TY - JOUR
T1 - A machine learning model to estimate myocardial stiffness from EDPVR
AU - Babaei, Hamed
AU - Mendiola, Emilio A.
AU - Neelakantan, Sunder
AU - Xiang, Qian
AU - Vang, Alexander
AU - Dixon, Richard A.F.
AU - Shah, Dipan J.
AU - Vanderslice, Peter
AU - Choudhary, Gaurav
AU - Avazmohammadi, Reza
N1 - Funding Information:
R.A. was supported by the NHLBI grant R00HL138288, and G.C. was supported by the NHLBI grants R01HL128661 and R01HL148727 and the VA CSRD grant I01CX001892. We would like to thank Ms. Dana Leichter, Mr. Owen leary and Dr. Richard Gilbert at the providence VA medical center for assisting us with the MRI of PH rats, and Mr. Vishal Kandala at Texas A&M University for assisting us with Supplementary Fig. 2. We would also like to thank Mr. Samer Merchant and Dr. Edward Hsu at the University of Utah for assisting with the MRI of the MI rats. The authors acknowledge the assistance of the Integrated Microscopy and Imaging Laboratory at the Texas A&M College of Medicine.
Funding Information:
R.A. was supported by the NHLBI grant R00HL138288, and G.C. was supported by the NHLBI grants R01HL128661 and R01HL148727 and the VA CSRD grant I01CX001892. We would like to thank Ms. Dana Leichter, Mr. Owen leary and Dr. Richard Gilbert at the providence VA medical center for assisting us with the MRI of PH rats, and Mr. Vishal Kandala at Texas A&M University for assisting us with Supplementary Fig. . We would also like to thank Mr. Samer Merchant and Dr. Edward Hsu at the University of Utah for assisting with the MRI of the MI rats. The authors acknowledge the assistance of the Integrated Microscopy and Imaging Laboratory at the Texas A&M College of Medicine.
Publisher Copyright:
© 2022, The Author(s).
PY - 2022/3/31
Y1 - 2022/3/31
N2 - In-vivo estimation of mechanical properties of the myocardium is essential for patient-specific diagnosis and prognosis of cardiac disease involving myocardial remodeling, including myocardial infarction and heart failure with preserved ejection fraction. Current approaches use time-consuming finite-element (FE) inverse methods that involve reconstructing and meshing the heart geometry, imposing measured loading, and conducting computationally expensive iterative FE simulations. In this paper, we propose a machine learning (ML) model that feasibly and accurately predicts passive myocardial properties directly from select geometric, architectural, and hemodynamic measures, thus bypassing exhaustive steps commonly required in cardiac FE inverse problems. Geometric and fiber-orientation features were chosen to be readily obtainable from standard cardiac imaging protocols. The end-diastolic pressure-volume relationship (EDPVR), which can be obtained using a single-point pressure-volume measurement, was used as a hemodynamic (loading) feature. A comprehensive ML training dataset in the geometry-architecture-loading space was generated, including a wide variety of partially synthesized rodent heart geometry and myofiber helicity possibilities, and a broad range of EDPVRs obtained using forward FE simulations. Latin hypercube sampling was used to create 2500 examples for training, validation, and testing. A multi-layer feed-forward neural network (MFNN) was used as a deep learning agent to train the ML model. The model showed excellent performance in predicting stiffness parameters af and bf associated with fiber direction (Raf2=99.471% and Rbf2=92.837%). After conducting permutation feature importance analysis, the ML performance further improved for bf (Rbf2=96.240%), and the left ventricular volume and endocardial area were found to be the most critical geometric features for accurate predictions. The ML model predictions were evaluated further in two cases: (i) rat-specific stiffness data measured using ex-vivo mechanical testing, and (ii) patient-specific estimation using FE inverse modeling. Excellent agreements with ML predictions were found for both cases. The trained ML model offers a feasible technology to estimate patient-specific myocardial properties, thus, bridging the gap between EDPVR, as a confounded organ-level metric for tissue stiffness, and intrinsic tissue-level properties. These properties provide incremental information relative to traditional organ-level indices for cardiac function, improving the clinical assessment and prognosis of cardiac diseases.
AB - In-vivo estimation of mechanical properties of the myocardium is essential for patient-specific diagnosis and prognosis of cardiac disease involving myocardial remodeling, including myocardial infarction and heart failure with preserved ejection fraction. Current approaches use time-consuming finite-element (FE) inverse methods that involve reconstructing and meshing the heart geometry, imposing measured loading, and conducting computationally expensive iterative FE simulations. In this paper, we propose a machine learning (ML) model that feasibly and accurately predicts passive myocardial properties directly from select geometric, architectural, and hemodynamic measures, thus bypassing exhaustive steps commonly required in cardiac FE inverse problems. Geometric and fiber-orientation features were chosen to be readily obtainable from standard cardiac imaging protocols. The end-diastolic pressure-volume relationship (EDPVR), which can be obtained using a single-point pressure-volume measurement, was used as a hemodynamic (loading) feature. A comprehensive ML training dataset in the geometry-architecture-loading space was generated, including a wide variety of partially synthesized rodent heart geometry and myofiber helicity possibilities, and a broad range of EDPVRs obtained using forward FE simulations. Latin hypercube sampling was used to create 2500 examples for training, validation, and testing. A multi-layer feed-forward neural network (MFNN) was used as a deep learning agent to train the ML model. The model showed excellent performance in predicting stiffness parameters af and bf associated with fiber direction (Raf2=99.471% and Rbf2=92.837%). After conducting permutation feature importance analysis, the ML performance further improved for bf (Rbf2=96.240%), and the left ventricular volume and endocardial area were found to be the most critical geometric features for accurate predictions. The ML model predictions were evaluated further in two cases: (i) rat-specific stiffness data measured using ex-vivo mechanical testing, and (ii) patient-specific estimation using FE inverse modeling. Excellent agreements with ML predictions were found for both cases. The trained ML model offers a feasible technology to estimate patient-specific myocardial properties, thus, bridging the gap between EDPVR, as a confounded organ-level metric for tissue stiffness, and intrinsic tissue-level properties. These properties provide incremental information relative to traditional organ-level indices for cardiac function, improving the clinical assessment and prognosis of cardiac diseases.
UR - http://www.scopus.com/inward/record.url?scp=85127395884&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85127395884&partnerID=8YFLogxK
U2 - 10.1038/s41598-022-09128-6
DO - 10.1038/s41598-022-09128-6
M3 - Article
C2 - 35361836
AN - SCOPUS:85127395884
VL - 12
JO - Scientific Reports
JF - Scientific Reports
SN - 2045-2322
IS - 1
M1 - 5433
ER -