group 1, 29.2 % in group 2 and 91.1 % in group 3; similarly, the prevalence of Lp(a) >180 mg/dl was 1.0 %, 3.0 % and 7.6 % respectively. Conclusions: A decision tree algorithm, performed by machine learning, identified patients with, or at high risk of having, CHD have higher probabilities of having elevated Lp(a). Competing Interests: Declaration of competing interest Alberto Cordero reports a) honoraria for lectures from AstraZeneca, AMGEN, Bristol-Myers Squibb, Ferrer, Boehringer Ingelheim, MSD, Daiichy Sankio, Novartis, Novo Nordisk, Sanofi and Amarin; b) consulting fees from AstraZeneca, Ferrer, Sanofi, AMGEN, Novartis, Lilly, Novo Nordisk and Amarin. (Copyright © 2024 Elsevier B.V. All rights reserved.) |
Databáze: | MEDLINE |
Externí odkaz: |