Noninvasive Estimation of Mean Pulmonary Artery Pressure Using MRI, Computer Models, and Machine Learning

Abstract

Pulmonary Hypertension (PH) is a severe disease characterized by an elevated pulmonary artery pressure. The gold standard for PH diagnosis is measurement of mean Pulmonary Artery Pressure (mPAP) during an invasive Right Heart Catheterization. In this paper, we investigate noninvasive approach to PH detection utilizing Magnetic Resonance Imaging, Computer Models and Machine Learning. We show using the ablation study, that physics-informed feature engineering based on models of blood circulation increases the performance of Gradient Boosting Decision Trees-based algorithms for classification of PH and regression of values of mPAP. We compare results of regression (with thresholding of estimated mPAP) and classification and demonstrate that metrics achieved in both experiments are comparable. The predicted mPAP values are more informative to the physicians than the probability of PH returned by classification models. They provide the intuitive explanation of the outcome of the machine learning model (clinicians are accustomed to the mPAP metric, contrary to the PH probability).

7Appendix

Acronyms used in Table 1 and their explanations: mPAP: mean pulmonary arterial pressure measured during RHC procedure, who: WHO functional PAH score [3], bsa: body surface area, $R_{d}$: distal resistance calculated from 0D model, $R_{c}$: proximal resistance, C: total pulmonary compliance, $R_{tot}$: total resistance, $W_{b}/W_{tot}$: backward pressure wave to the total wave power, rac_fiesta: pulmonary arterial relative area change from bSSFP MRI, systolic_area_fiesta: syst area of MPA from bSSFP, diast_area_fiesta: diastolic area of MPA from bSSFP, rvedv: right ventricle end diastolic volume, rvedv_index: rv end diastolic volume index, rvesv: rv end systolic volume, rvesv_index: rv end systolic volume index, rvef: right ventricle ejection fraction, rvsv: rv stroke volume, rvsv_index: rvsv index, lvedv: left ventricle end diastolic volume, lvedv_index: lvedv index, lvesv: lv end systolic volume, lvesv_index: lvesv index, lvef: lv ejection fraction, lvsv: lv stroke volume, lvsv_index: lvsv index, rv_dia_mass: rv diastolic mass, lv_dia_mass: lv diastolic mass, lv_syst_mass: lv systolic mass, rv_mass_index: rv diastolic mass index, lv_mass_index: lv diastolic mass index, sept_angle_syst: systolic septal angle, sept_angle_diast: diastolic septal angle, 4ch_la_area: left atrium area 4 chamber, 4ch_la_length: la length 4 chamber, 2ch_la_area: left atrium area 2 chamber, 2ch_la_length: la length 2 chamber, la_volume: la volume, la_volume_index: la volume index, ao_qflowpos: aortic positive flow, ao_qfp_ind: aortic positive flow index, pa_qflowpos: PA positive flow, pa_qflowneg: PA negative flow, pa_qfn_ind: PA negative flow index, systolic_area_pc: systolic MPA area from PC, diastolic_area_pc: diastolic MPA area from PC, rac_pc: relative area change of MPA from PC.