Title
题目
Biomechanical modeling combined with pressure-volume loop analysis to aid surgical planning in patients with complex congenital heart disease
结合生物力学建模和压力-容积环分析，辅助复杂先天性心脏病患者的手术规划
01
文献速递介绍
患有先天性矫正型大动脉转位（ccTGA）的患者具有异常的心房-心室和心室-动脉连接。在这种病症中，形态上的左心室（LV）连接到右心房和肺动脉（PA），而右心室（RV）连接到左心房和主动脉（AO）。由于体循环中的高压力，形态上的右心室需要超出其生理极限工作，这在成年后会导致高发的系统性右心室衰竭（Cui等，2021）。为避免这一并发症，这些患者可以通过心房和动脉转换手术——即双转换手术（DSO）进行治疗，以恢复形态上的左心室和右心室分别执行系统性和亚肺循环功能（Marathe等，2022）。
由于左心室长期作为低压的亚肺心室工作，其功能逐渐退化。这会导致DSO后左心室因无法承受系统性后负荷而发生衰竭的风险。为应对此问题，患者通常接受手术肺动脉束带术（PAB），通过增加后负荷对亚肺左心室进行再训练。PAB会引发左心室的适应性重塑，通常保留6至18个月（Mainwaring等，2018）。在此期间结束后，对左心室进行临床评估以判断其是否具备进行DSO的准备。标准的临床生物标志物通过心导管检查和心脏磁共振成像（MRI）或三维超声心动图获得，通常包括左心室压力、容积、心肌质量和质量-容积比的群体平均“临界值”（Marathe等，2022）。然而，仅依赖于分离的压力（心导管检查）和体积（超声心动图和心脏MRI）标志物对心室功能的理解是有限的。需要将两个心室作为一个整体系统来评估，以维持心输出量的能力。特别是，左心室的特性应与右心室及系统性循环的特性进行比较。然而，标准的临床指标显示出有限的预测能力，即使使用这些指标来评估DSO前左心室的准备情况，DSO的长期结果仍显示~15%的左心室衰竭率（Cui等，2021；Marathe等，2022）。压力-容积（PV）环分析可以评估两个心室的机械性能和能量学特性（Bastos等，2020；Hiremath等，2023）。通过使用电导导管可以获得同步的压力-容积数据，并进一步导出心功能的PV环指标。然而，这种方法存在一些局限性。首先，电导导管的基线校准通常与在不同时间点获取的影像数据对齐，这可能导致舒张末期或收缩末期容积的错配。其次，通过房室（AV）瓣放置导管可能会引发非生理性水平的房室瓣返流。最后，部分儿童的体型或心室大小较小，可能无法放置这些导管。一个替代方法是结合非同步获取的心脏MRI和心导管检查数据来重建双心室的压力-容积环。
Aastract
摘要
Patients with congenitally corrected transposition of the great arteries (ccTGA) can be treated with a double switch operation (DSO) to restore the normal anatomical connection of the left ventricle (LV) to the systemic circulation and the right ventricle (RV) to the pulmonary circulation. The subpulmonary LV progressively deconditions over time due to its connection to the low pressure pulmonary circulation and needs to be retrained using a surgical pulmonary artery band (PAB) for 6–12 months prior to the DSO. The subsequent clinical followup, consisting of invasive cardiac pressure and non-invasive imaging data, evaluates LV preparedness for the DSO. Evaluation using standard clinical techniques has led to unacceptable LV failure rates of ~15 % after DSO. We propose a computational modeling framework to (1) reconstruct LV and RV pressure-volume (PV) loops from non-simultaneously acquired imaging and pressure data and gather model-derived mechanical indicators of ventricular function; and (2) perform in silico DSO to predict the functional response of the LV when connected to the high-pressure systemic circulation.Clinical datasets of six patients with ccTGA after PAB, consisting of cardiac magnetic resonance imaging (MRI) and right and left heart catheterization, were used to build patient-specific models of LV and RV – M LV baseline and M RV baseline. For in silico DSO the models of M LV baseline and M RV baseline were used while imposing the afterload of systemic and pulmonary circulations, respectively. Model-derived contractility and Pressure-Volume Area (PVA) – i.e., the sum of stroke work and potential energy – were computed for both ventricles at baseline and after in silico DSO.In silico* DSO suggests that three patients would require a substantial augmentation of LV contractility between 54 % and 80 % and an increase in PVA between 38 % and 79 % with respect to the baseline values to accommodate the increased afterload of the systemic circulation. On the contrary, the baseline functional state of the remaining three patients is predicted to be adequate to sustain cardiac output after the DSO.This work demonstrates the vast variation of LV function among patients with ccTGA and emphasizes the importance of a biventricular approach to assess patients’ readiness for DSO. Model-derived predictions have the potential to provide additional insights into planning of complex surgical interventions.
