Model & case documentation

HFrEF · chronic LV-dilated case

One resting construction with a larger left ventricle and reduced ejected fraction.

A case for research and education, not for patient diagnosis or treatment.

Case overview

HFrEF identifies heart failure with reduced EF; it does not itself specify chronicity, dilation or etiology. This case selects a chronic LV-dilated phenotype, not acute infarction, regional ischemia or the time course of remodeling.

The intended lesson is greater residual LV blood, retained resting forward output and elevated left atrial pressure. RV behavior and filling need not match a universal HFrEF pattern. Intentions and observed results are separated below.

Circuit and pressure interpretation

Four cavities and eleven main-circuit vascular compartments connect to sixteen coronary compartments. Compartments store blood; links carry hydraulic losses. Names denote lumped regions, not exact catheter positions. The septum is a material wall, not a blood-storage node.

Systemic path

  1. Left ventricle (LV)
  2. Proximal aorta (Ao)
  3. Systemic arteries (SA)
  4. Systemic resistance-vessel side (Art)
  5. Systemic capillary bed (Cap)
  6. Systemic veins (SV)
  7. Vena cava (VC)
  8. Right atrium (RA)

Pulmonary path

  1. Right ventricle (RV)
  2. Proximal pulmonary artery (PA)
  3. Pulmonary resistance-vessel side (PArt)
  4. Pulmonary capillary bed (PCap)
  5. Pulmonary venular side (PVen)
  6. Pulmonary vein / LA inlet (PVein)
  7. Left atrium (LA)
LA —MV→ LV and RA —TV→ RV close the circuit. Arrows define positive flow; permitted reverse flow enters continuity with its sign. Coronaries branch from Ao and return to RA.

Ao/SA/Art separate proximal storage from pressure loss toward the periphery, without representing pulse transit or reflection between compartments. SV is the main venous reservoir; VC is the thoracic-pressure-exposed compartment before RA. PCap receives alveolar pressure, while PVen/PVein receive thoracic pressure, separating pulmonary storage and resistance along the path to LA. This is a functional partition, not a vessel-by-vessel anatomical reconstruction.

All main-circuit compartments: law and pressure reference
CompartmentPressure relationExternal pressureDisplayed counterpart
Left ventricle (LV)Material / geometry balancePth+PperiP_{\mathrm{th}}+P_{\mathrm{peri}}LVP
Left atrium (LA)Material / geometry balancePth+PperiP_{\mathrm{th}}+P_{\mathrm{peri}}LAP
Right ventricle (RV)Material / geometry balancePth+PperiP_{\mathrm{th}}+P_{\mathrm{peri}}RVP
Right atrium (RA)Material / geometry balancePth+PperiP_{\mathrm{th}}+P_{\mathrm{peri}}RAP / CVP
Proximal aorta (Ao)Exponential arterial0 (reference)AoP
Systemic arteries (SA)Exponential arterial0 (reference)ABP
Systemic resistance-vessel side (Art)Exponential arterial0 (reference)
Systemic capillary bed (Cap)Linear storage0 (reference)
Systemic veins (SV)Nonlinear venous-type0 (reference)
Vena cava (VC)Nonlinear venous-typePthP_{\mathrm{th}}
Proximal pulmonary artery (PA)Exponential arterialPthP_{\mathrm{th}}PAP
Pulmonary resistance-vessel side (PArt)Exponential arterialPthP_{\mathrm{th}}
Pulmonary capillary bed (PCap)Nonlinear venous-typePalvP_{\mathrm{alv}}
Pulmonary venular side (PVen)Nonlinear venous-typePthP_{\mathrm{th}}
Pulmonary vein / LA inlet (PVein)Nonlinear venous-typePthP_{\mathrm{th}}

AoP/PAP are Ao/PA intravascular pressures, with no display-only ZcQ. ABP is SA pressure, not a simulated brachial cuff. CVP uses mean RA; PCWP uses mean LA as a proxy, without simulating catheter wedging. PV loops use ventricular transmural pressure.

V˙i=jNijQj,Nij={+1j enters i1j leaves i0otherwise,i=131Vi=TBV\dot V_i=\sum_j N_{ij}Q_j,\qquad N_{ij}=\begin{cases}+1&j\text{ enters }i\\-1&j\text{ leaves }i\\0&\text{otherwise}\end{cases},\qquad\sum_{i=1}^{31}V_i=TBV

The link tables define incidence matrix N. Each flow leaves one compartment and enters another, conserving total blood. Myocardium and pericardial fluid are excluded. Blood is not added or removed to match pressure or output at fixed TBV.

