EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Interpretable Machine Learning Using Accumulated Local Effects to Characterize Predictors of Subclinical Leaflet Thrombosis After Self-Expanding Transcatheter Aortic Valve Implantation

Marco Moscarelli iD, Thanos Athanasiou iD, Roberto Casula, Vincenzo Pernice, Massimo Salardino, Gregorio Zaccone, Adriana Zlahoda-Huzior iD, Dariusz Dudek, Sabrina Milo, Giuseppe Speziale, Khalil Fattouch iD

Interdisciplinary CardioVascular and Thoracic Surgery · 2026

DOI: 10.1093/icvts/ivag144 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Interdisciplinary CardioVascular and Thoracic Surgery, 2026. Verfasst von Marco Moscarelli und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Cardiovascular Surgery, Maria Eleonora Hospital, GVM Care & Research, Palermo 90100, Italy. Hauptthemen laut MEDLINE: Transcatheter Aortic Valve Replacement, Thrombosis, Heart Valve Prosthesis, Aortic Valve. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European grant.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

Subclinical leaflet thrombosis (SLT) is an early form of bioprosthetic valve dysfunction after transcatheter aortic valve implantation. Predicting SLT remains challenging. We aimed to apply machine learning not only to identify predictors but also to interpret their dynamic, non-linear effects on SLT risk using accumulated local effects (ALEs) curves, a robust method that accounts for variable collinearity.

Methods

A prospective cohort of 128 consecutive patients receiving a self-expanding transcatheter heart valve underwent multimodality imaging and haematological profiling. The primary outcome was SLT on 6-month CT. An extreme gradient boosting classifier was trained on 126 variables. Model performance was evaluated via nested cross-validation. Interpretability used SHapley Additive exPlanations and ALEs plots.

Results

Subclinical thrombosis was detected in 22 patients (17.1%). Extreme gradient boosting model demonstrated excellent discrimination (AUC: 0.91, 95% CI, 0.87-0.94) and calibration (Brier score: 0.09). SHapley Additive exPlanations analysis identified the top predictors: bicuspid valve anatomy (mean: 0.45), baseline haemoglobin (0.28), periprocedural Δhaematocrit (0.19), prosthesis eccentricity (0.14), and post-procedural platelet nadir (0.09). ALEs curves revealed a U-shaped association for baseline haemoglobin (lowest risk around 13 g/dL) a monotonic rise in SLT risk with increasing prosthesis eccentricity and greater periprocedural declines in haematocrit, and a progressively higher risk with post-procedural thrombocytopaenia.

Conclusions

This study introduces ALEs plots to cardiovascular medicine, translating predictions into interpretable risk curves that show how leaflet thrombosis risk evolves across key predictors, linking valve geometry and periprocedural haematological changes to guide patient surveillance after transcatheter aortic valve replacement.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Transcatheter Aortic Valve ReplacementThrombosisHeart Valve ProsthesisAortic ValveMachine LearningAortic Valve StenosisHumansProspective StudiesPredictive Learning ModelsFemaleMaleRisk FactorsAged, 80 and overAgedTreatment OutcomeRisk AssessmentProsthesis DesignPredictive Value of TestsBoosting Machine Learning AlgorithmsTime Factors

Stichwörter der Autor:innen

SHAPaccumulated local effectsextreme gradient boostingmachine learningsubclinical leaflet thrombosistranscatheter aortic valve implantation

Autor:innen und Institutionen

  • Department of Cardiovascular Surgery, Maria Eleonora Hospital, GVM Care & Research, Palermo 90100, Italy.
  • Department of Surgery & Cancer, Faculty of Medicine, Imperial College London, London 0HS0HS w12 0HS, United Kingdom.
  • Department of Cardiothoracic Surgery, Hammersmith Hospital, Imperial College Healthcare NHS Trust, London w12 0Hs, United Kingdom.
  • Simhub, Virmed, Krakow 33332, Poland.
  • Center for Digital Medicine and Robotics, Jagiellonian University Medical College, Krakow 31034, Poland.
  • Department of Radiology, Maria Eleonora Hospital, GVM Care & Research, Palermo 90100, Italy.

ORCID-Profile: Marco Moscarelli, Thanos Athanasiou, Adriana Zlahoda-Huzior, Khalil Fattouch

Förderung

  • European grant

Zitieren

APA

Moscarelli, M., Athanasiou, T., Casula, R., Pernice, V., Salardino, M., Zaccone, G., Zlahoda-Huzior, A., Dudek, D., Milo, S., Speziale, G., & Fattouch, K. (2026). Interpretable Machine Learning Using Accumulated Local Effects to Characterize Predictors of Subclinical Leaflet Thrombosis After Self-Expanding Transcatheter Aortic Valve Implantation. Interdisciplinary CardioVascular and Thoracic Surgery. https://doi.org/10.1093/icvts/ivag144

Vancouver

Moscarelli M, Athanasiou T, Casula R, Pernice V, Salardino M, Zaccone G, et al. Interpretable Machine Learning Using Accumulated Local Effects to Characterize Predictors of Subclinical Leaflet Thrombosis After Self-Expanding Transcatheter Aortic Valve Implantation. Interdisciplinary CardioVascular and Thoracic Surgery. 2026. doi:10.1093/icvts/ivag144

BibTeX
@article{Moscarelli2026,
  author = {Moscarelli, Marco and Athanasiou, Thanos and Casula, Roberto and Pernice, Vincenzo and Salardino, Massimo and Zaccone, Gregorio and Zlahoda-Huzior, Adriana and Dudek, Dariusz and Milo, Sabrina and Speziale, Giuseppe and Fattouch, Khalil},
  title = {Interpretable Machine Learning Using Accumulated Local Effects to Characterize Predictors of Subclinical Leaflet Thrombosis After Self-Expanding Transcatheter Aortic Valve Implantation},
  journal = {Interdisciplinary CardioVascular and Thoracic Surgery},
  year = {2026},
  doi = {10.1093/icvts/ivag144}
}
Kostenlos starten →