患有先天性矫正型大动脉转位（ccTGA）的患者可以通过双转换手术（DSO）恢复左心室（LV）与体循环和右心室（RV）与肺循环的正常解剖连接。然而，由于左心室与低压肺循环相连，其功能会随着时间逐渐减弱，因此需要在进行DSO之前，通过手术肺动脉束带（PAB）进行为期6至12个月的功能再训练。后续的临床随访包括侵入性心脏压力监测和非侵入性影像学数据，用于评估左心室为DSO做好准备的情况。然而，采用标准临床技术进行评估后，DSO后的左心室衰竭发生率仍高达约15%。我们提出了一种计算建模框架，旨在：（1）通过非同时获取的影像和压力数据重建左心室和右心室的压力-容积（PV）环，并利用模型导出的机械指标评估心室功能；（2）进行计算机模拟（in silico）的DSO，以预测左心室连接到高压体循环后的功能响应。利用来自6例接受PAB治疗的ccTGA患者的临床数据集，这些数据包括心脏磁共振成像（MRI）以及左、右心导管检查数据，建立了患者特异性的左心室和右心室模型（M LV baseline和M RV baseline）。在计算机模拟DSO过程中，M LV baseline和M RV baseline模型被用于分别施加体循环和肺循环的后负荷。计算了模型导出的收缩力和压力-容积面积（PVA，即搏功和潜在能量的总和），用于评估两种状态下的心室功能：基线状态和计算机模拟DSO状态。计算机模拟DSO表明，三名患者需要显著增强左心室收缩力，增加幅度在54%至80%之间，同时PVA需要增加38%至79%，以适应体循环的高后负荷。而其余三名患者的基线功能状态被预测为足以在DSO后维持心输出量。该研究显示了ccTGA患者中左心室功能的巨大差异，并强调了采用双心室方法评估患者准备状态的重要性。基于模型的预测可能为复杂手术干预的规划提供额外的见解。
Method
方法
Six datasets of patients with ccTGA were included in the study. Thedata were acquired at Boston Children’s Hospital under IRB-P00044532.Patients underwent cardiac catheterization on average 8 months afterPAB was performed. Left and right heart pressures (LV, RV, PA, and AO)were measured with a fluid-filled catheter. One day before the catheterization, cardiac MRI data were collected: retrospective ECG-gatedcine bSSFP sequence, 1.5T scanner (Achieva, Philips Healthcare, Best,the Netherlands), SENSE=2, spatial resolution 1.5 × 1.5 × 5 mm, 30phases per cardiac cycle (temporal resolution of 17–31 ms per phase at the observed heart rates of 63 to 114 beats per minute). These catheterization and MRI data, acquired after PAB, are further referred to as baseline data. Ventricular end-systolic/-diastolic volumes weresegmented in CVI42 software (Circle Cardiovascular Imaging Inc., Calgary, Canada) and time vs. volume signals were obtained by using amotion tracking algorithm (Genet et al., 2018). Note that motiontracking was used solely to extract the time-variant ventricular volumesfrom the MRI data, and no parameters from the tracking tool were usedfor patient functional assessment.