Model mechanisms

Myocardium generates tension; geometry and equilibrium determine pressure and blood flow. Waveforms are not prescribed. Shared equations and this case's effective coefficients follow.

Activation and calcium

Atrial and ventricular activation events drive calcium transients. Calcium generates pressure through myofilaments and chamber mechanics, not through a prescribed pressure waveform.

Equations and assumptions

Regular sinus rhythm includes atrial capture and ventricular conduction. Ventricles use a calibrated biexponential event source; atria retain a separate source. This is not a full action-potential or intracellular calcium-cycling model. Ca₀ is the long event-free limit, not the periodic trough.

x˙r=xr/τr,x˙d=xd/τd,[Ca]=Ca0+g(xdxr)\dot x_r=-x_r/\tau_r,\quad \dot x_d=-x_d/\tau_d,\qquad [Ca]=\mathrm{Ca}_0+g(x_d-x_r)

Implemented subset for distinct time constants. Events increment both drives equally; their difference gives a rise and subsequent decay.

xr,  xdx_r,\;x_d
Dimensionless calcium-drive states
τr,  τd\tau_r,\;\tau_d
Decay time constants (s)
Ca0,  g\mathrm{Ca}_0,\;g
Event-free calcium limit and amplitude coefficient (µM)

Activation times and state updates

Time is in seconds. In regular sinus rhythm T=60/HR; ventricular activation follows atrial activation by 120 ms (80 ms AV + 40 ms distal conduction). Calcium deposits occur another 12 ms after the corresponding activation. Each wall has two calcium states.

q is deposit strength, equal to 1 for all five walls at baseline. States decay analytically between events and both receive q at an event. The periodic formula below is for regular unit deposits; it does not initialize chamber volumes or Land states.

tA,k=tA,0+kT,tV,k=tA,k+0.120,tCa,w,k=tA or V,k+0.012t_{A,k}=t_{A,0}+kT,\quad t_{V,k}=t_{A,k}+0.120,\quad t_{Ca,w,k}=t_{A\text{ or }V,k}+0.012
xj(t+Δt)=xj(t)eΔt/τj,xj(tk+)=xj(tk)+qk(j=r,d)x_j(t+\Delta t)=x_j(t)e^{-\Delta t/\tau_j},\quad x_j(t_k^+)=x_j(t_k^-)+q_k\quad(j=r,d)
xj(tk+)=11eT/τj(periodic, qk=1)x_j(t_k^+)=\frac{1}{1-e^{-T/\tau_j}}\quad\text{(periodic, }q_k=1\text{)}

Interval-dependent calcium strength

Ventricles carry a discrete interval-dependent normalized load L, not measured SR calcium concentration. I is the preceding ventricular interval, a recovery and q the next calcium deposit. Coefficients and reference states are tabulated below. L and a at the reference heart rate form a fixed point with q=1.

The construction specified here is regular sinus rhythm without mechanical support. Optional ectopy, pacing and support-device modes are not included in this baseline equation system.

ak=1eIk/τrec,qk=akβLk,Lk+1=Lk(1r)qk+γ(1hak)a_k=1-e^{-I_k/\tau_{\mathrm{rec}}},\quad q_k=a_k\beta L_k,\qquad L_{k+1}=L_k-(1-r)q_k+\gamma(1-ha_k)
Adopted calcium source
Wallτr\tau_r (s)τd\tau_d (s)Ca0\mathrm{Ca}_0 (µM)g (µM)
Left atrium (LA)0.01250.30.06315457068130.604696429446
LV free wall (LVFW)0.118536374220.1317070824670.12070615927212.0780349398
Ventricular septum (SEP)0.118536374220.1317070824670.12070615927212.0780349398
RV free wall (RVFW)0.118536374220.1317070824670.12070615927212.0780349398
Right atrium (RA)0.01250.30.06315457068130.604696429446
Interval-strength coefficients and reference state
Symbol / meaningAdopted value
τrec (s)0.5
β0.8
r0.5
h0.2
γ0.598072916413
TrefT_{\mathrm{ref}} (s)0.857142857143
aref0.819907687852
Lref1.52456187266

Myofilaments: calcium to tension

Calcium binding, crossbridge populations, stretch and shortening velocity determine active tension. Force depends on length and loading history even at the same calcium level.

