Beyond thrombus vs tumor: what computed tomography radiomics of cardiac masses must show next

Q: What is the clinical evidence from this study?

Study design: Editorial. Population: Cardiac masses. Intervention: Computed tomography radiomics vs. Conventional imaging.

Artur Dziewierz; Francesco Pelliccia

doi:10.20452/pamw.17404

September 29, 2026Polskie Archiwum Medycyny WewnętrznejOpen Access

Beyond thrombus vs tumor: what computed tomography radiomics of cardiac masses must show next

View Full Paper

Ask AI

Bookmark

Share

Key result

Cardiac mass CT radiomics requires multiclass phenotyping and external validation to prove clinical value.

Key Points

Assess the diagnostic performance, methodological limitations, and clinical requirements for computed tomography radiomics in distinguishing complex cardiac masses beyond binary tumor-versus-thrombus classification.
Analyzed evidence from studies applying CT radiomics to cardiac masses, focusing on a 3-class model by Trochimiuk et al (N=89; 72 training, 17 validation) alongside cohorts from Lee et al (N=255; 192 training, 63 external) and Qian et al (N=109; 76 training, 33 testing).
Compared feature extraction, reduction methods (such as LASSO regularization), reference standards (including response to anticoagulation in all 37 thrombus cases in Trochimiuk et al), and validation metrics.
In Trochimiuk et al, reduction of 1691 features yielded a 4-feature multinomial model with one-vs-rest validation AUCs ranging from 0.771 to 0.933 and an accuracy of 88.2% (15 of 17), though neither of the 2 lesions in the heterogeneous 'other' validation category was correctly classified.
In independent external validation by Lee et al, radiomics alone was not significantly superior to a conventional CT model (AUC 0.872 vs 0.753; P=0.12), though combining radiomics with clinical and CT variables proved additive (AUC 0.911 vs 0.802; P=0.04).

PICO

P

Population

Cardiac masses

I

Intervention / Comparator

Computed tomography radiomics vs Conventional imaging

Limitations

Radiomic features vary depending on scanner, acquisition parameters, reconstruction kernel, and preprocessing.
Using response to anticoagulation as a reference standard may underrepresent organized thrombi that mimic neoplasia.
Rare tumors and difficult mimics require multicenter registries to accumulate sufficient numbers for robust model training.

Implication

Editorial review evaluates CT radiomics for classifying cardiac masses beyond binary tumor-versus-thrombus models, highlighting clinical challenges and performance metric limitations.

Abstract

No abstract yet. The journal hasn't published one we can show.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Dziewierz et al. (2026) conducted an editorial in Cardiac masses. Computed tomography radiomics vs. Conventional imaging was evaluated. Computed tomography radiomics for cardiac masses must evolve beyond binary classification toward multiclass phenotyping with external validation to demonstrate incremental clinical value.

synapsesocial.com/papers/6abdfac166f3754618517abd https://doi.org/10.20452/pamw.17404

View Full Paper

Ask AI

Bookmark

Share

Beyond thrombus vs tumor: what computed tomography radiomics of cardiac masses must show next

Key Points

PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Beyond thrombus vs tumor: what computed tomography radiomics of cardiac masses must show next

Key Points

PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Keep reading

More from Polskie Archiwum Medycyny Wewnętrznej

Keep reading

More from Polskie Archiwum Medycyny Wewnętrznej