本研究包含6例ccTGA患者的数据。这些数据在波士顿儿童医院根据IRB批准编号P00044532收集。患者在进行PAB手术后平均8个月接受了心导管检查。通过液体填充导管测量左心室（LV）、右心室（RV）、肺动脉（PA）和主动脉（AO）的压力。在心导管检查前一天，收集了心脏MRI数据：采用回顾性心电图门控的cine bSSFP序列，使用1.5T扫描仪（Achieva，飞利浦医疗，荷兰Best），SENSE因子为2，空间分辨率为1.5 × 1.5 × 5 mm，每心动周期30个相位（在观察到的心率范围63到114次/分钟内，每相位的时间分辨率为17–31 ms）。这些在PAB手术后收集的心导管检查和MRI数据统称为基线数据。
心室的收缩末期和舒张末期容积在CVI42软件（Circle Cardiovascular Imaging Inc., 加拿大卡尔加里）中进行分割，并通过运动追踪算法获得时间-容积信号（Genet等，2018）。需要注意的是，运动追踪仅用于从MRI数据中提取时间变化的心室容积，追踪工具的任何参数均未用于患者功能评估。
Conclusion
结论
This study underscores the vast variation in the function of subpulmonary LV among patients with ccTGA prior to DSO, as assessed from imaging and pressure data using biomechanical models. The study emphasizes the importance of a biventricular assessment for evaluating patients’ readiness for DSO. Model-derived predictions from *in silico* surgeries offer valuable insights that enhance understanding of physiology and has the potential to assist in guiding treatment strategies. Computational modeling provides an optimized framework to exploit cardiac MRI and catheterization data in clinical settings, e.g., allowing the reconstruction of both RV and LV PV loops from non-simultaneously acquired signals. This optimized framework provides a more comprehensive means of evaluating patient physiology, improving the insights into the challenges posed by ccTGA and LV preparedness for DSO. The integration of modeling techniques with clinical data holds promises for refining patient-specific strategies for interventions in a wide variety of congenital or acquired heart conditions.
本研究强调了在双转换手术（DSO）之前，通过生物力学模型分析影像和压力数据，评估ccTGA患者亚肺左心室（LV）功能的显著差异性。本研究突出了双心室评估在判断患者是否为DSO做好准备方面的重要性。基于*计算机模拟*手术的模型预测提供了有价值的见解，增强了对生理学的理解，并有潜力指导治疗策略。计算建模为在临床环境中充分利用心脏MRI和心导管检查数据提供了优化框架，例如允许从非同时获取的信号中重建右心室（RV）和左心室（LV）的压力-容积环（PV环）。这种优化框架提供了一种更全面的方法来评估患者的生理状况，从而更深入地了解ccTGA所面临的挑战以及LV在DSO中的准备情况。将建模技术与临床数据相结合，为优化各种先天性或后天性心脏疾病中的个体化干预策略提供了广阔的前景。
Results
结果
Patients’ clinical data obtained during catheterization are shown in Tables 2 and 3. Note that Patient 1 had a large atrial septal defect (ASD) leading to a significantly higher subpulmonary LV stroke volume at baseline compared to the RV (Fig. 4). Changes in LV and RV contractility, stroke work, potential energy, and PVA are summarized in Tables 6 and 7.Figs. 2 and 3 show an example of LV and RV model calibration for Patient 3 to the baseline data – M LV baseline and M RV baseline – and a model output of in silico DSO for LV and RV – M LV switch and M RV switch, respectively. RV and LV model calibrations for remaining 5 patients can be found in Supplementary Figures S1-S10.