Equations and assumptions

Based on Land 2017, but not an unchanged reproduction. Ventricular stress scale, calcium affinity and transition rates are calibrated. An extension returns excess strong-bound population to the unbound pool at low calcium. Population conservation does not transfer the original paper's validation to this extension. Passive and viscoelastic components complete wall stress.

Ta=h(λ)Trefrs[S(1+ζs)+Wζw]T_a=\frac{h(\lambda)T_{\mathrm{ref}}}{r_s}\left[S(1+\zeta_s)+W\zeta_w\right]

Land-derived active fiber stress output (Pa): W/S are weak/strong populations, ζ their distortions and h a length-dependent factor. This is not cavity pressure.

Ta,  TrefT_a,\;T_{\mathrm{ref}}
Active fiber stress and its scale (Pa)
λ,  h(λ)\lambda,\;h(\lambda)
Fiber stretch and length-dependent factor (dimensionless)
W,  S,  rsW,\;S,\;r_s
Weak/strong populations and reference strong fraction
ζw,  ζs\zeta_w,\;\zeta_s
Dimensionless crossbridge distortions, dependent on shortening and history
Jexit=kmax(θnθn+cn)pmax(SrW,0),S˙exit=Jexit,U˙exit=JexitJ_{\mathrm{exit}}=k_{\mathrm{max}}\left(\frac{\theta^n}{\theta^n+c^n}\right)^p\max(S-rW,0),\quad \dot S|_{\mathrm{exit}}=-J_{\mathrm{exit}},\quad \dot U|_{\mathrm{exit}}=J_{\mathrm{exit}}

Added exit flux: c is calcium-bound troponin fraction, θ its reference, n/p exponents, kmax a rate (s⁻¹), r=kws/ksu the zero-distortion strong/weak ratio, and U the unbound fraction. Only positive excess is transferred.

Six states and conservation

Each wall has c (Ca-bound troponin), b (blocked), W/S (weak/strong binding) and distortions ζw/ζs. U=1−b−W−S is dependent. Valid states require 0<c≤1 and b,U,W,S≥0. b is unrelated to a hydraulic quadratic coefficient.

Ca is free calcium (µM); λ is the Land stretch. gwu/gsu are distortion-dependent detachment rates (s⁻¹). The added exit flux above applies to ventricular walls; Jexit=0 in atria. Detached population returns to U.

c˙=kTRPN{(Ca/Ca50)nTRPN(1c)c},b˙=kbmin(cnTm/2,100)UkucnTm/2b,W˙=kuwU(kwu+kws+gwu)W,S˙=kwsW(ksu+gsu)SJexit,ζ˙w=Awλ˙cwζw,ζ˙s=Asλ˙csζs.\begin{aligned}\dot c&=k_{\mathrm{TRPN}}\{(\mathrm{Ca}/\mathrm{Ca}_{50})^{n_{\mathrm{TRPN}}}(1-c)-c\},\\\dot b&=k_b\min(c^{-n_{\mathrm{Tm}}/2},100)U-k_u c^{n_{\mathrm{Tm}}/2}b,\\\dot W&=k_{\mathrm{uw}}U-(k_{\mathrm{wu}}+k_{\mathrm{ws}}+g_{\mathrm{wu}})W,\\\dot S&=k_{\mathrm{ws}}W-(k_{\mathrm{su}}+g_{\mathrm{su}})S-J_{\mathrm{exit}},\\\dot\zeta_w&=A_w\dot\lambda-c_w\zeta_w,\qquad\dot\zeta_s=A_s\dot\lambda-c_s\zeta_s.\end{aligned}
λc=min(λ,1.2),Ca50=Ca50,ref+β1(λc1),h(λ)=max{0,1+β0[λc+min(λc,0.87)1.87]},gwu=γwζw,gsu=γsmax(ζs1,ζs,0).\begin{aligned}\lambda_c&=\min(\lambda,1.2),\\\mathrm{Ca}_{50}&=\mathrm{Ca}_{50,\mathrm{ref}}+\beta_1(\lambda_c-1),\\h(\lambda)&=\max\{0,1+\beta_0[\lambda_c+\min(\lambda_c,0.87)-1.87]\},\\g_{\mathrm{wu}}&=\gamma_w|\zeta_w|,\quad g_{\mathrm{su}}=\gamma_s\max(-\zeta_s-1,\zeta_s,0).\end{aligned}

Derived rates and wall coupling

The following rates derive from the independent parameters below. rₛ/r𝓌 are reference fractions, θ=TRPN50. k/c rates are in s⁻¹ and A is dimensionless. Ca50 must remain positive. The λ=1.2 saturation and factor limit 100 are part of the adopted law.

