What is the clinical evidence from this study?

Study design: Observational. Population: Heart failure (n=1204). Intervention: Pericardial adipose tissue (PAT) radiomics phenotyping vs. Non-heart failure controls and PAT area alone. Primary outcome: Discrimination of prevalent heart failure and prediction of incident heart failure (AUC 0.76 (prevalent HF); AUC 0.74 (incident HF)).

What does this research mean for the field?

Pericardial adipose tissue radiomics phenotyping discriminates prevalent heart failure and predicts incident heart failure, performing marginally better than pericardial fat area alone. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.ESTABLISHES_NEW_DIRECTION.

November 21, 2023European RadiologyOpen Access

Radiomics of pericardial fat: a new frontier in heart failure discrimination and prediction

View Full Paper

Ask AI

Bookmark

Key result

Pericardial fat radiomics predicts heart failure with 0.76 AUC, marginally outperforming fat area alone.

AUC 0.76 (prevalent HF); AUC 0.74 (incident HF)
n=1,204

Why the study?

The potential of pericardial adipose tissue radiomics phenotyping to differentiate and predict heart failure cases was unexplored in a large population cohort.

Does pericardial adipose tissue radiomics phenotyping improve the discrimination of prevalent and prediction of incident heart failure compared to PAT area alone in UK Biobank participants?

Population

1204 UK Biobank participants including 297 prevalent and 305 incident heart failure cases

Comparison

Radiomics models using 28 features plus sex and age versus PAT area alone

Design

Observational cohort study using cardiovascular magnetic resonance and machine learning classifiers

Follow-up

Not explicitly stated

Authors

LSLiliána SzabóSemmelweis University ASAhmed SalihSt Bartholomew's Hospital EPEsmeralda Ruiz PujadasBarcelona Graduate School of Mathematics

Discussion

Member takes

Implication

May enhance HF discrimination on CMR; leaves open clinical utility without prospective validation.

Study Design

Type

Observational (n=1,204)

Multicenter

Yes

Structured PICO

Does pericardial adipose tissue radiomics phenotyping improve the discrimination of prevalent and prediction of incident heart failure compared to PAT area alone in UK Biobank participants?

Population

1,204 participants from the UK Biobank Imaging Study, including 297 with prevalent heart failure, 305 with incident heart failure, and 602 non-heart failure controls, followed for an average of 3.7 years.

Exposure

Pericardial adipose tissue (PAT) radiomics phenotyping derived from cardiovascular magnetic resonance (CMR) using machine learning classification models (incorporating 28 radiomics features, sex, and age).

Comparator

Predictive models using PAT area alone, sex, and age.

Outcome

Discrimination of prevalent heart failure and prediction of incident heart failure.surrogate

Main Result

Effect estimate: AUC 0.76 (prevalent HF); AUC 0.74 (incident HF)

Pericardial adipose tissue radiomics phenotyping from CMR scans can effectively discriminate prevalent heart failure and predict incident heart failure, performing marginally better than simple PAT area measurements.

Limitations

The models are preliminary and need further independent external validation.
Random undersampling was used due to extreme dataset imbalance, potentially leading to bias and overoptimistic performance estimates.
Other relevant clinical factors were not included in the models, warranting comprehensive patient information in future research.

Cite This Study

Szabó et al. (2023) conducted an observational in Heart failure (n=1,204). Pericardial adipose tissue (PAT) radiomics phenotyping vs. Non-heart failure controls and PAT area alone was evaluated on Discrimination of prevalent heart failure and prediction of incident heart failure (AUC 0.76 (prevalent HF); AUC 0.74 (incident HF)). Pericardial adipose tissue radiomics phenotyping demonstrated good discriminative performance for detecting prevalent heart failure (AUC 0.76) and predicting incident heart failure (AUC 0.74), performing marginally better than pericardial fat area alone.

synapsesocial.com/papers/6aaf41ae66ccc072b425d1f4 https://doi.org/10.1007/s00330-023-10311-0

Topics

Heart failure HFrEF treatment

View Full Paper

Ask AI

Bookmark

Radiomics of pericardial fat: a new frontier in heart failure discrimination and prediction

Discussion

Member takes

Implication

Study Design

Structured PICO

Main Result

Limitations

Cite This Study

Topics

Radiomics of pericardial fat: a new frontier in heart failure discrimination and prediction

Discussion

Member takes

Implication

Study Design

Structured PICO

Main Result

Limitations

Cite This Study

Also Consider

Topics

Also Consider