患者在心导管检查期间获得的临床数据见表2和表3。注意，患者1存在较大的房间隔缺损（ASD），导致基线状态下亚肺左心室（LV）的搏出量显著高于右心室（RV）（见图4）。左心室和右心室的收缩力、搏功、势能和压力-容积面积（PVA）的变化汇总于表6和表7中。图2和图3展示了患者3的左心室（LV）和右心室（RV）模型在基线数据（M LV baseline 和 M RV baseline）下的校准实例，以及计算机模拟DSO后左心室和右心室模型的输出（M LV switch 和 M RV switch）。其余5名患者的右心室和左心室模型校准可参见补充图S1-S10。
Figure
图
Fig. 1. (a) Schematics of a biomechanical model of a single ventricular cavity with thickness d, and radius R coupled with the 2-stage Windkessel model of the circulation. Atrioventricular and outflow valves are represented by the system of diodes with forward and backward resistances. (b) Pressure-volume (PV) loop analysis of myocardial energetics presented is a model-derived PV loop at baseline of LV of Patient 2. ESPVR: end-systolic pressure volume relationship, SW: stroke work, PE: potential energy. The Pressure-Volume Area (PVA) is given by summing up PE and SW.
图1. (a) 单心室腔的生物力学模型示意图，其中腔壁厚度为d，半径为R，并与循环系统的双阶段Windkessel模型耦合。房室瓣和流出瓣由具有正向和反向阻力的二极管系统表示。(b) 心肌能量学的压力-容积（PV）环分析，展示的是模型导出的Patient 2左心室基线状态下的PV环。ESPVR：收缩末期压力-容积关系，SW：搏功，PE：势能。压力-容积面积（PVA）通过将PE和SW相加得到。
Fig. 2. Graphical illustration of the experimental setup for each individual gastrointestinal endoscopic application. For each set of experiments, the section numbers are indicated in which the experimental results are presented and discussed.
图2. 每个胃肠道内镜应用的实验设置图示。对于每组实验，标注了实验结果展示和讨论所在的章节编号。
Fig. 3. The schematic of the design of (a) TransCAN and the details of (b) Contour Extractor module and © SB-Matching module in TransCAN.
图3. (a) TransCAN设计示意图，以及(b) 轮廓提取模块和© TransCAN中侧支匹配模块的详细结构。
Fig. 4. Visual description of key challenges in fine-alignment. (a) The straighteningof 3D-XA with cubes representing convolution block; (b) Example slices p and d havesame area but differ in anatomical and topological characteristics. From left to rightrepresent the two slices have spatial distortions, locating on opposite directions ofa SB, and different contour shapes; © The SCT1D transforms 3D points into a 1Drepresentation.
图4. 精细对准中的关键挑战的可视化描述。(a) 三维XA的直化处理，其中方块代表卷积块；(b) 示例切片p和d具有相同的面积，但在解剖学和拓扑特性上存在差异。从左到右分别表示两个切片存在空间扭曲、位于侧支的相反方向以及轮廓形状不同；© SCT1D将三维点转换为一维表示。
Fig. 5. Branch Position Encoder Generator (BPEG) module. 𝐿 and 𝑑 represent thelength of sequence and contour feature 𝜒, respectively. The (⋅) consists of a flatteningoperation followed by a 1D-Depthwise Separable Convolution with a kernel size of ℎ×1.
图5. 分支位置编码生成器（BPEG）模块。𝐿 和 𝑑 分别表示序列长度和轮廓特征 𝜒。(⋅) 包括一个展平操作，随后进行核大小为 ℎ×1 的一维深度可分卷积。
Fig. 6. Prediction of LV contractility by the in silico double switch operation (black dot) and various levels of in silico pulmonary artery band (PAB) tightening (black stars).
图6. 计算机模拟双转换手术（黑点）和不同程度的计算机模拟肺动脉束带（PAB）收紧（黑星）对左心室（LV）收缩力的预测。
Table
表
Table 1 Overview of model calibration procedure. EDP: end-diastolic pressure, EDV: end-diastolic volume, PSP: peak systolic pressure, AV valve: atrioventricular valve, SV: stroke volume.