With geometric log strain e, λ=s₀ exp(e), where s₀=1 for all baseline walls. The discrete velocity is (λn+1−λn)/Δt. Land Tₐ is multiplied by λ, orientation fo and active fraction fa to obtain Kirchhoff stress. fo=fa=1. Active controls scale Tref; passive controls scale passive stress and viscous modulus.

kb=kuθnTm(1rs)(1rw),kwu=kuw(1/rw1)kws,ksu=kwsrw(1/rs1),Aw=As=Aeffrs(1rs)rw+rs,cw=ϕkuw(1rw)rw,cs=ϕkws(1rs)rwrs,τf=λfofaTa+τpass+τvis.\begin{aligned}k_b&=\frac{k_u\theta^{n_{\mathrm{Tm}}}}{(1-r_s)(1-r_w)},\quad k_{\mathrm{wu}}=k_{\mathrm{uw}}(1/r_w-1)-k_{\mathrm{ws}},\\k_{\mathrm{su}}&=k_{\mathrm{ws}}r_w(1/r_s-1),\quad A_w=A_s=\frac{A_{\mathrm{eff}}r_s}{(1-r_s)r_w+r_s},\\c_w&=\frac{\phi k_{\mathrm{uw}}(1-r_w)}{r_w},\quad c_s=\frac{\phi k_{\mathrm{ws}}(1-r_s)r_w}{r_s},\\\tau_f&=\lambda f_o f_aT_a+\tau_{\mathrm{pass}}+\tau_{\mathrm{vis}}.\end{aligned}

These are model reference coefficients. Active multipliers apply to each wall's Tref; effective values are also tabulated in the passive/viscoelastic section.

Independent Land parameters: ventricles and atria
Symbol / meaningVentricular wallsAtriaUnit
kTRPNk_{\mathrm{TRPN}}Troponin rate100100s⁻¹
nTRPNn_{\mathrm{TRPN}}Ca binding exponent221
Ca50,ref\mathrm{Ca}_{50,\mathrm{ref}}Ca50,ref0.60.86µM
kuk_uThin-filament activation rate10001000s⁻¹
nTmn_{\mathrm{Tm}}Thin-filament exponent551
θ\thetareference bound fraction0.350.351
kuwk_{\mathrm{uw}}Unbound→weak104182s⁻¹
kwsk_{\mathrm{ws}}Weak→strong4.836s⁻¹
rwr_wreference weak fraction0.50.51
rsr_sreference strong fraction0.250.251
γs\gamma_sstrong distortion detachment8.58.5s⁻¹
γw\gamma_wweak distortion detachment615615s⁻¹
ϕ\phidistortion relaxation scale2.232.231
AeffA_{\mathrm{eff}}velocity sensitivity26.5251
β0\beta_0force-length dependence2.32.31
β1\beta_1affinity-length dependence-1.2-2.4µM
TrefT_{\mathrm{ref}}tension scale238816.54628111661.1510106Pa
TTSource temperature (not a variable kinetic factor)310.15310.15K
Derived Land coefficients
VentricleAtrium
kbk_b14.005833333314.0058333333
AwA_w10.610
AsA_s10.610
kwuk_{\mathrm{wu}}99.2146
ksuk_{\mathrm{su}}7.254
cwc_w231.92405.86
csc_s16.056120.42

Ventricular added exit: kmax=60 s⁻¹, p=16; atria have no added exit. Baseline settings compare adopted values with source values and calibration provenance. These are not independently measured constants from one subject.

Land et al. 2017

Passive mechanics and deformation history

Passive stretch resistance and time-dependent viscoelastic relaxation contribute to diastolic pressure alongside active stress.

