Advancing Heart Failure Classification Beyond Left Ventricular Ejection Fraction: Unsupervised Machine Learning Insights From the Japanese Nationwide Acute Decompensated Heart Failure Cohort: JROADHF
5 affiliations
- 1Nara Medical University
- 2Massachusetts Institute of Technology
- 3Broad Institute of MIT and Harvard
- 4Kyushu University
- 5International University of Health and Welfare
Abstract
BACKGROUND: Heart failure (HF) classification based on left ventricular ejection fraction (LVEF) inadequately captures HF heterogeneity, limiting phenotype-specific therapies. While machine learning (ML) has been applied to HF with preserved ejection fraction, comprehensive approaches across the full LVEF spectrum remain scarce. This study aimed to identify HF phenogroups across the entire LVEF spectrum using unsupervised ML, moving beyond the LVEF-centric and fragmented understanding of HF. METHODS: We analyzed 13 238 patients with acute HF from a nationwide, multicenter Japanese cohort (JROADHF [Japanese Registry of Acute Decompensated Heart Failure]). Unsupervised latent class analysis identified phenogroups from 46 clinically important admission variables. The primary end point was the 5-year composite of cardiovascular death or HF readmission. RESULTS: =0.267). Compared with phenogroup 3, phenogroup 1 had a higher age-adjusted risk by Cox regression model (hazard ratio [HR], 1.60 [95% CI, 1.49-1.71]) and phenogroup 2 had a lower risk (HR, 0.65 [95% CI, 0.60-0.70]). Notably, variable-importance analysis by discriminative power for phenogroup assignment indicated that total bilirubin, left ventricular diastolic diameter, and hemoglobin rather than LVEF principally drove classification. CONCLUSIONS: Unsupervised ML identified novel HF phenogroups with distinct 5-year outcomes. This framework moves beyond an LVEF-centric paradigm, deepens understanding of HF pathophysiology, and may facilitate phenotype-specific therapeutic development.
Related
- Perioperative Management of Decompensated Peripartum Cardiomyopathy With Severe Left Ventricular Dysfunction: A Case Report.
- Advancing Lumbar Disc Herniation Classification from MRI Using an Efficient and Reproducible Classical Machine Learning Framework
- Curriculum-Based Deep Learning for Automated Segmental Left Ventricular Hypertrophy Classification Using Short-Axis Echocardiography
- Impact of Right Ventricular Involvement on Bleeding and Mortality in Patients With Right Coronary Artery-Related Myocardial Infarction ― Insights From Nationwide Registry ―
- Nationwide Evolution of Durable Left Ventricular Assist Device Outcomes Following Introduction of the HeartMate 3 and Destination Therapy in Japan ― The Third Official Report From J-MACS ―
BibTeX
@article{kyodo2026advancing,
title = {{Advancing Heart Failure Classification Beyond Left Ventricular Ejection Fraction: Unsupervised Machine Learning Insights From the Japanese Nationwide Acute Decompensated Heart Failure Cohort: JROADHF}},
author = {Atsushi Kyodo and Yasuki Nakada and Kazutaka Nogi and Satomi Ishihara and Tomoya Ueda and Takeshi Tohyama and Nobuyuki Enzan and Shouji Matsushima and Hiroyuki Tsutsui and Shungo Hikoso},
journal = {Journal of the American Heart Association},
year = {2026},
month = oct,
doi = {10.1161/jaha.126.050615},
url = {https://www.ahajournals.org/doi/10.1161/JAHA.126.050615},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Advancing Heart Failure Classification Beyond Left Ventricular Ejection Fraction: Unsupervised Machine Learning Insights From the Japanese Nationwide Acute Decompensated Heart Failure Cohort: JROADHF
AU - Atsushi Kyodo
AU - Yasuki Nakada
AU - Kazutaka Nogi
AU - Satomi Ishihara
AU - Tomoya Ueda
AU - Takeshi Tohyama
AU - Nobuyuki Enzan
AU - Shouji Matsushima
AU - Hiroyuki Tsutsui
AU - Shungo Hikoso
PY - 2026
DA - 2026/10/20
JO - Journal of the American Heart Association
DO - 10.1161/jaha.126.050615
UR - https://www.ahajournals.org/doi/10.1161/JAHA.126.050615
AB - BACKGROUND: Heart failure (HF) classification based on left ventricular ejection fraction (LVEF) inadequately captures HF heterogeneity, limiting phenotype-specific therapies. While machine learning (ML) has been applied to HF with preserved ejection fraction, comprehensive approaches across the full LVEF spectrum remain scarce. This study aimed to identify HF phenogroups across the entire LVEF spectrum using unsupervised ML, moving beyond the LVEF-centric and fragmented understanding of HF. METHODS: We analyzed 13 238 patients with acute HF from a nationwide, multicenter Japanese cohort (JROADHF [Japanese Registry of Acute Decompensated Heart Failure]). Unsupervised latent class analysis identified phenogroups from 46 clinically important admission variables. The primary end point was the 5-year composite of cardiovascular death or HF readmission. RESULTS: =0.267). Compared with phenogroup 3, phenogroup 1 had a higher age-adjusted risk by Cox regression model (hazard ratio [HR], 1.60 [95% CI, 1.49-1.71]) and phenogroup 2 had a lower risk (HR, 0.65 [95% CI, 0.60-0.70]). Notably, variable-importance analysis by discriminative power for phenogroup assignment indicated that total bilirubin, left ventricular diastolic diameter, and hemoglobin rather than LVEF principally drove classification. CONCLUSIONS: Unsupervised ML identified novel HF phenogroups with distinct 5-year outcomes. This framework moves beyond an LVEF-centric paradigm, deepens understanding of HF pathophysiology, and may facilitate phenotype-specific therapeutic development.
SN - 2047-9980
ER -