表1模型校准过程概览。 EDP：舒张末期压力，EDV：舒张末期容积，PSP：收缩峰值压力，AV瓣：房室瓣，SV：搏出量。
Table 2 Patient characteristics and left ventricular (LV) clinical data derived from catheterization and cardiac magnetic resonance imaging (MRI) at baseline. Note that there was no or trivial aortic and pulmonary valve regurgitation. BSA: body surface area, EDP: end-diastolic pressure, PSP: peak systolic pressure, LVFW: left ventricular free wall, PA: pulmonary artery; PA Qfor: forward flow in PA; RF: regurgitation fraction; HR: heart rate. Note that, heart rate was recorded from LV pressure waveform.、
表2患者特征及基线状态下通过心导管检查和心脏磁共振成像（MRI）获取的左心室（LV）临床数据。注意，主动脉和肺动脉瓣无或仅有轻微返流。 BSA：体表面积，EDP：舒张末期压力，PSP：收缩峰值压力，LVFW：左心室游离壁，PA：肺动脉，PA Qfor：肺动脉前向流量，RF：返流分数，HR：心率。注意，心率是从左心室压力波形中记录的。
Table 3 Right ventricular (RV) clinical data derived from catheterization and cardiac magnetic resonance imaging (MRI) at baseline. Note that there was no or trivial aortic and pulmonary valve regurgitation. EDP: end-diastolic pressure, PSP: peak systolic pressure, RVFW: right ventricular free wall, AO: aorta; AO Qfor: forward flow in aorta; RF: regurgitation fraction.
表3基线状态下通过心导管检查和心脏磁共振成像（MRI）获取的右心室（RV）临床数据。注意，主动脉和肺动脉瓣无或轻微返流。 EDP：舒张末期压力，PSP：收缩峰值压力，RVFW：右心室游离壁，AO：主动脉，AO Qfor：主动脉前向流量，RF：返流分数。
Table 4 Patient-specific mechanical parameters of the left ventricular (LV) heart and pulmonary circulation model at baseline. PA: pulmonary artery; MV: mitral valve; σ0: myocardial contractility; K: relative stiffness; RVOT for : forward resistance of the pulmonary valve; RAV back: backward resistance of the mitral valve; Rp: proximal resistance of the pulmonary circulation;Rd: distal resistance of the pulmonary circulation; Cd: distal capacitance of the pulmonary circulation. Note that, RAV back=20,000×107 Pas/m3 is a default value corresponding to no MV regurgitation
表4基线状态下患者特异性的左心室（LV）与肺循环模型的机械参数。PA：肺动脉；MV：二尖瓣；σ₀：心肌收缩力；K：相对刚度；Rᴠᴏᴛ ᶠᵒʳ：肺动脉瓣的正向阻力；Rᴀᴠ ʙᴀᴄᴋ：二尖瓣的反向阻力；Rₚ：肺循环的近端阻力；Rᵈ：肺循环的远端阻力；Cᵈ：肺循环的远端电容。注意，Rᴀᴠ ʙᴀᴄᴋ=20,000×10⁷ Pa·s/m³是默认值，对应于无二尖瓣返流的情况。
Table 5 Patient-specific mechanical parameters of the right ventricular (RV) heart and systemic circulation model at baseline. AO: aorta; TV: tricuspid valve; σ0: myocardial contractility; K: relative stiffness; RVOT for : forward resistance of the aortic valve; RAV back: backward resistance of the tricuspid valve; Rp: proximal resistance of the systemic circulation; Rd: distal resistance of the systemic circulation; Cd: distal capacitance of the systemic circulation. Note that RAV back=20,000 × 107 Pas/m3 is a default value corresponding to no TV regurgitation, and RVOT for =0.0077×107 Pas/m3 is a default value corresponding to no ventricular-arterial pressure gradient.
表5基线状态下患者特异性的右心室（RV）与体循环模型的机械参数。 AO：主动脉；TV：三尖瓣；σ₀：心肌收缩力；K：相对刚度；Rᴠᴏᴛ ᶠᵒʳ：主动脉瓣的正向阻力；Rᴀᴠ ʙᴀᴄᴋ：三尖瓣的反向阻力；Rₚ：体循环的近端阻力；Rᵈ：体循环的远端阻力；Cᵈ：体循环的远端电容。 注意，Rᴀᴠ ʙᴀᴄᴋ=20,000 × 10⁷ Pa·s/m³是默认值，对应于无三尖瓣返流的情况，Rᴠᴏᴛ ᶠᵒʳ=0.0077 × 10⁷ Pa·s/m³是默认值，对应于无心室-动脉压力梯度的情况。
Table 6 Prediction of left ventricular (LV) contractility (kPa), stroke work (J), potential energy (J), and Pressure-Volume Area (J) by the model of in silico double switch operation at various levels of LV end-diastolic pressures (EDP). AO: aorta; PA: pulmonary artery; DSO: double-switch operation. * Averaged values over the LV EDP range of 6–15 mmHg were taken to compute the relative changes with the baseline LV parameters. The Pressure-Volume Area (PVA) is given by summing up PE and SW.