Equations and assumptions

Ventricular passive stress derives from convex strain energy; atria use a separate law. A one-state Maxwell branch represents viscoelastic history. The ventricular prior references organ-level Klotz EDPVR, not direct tissue measurements.

e=lnλg,τpass=dΨde,τvis=Ev(eα),α˙=eατve=\ln\lambda_g,\quad \tau_{\mathrm{pass}}=\frac{d\Psi}{de},\quad \tau_{\mathrm{vis}}=E_v(e-\alpha),\quad \dot\alpha=\frac{e-\alpha}{\tau_v}

Passive and viscous relations. Land active stress is converted using its stretch and orientation/viability fractions before addition in the same stress convention.

λg,  e,  α\lambda_g,\;e,\;\alpha
Geometric stretch, logarithmic strain, and viscous strain
Ψ,  τpass,  τvis\Psi,\;\tau_{\mathrm{pass}},\;\tau_{\mathrm{vis}}
Energy density (J/m³) and passive/viscous Kirchhoff stresses (Pa)
Ev,  τvE_v,\;\tau_v
Viscoelastic modulus (Pa) and material time constant (s), distinct from pressure-derived LV τ

Ventricular elastic energy

Let e=ln λg, p=Hδ(e), q=Hδ(−e). K0 is central stiffness; Kt/a govern tensile stiffening; Kc adds compression stiffness. Hδ is a smooth positive part with u=z/δ. The three ventricular walls share this law; energy and derivatives are multiplied by each wall's passive scale sₚ,w, listed in the baseline settings.

Hδ(z)={0z0δ(u312u4)0<z<δzδ/2zδH_\delta(z)=\begin{cases}0&z\le0\\\delta(u^3-\tfrac12u^4)&0<z<\delta\\z-\delta/2&z\ge\delta\end{cases}
Ψ(e)=12K0e2+Kta2(eap1ap)+12Kcq2,τpass=sp,wdΨde\Psi(e)=\tfrac12K_0e^2+\frac{K_t}{a^2}(e^{ap}-1-ap)+\tfrac12K_cq^2,\qquad \tau_{\mathrm{pass}}=s_{p,w}\,\frac{d\Psi}{de}

Atrial passive law and viscoelasticity

Atria use a separate equibiaxial reduction; sₚ,w below is each wall's passive multiplier. λg=exp(e); C1/C2/C3 are Pa, C4 dimensionless. The fiber term is tension-only. Energy is Ψ(e)=∫₀ᵉτpass(s)ds. Supported strain is −0.5≤e≤0.5; applying the LA-derived material to RA is an extrapolation.

Each wall has a one-state Maxwell branch: α is viscous strain, Ev modulus, τv relaxation time. The implemented backward-Euler update below is applied once per mechanics step. Listed Ev already includes the passive scale.

τpass=sp,w[4C1(λg2λg4)+4C2(λg4λg2)+{C3(eC4(λg1)1)e>00e0]\tau_{\mathrm{pass}}=s_{p,w}\left[4C_1(\lambda_g^2-\lambda_g^{-4})+4C_2(\lambda_g^4-\lambda_g^{-2})+\begin{cases}C_3(e^{C_4(\lambda_g-1)}-1)&e>0\\0&e\le0\end{cases}\right]
αn+1=αn+(Δt/τv)en+11+Δt/τv,τvis,n+1=Ev(en+1αn+1)\alpha_{n+1}=\frac{\alpha_n+(\Delta t/\tau_v)e_{n+1}}{1+\Delta t/\tau_v},\qquad \tau_{vis,n+1}=E_v(e_{n+1}-\alpha_{n+1})
Reference passive coefficients (wall multipliers applied separately)
Symbol / meaningAdopted value
K0K_0 (Pa)1200
KtK_t (Pa)22345.9396953
a18
KcK_c (Pa)0
δ0.001
C1C_1 (Pa)1650
C2C_2 (Pa)0
C3C_3 (Pa)15
C4C_413.37
Wall multipliers and effective material coefficients
WallActive scalePassive scaleTrefT_{\mathrm{ref}} (Pa)EvE_v (Pa)τv\tau_v (s)
Left atrium (LA)1111661.151010659470.3
LV free wall (LVFW)0.351.0483585.791198519307.60.3
Ventricular septum (SEP)0.351.0483585.791198519307.60.3
RV free wall (RVFW)11.04238816.54628119307.60.3
Right atrium (RA)1111661.151010659470.3

The 0.3 s branch is a reduction informed by healthy ovine RV Prony data, not identified in every human wall. Acta Biomaterialia 2022

Klotz et al. 2006Organ EDPVR context, not the source of the adopted tissue constitutive law.

Moyer et al. 2015Starting point for atrial passive material, reduced here to equibiaxial deformation.

Chambers: tension to pressure

LV free wall, septum and RV free wall interact through shared geometry, with two atrial walls completing the five-wall model. Pressure follows stress and geometry, rather than a prescribed time waveform.

Equations and assumptions

Ventricles use spherical-cap TriSeg geometry and an energy-conjugate pressure mapping. Pericardial/thoracic external pressure is distinct from transmural pressure. Geometry is not patient-specific 3D imaging and does not resolve local stress or torsion. PV loops use transmural pressure while pressure waveforms use cavity pressure, so they differ when external pressure is present.

δW=wVm,wτf,wδlnλw,Pcavity=Ptm+Pext\delta W=\sum_w V_{\mathrm{m},w}\tau_{\mathrm{f},w}\,\delta\ln\lambda_w,\qquad P_{\mathrm{cavity}}=P_{\mathrm{tm}}+P_{\mathrm{ext}}

Wall virtual work is matched to cavity pressure–volume work. Septal position and junction radius also satisfy force balance.

δW,  Vm,w\delta W,\;V_{\mathrm{m},w}
Virtual work (J) and material volume of wall w (m³)
τf,w,  λw\tau_{\mathrm{f},w},\;\lambda_w
Fiber Kirchhoff stress (Pa) and geometric stretch; this stress τ is distinct from the LVP relaxation time
Pcavity,  Ptm,  PextP_{\mathrm{cavity}},\;P_{\mathrm{tm}},\;P_{\mathrm{ext}}
Cavity, transmural, and external pressures, added in the same units
LVRVSEPyJunction circlePositive toward RV
Schematic of spherical caps sharing a junction circle, not measured geometry or wall thickness. h is signed height from the junction plane to each cap apex.

Three spherical caps and fiber length

Mechanics uses SI units: m, m², m³, Pa. VL/VR are cavity blood volumes, Mw wall material volumes, vS the signed septal cap volume, y>0 the junction-circle radius. Wall material is not part of TBV. Cap height h is positive toward RV, usually negative for LV free wall.

For each wall solve h from cap volume, then area Aw, curvature κw and thickness correction zw. Aref,w is reference midwall area. zw is unrelated to Land crossbridge distortion ζ. Geometric ew enters the material law.

vL=VL12(ML+MS)+vS,vR=VR+12(MR+MS)+vSv_L=-V_L-\tfrac12(M_L+M_S)+v_S,\qquad v_R=V_R+\tfrac12(M_R+M_S)+v_S
vw=πhw(hw2+3y2)6,Aw=π(hw2+y2),κw=2hwhw2+y2v_w=\frac{\pi h_w(h_w^2+3y^2)}6,\quad A_w=\pi(h_w^2+y^2),\quad\kappa_w=\frac{2h_w}{h_w^2+y^2}
zw=3κwMw2Aw,ew=12lnAwAref,wzw2120.019zw4z_w=\frac{3\kappa_wM_w}{2A_w},\qquad e_w=\frac12\ln\frac{A_w}{A_{\mathrm{ref},w}}-\frac{z_w^2}{12}-0.019z_w^4

Internal balance and cavity pressures

F/G are generalized forces for cap volume and junction radius. Multiply each current material stress by geometric strain derivatives; do not differentiate active stress as though it were a conservative potential.

For supplied cavity volumes solve vS/y simultaneously from force balance, then obtain LV/RV transmural pressures. Atria use a spherical one-fiber reduction with reference cavity volume Vref. No prescribed pressure waveform or wall-specific pressure gain is added.