表6模型预测在不同左心室舒张末期压力（EDP）水平下，计算机模拟双转换手术（DSO）后的左心室（LV）收缩力（kPa）、搏功（J）、势能（J）和压力-容积面积（PVA，单位：J）。 AO：主动脉；PA：肺动脉；DSO：双转换手术。平均值为LV EDP在6–15 mmHg范围内的取值，用于计算相对于基线LV参数的相对变化。压力-容积面积（PVA）通过将势能（PE）和搏功（SW）相加得到。
Table 7 Prediction of right ventricular (RV) contractility (kPa), stroke work (J), potential energy (J), and Pressure-Volume Area (J) by the model of in silico double switch operation at various levels of RV end-diastolic pressures (EDP). AO: aorta; PA: pulmonary artery; DSO: double-switch operation. * Averaged values over the RV EDP range of 6–15 mmHg were taken to compute the relative changes with the baseline RV parameters.
表7模型预测在不同右心室舒张末期压力（EDP）水平下，计算机模拟双转换手术（DSO）后的右心室（RV）收缩力（kPa）、搏功（J）、势能（J）和压力-容积面积（PVA，单位：J）。 AO：主动脉；PA：肺动脉；DSO：双转换手术。平均值为RV EDP在6–15 mmHg范围内的取值，用于计算相对于基线RV参数的相对变化。
Table 8 LV and RV Pressure-Volume Area (PVA) comparison. The in silico DSO LV PVA displayed is the average of values obtained over a range of end-diastolic pressures from 6 to 15 mmHg as shown in Table 2. LV, left ventricle; RV, right ventricle; DSO, double switch operation.
表8左心室（LV）和右心室（RV）压力-容积面积（PVA）比较。计算机模拟双转换手术（DSO）中的LV PVA为在舒张末期压力范围6至15 mmHg内获取的值的平均值（见表2）。 LV：左心室；RV：右心室；DSO：双转换手术。
Table 9 Left and right ventricular (LV, RV) cardiac output (CO) in baseline and in silico double switch operation (DSO). Averaged values over the LV/RV EDP range of 6–15 mmHg were taken to compute the relative change between baseline and in silico DSO models, where negative and positive errors represent a decrease and increase of in silico DSO value, respectively. Note that, heart rate was assumed to be the same in baseline and for in silico DSO.
表9左心室（LV）和右心室（RV）在基线状态和计算机模拟双转换手术（DSO）中的心输出量（CO）。LV/RV舒张末期压力（EDP）范围为6至15 mmHg的平均值用于计算基线状态与计算机模拟DSO模型之间的相对变化，其中负误差和正误差分别表示计算机模拟DSO值的减少和增加。注意，心率在基线状态和计算机模拟DSO中假设相同。
Table 10 Left and right ventricular (LV, RV) arterial forward flow in baseline and in silico double switch operation (DSO). Averaged values over the LV/RV EDP range of 6–15 mmHg were taken to compute the relative change between baseline and in silico* DSO models, where negative and positive errors represent a decrease and increase for in silico DSO value, respectively.、
表10左心室（LV）和右心室（RV）在基线状态和计算机模拟双转换手术（DSO）中的动脉前向流量。LV/RV舒张末期压力（EDP）范围为6至15 mmHg的平均值用于计算基线状态与计算机模拟DSO模型之间的相对变化，其中负误差和正误差分别表示计算机模拟DSO值的减少和增加。