Fw=Mwτf,wewvwy,Gw=Mwτf,wewyvwF_w=M_w\tau_{\mathrm{f},w}\left.\frac{\partial e_w}{\partial v_w}\right|_y,\qquad G_w=M_w\tau_{\mathrm{f},w}\left.\frac{\partial e_w}{\partial y}\right|_{v_w}
FL+FS+FR=0,GL+GS+GR=0,Ptm,L=FL,Ptm,R=FRF_L+F_S+F_R=0,\quad G_L+G_S+G_R=0,\qquad P_{\mathrm{tm},L}=-F_L,\quad P_{\mathrm{tm},R}=F_R
eA=13lnVA+MA/2Vref,A+MA/2,Ptm,A=MAτf,A3(VA+MA/2)e_A=\frac13\ln\frac{V_A+M_A/2}{V_{\mathrm{ref},A}+M_A/2},\qquad P_{\mathrm{tm},A}=\frac{M_A\tau_{\mathrm{f},A}}{3(V_A+M_A/2)}

Thoracic and pericardial pressure

Occupied heart volume VH sums four cavity blood volumes, five wall volumes and prescribed pericardial fluid. Coronary blood is not added again to this occupancy formula. All cavities share the same external pressure. V0 is reference capacity, P* pressure scale and k stiffness.

H smooths engagement, δ=0.001 and u=(x+δ)/(2δ). H′ differentiates with respect to x. Baseline pressure offset/fluid/respiratory amplitudes are zero. With no engagement and zero thoracic pressure, cavity and transmural pressures coincide.

VH=c=LA,LV,RA,RVVc+wMw+Vfluid,x=(VHV0)/V0V_H=\sum_{c=LA,LV,RA,RV}V_c+\sum_w M_w+V_{\mathrm{fluid}},\quad x=(V_H-V_0)/V_0
H(x)={0xδδ(2u3u4)x<δxxδH(x)=\begin{cases}0&x\le-\delta\\\delta(2u^3-u^4)&|x|<\delta\\x&x\ge\delta\end{cases}
Pperi=Poffset+P[ekH(x)1]H(x),Pc=Ptm,c+Pth+PperiP_{peri}=P_{\mathrm{offset}}+P_*[e^{kH(x)}-1]H'(x),\qquad P_c=P_{\mathrm{tm},c}+P_{\mathrm{th}}+P_{peri}
Ventricular wall geometry constants
WallM (m³)ArefA_{\mathrm{ref}} (m²)
LV free wall (LVFW)0.00008384429580080.0107575058279
Ventricular septum (SEP)0.00004471695776040.00455984429148
RV free wall (RVFW)0.00003608736942070.012911294586
Atrial geometry constants
WallM (mL)VrefV_{\mathrm{ref}} (mL)
Left atrium (LA)25.982905982922.8900425619
Right atrium (RA)23.399810066532.5968711181
Pericardial constants
Symbol / meaningAdopted value
V0V_0 (m³)0.000600126542735
PP_* (Pa)500
k8
PoffsetP_{\mathrm{offset}} (Pa)0
VfluidV_{\mathrm{fluid}} (m³)0

BSA 1.9 m²; density 1053 kg/m³. LV free-wall/septal reference area and tissue volume are fixed at 1.15 and 1.25 times reference. RV free wall and atria are unchanged. This is neither evolving remodeling nor measured patient anatomy. Actual tissue occupies the unchanged pericardial bag; capacity, stiffness and fluid are not retuned.

Lumens et al. 2009 · TriSegBasis for three-wall geometry; the present constitutive laws and pressure mapping are not a reproduction of the entire original model.

Four valves: pressure difference to flow

Mitral, aortic, tricuspid and pulmonary valves share an opening and pressure-loss structure, with valve-specific forward and regurgitant areas.

Equations and assumptions

Flow is solved algebraically without a separate inertial flow state, while a bounded leaflet-opening memory remains. EOA already accounts for contraction/discharge; no extra Cd is applied. This construction has no local pressure-recovery correction. Raw AV/PV node differences are not Doppler maximum-jet or catheter peak-to-peak gradients.

ΔP=RQ+B(A)QQ,B(A)ρ/A2\Delta P=R Q+B(A)Q|Q|,\qquad B(A)\propto \rho/A^2

Constitutive law on the flowing branch; coefficients include unit conversions. Exact closure supports a pressure difference at Q=0, so this equation alone does not describe every branch.

ΔP,  Q\Delta P,\;Q
Simultaneous upstream-minus-downstream pressure (mmHg) and forward-positive flow (mL/s)
A,  ρA,\;\rho
Current effective orifice area and blood density; area depends on maximum area and opening fraction
R,  BR,\;B
Linear resistance (mmHg·s/mL) and quadratic loss coefficient (mmHg·s²/mL²)

Opening-state dynamics

ξ is opening fraction (0–1), ΔP upstream minus downstream pressure, d deadband, pₒ offset (0 mmHg in all baseline valves) and kₒ opening sensitivity. Fε smooths positive opening drive with ε=0.1 mmHg. Use the opening time constant when target exceeds previous ξ, otherwise the closing constant.

Fϵ(z)={0z0z2/(2ϵ)0<z<ϵzϵ/2zϵ,ξ=1ekoFϵ(ΔPdpo)F_\epsilon(z)=\begin{cases}0&z\le0\\z^2/(2\epsilon)&0<z<\epsilon\\z-\epsilon/2&z\ge\epsilon\end{cases},\quad \xi_\infty=1-e^{-k_oF_\epsilon(\Delta P-d-p_o)}
ξ˙=ξξτξ,ξn+1=ξn+(Δt/τξ)ξ(ΔPn+1)1+Δt/τξ\dot\xi=\frac{\xi_\infty-\xi}{\tau_\xi},\qquad \xi_{n+1}=\frac{\xi_n+(\Delta t/\tau_\xi)\xi_\infty(\Delta P_{n+1})}{1+\Delta t/\tau_\xi}

Area, direction and exact closure

Amax is maximal forward EOA; Ar closed regurgitant EOA. Forward area includes the residual gap plus the opening-dependent part; reverse area is Ar independent of ξ. Changing Ar also affects near-closure forward flow. Baseline Ar=0 in all valves.

Q is mL/s, A cm², pressure mmHg, ρ=1060 kg/m³, cP=133.322387415 Pa/mmHg. Zero area gives exactly Q=0 with supported pressure difference. No area floor or flow smoothing is introduced. R is not rescaled with area.

A={Ar+ξ(AmaxAr)ΔP0ArΔP<0,B(A)=ρ2cP(106104A)2A=\begin{cases}A_r+\xi(A_{\mathrm{max}}-A_r)&\Delta P\ge0\\A_r&\Delta P<0\end{cases},\quad B(A)=\frac{\rho}{2c_P}\left(\frac{10^{-6}}{10^{-4}A}\right)^2
Q={0A=0 or ΔP=0sgn(ΔP)2ΔPR+R2+4B(A)ΔPotherwiseQ=\begin{cases}0&A=0\text{ or }\Delta P=0\\\operatorname{sgn}(\Delta P)\frac{2|\Delta P|}{R+\sqrt{R^2+4B(A)|\Delta P|}}&\text{otherwise}\end{cases}
Four-valve coefficients (L=0)
ValveR (mmHg·s/mL)AmaxA_{\mathrm{max}} (cm²)ArA_r (cm²)kok_o (mmHg⁻¹)d (mmHg)τopen\tau_{\mathrm{open}} (s)τclose\tau_{\mathrm{close}} (s)
MV0.00275.5020.60.0240.016
AV0.00153.50300.0060.008
TV0.003580200.0180.01
PV0.00540200.010.006

Vessels: storage and flow

Compliance stores blood and resistance opposes flow. Systemic and pulmonary vessels form a closed circuit in which volume and pressure are solved together.

Equations and assumptions

Aortic and pulmonary roots use algebraic flow without local L. AoP and PAP are the Ao and PA node pressures, without a displayed ZcQ addition; SA is a downstream systemic-arterial compartment. Travelling/reflected waves, propagation delay and particular cuff/arterial-line sites are not simulated. Venous tone and TBV are not interchangeable controls: they affect filling and blood distribution.

dVidt=Qin,iQout,i,Ci(Ptm)=dVidPtm,i\frac{dV_i}{dt}=\sum Q_{\mathrm{in},i}-\sum Q_{\mathrm{out},i},\qquad C_i(P_{\mathrm{tm}})=\frac{dV_i}{dP_{\mathrm{tm},i}}

Volume conservation and differential compliance. Pressure–volume relations, including venous vessels, need not be linear with constant C.

Vi,  Qin,  QoutV_i,\;Q_{\mathrm{in}},\;Q_{\mathrm{out}}
Compartment volume (mL) and inflow/outflow (mL/s)
Ci,  Ptm,iC_i,\;P_{\mathrm{tm},i}
Pressure-dependent compliance (mL/mmHg) and transmural vascular pressure (mmHg)

Compartment pressure–volume laws

Pressure is mmHg, volume mL and flow mL/s. p=P−Pext is transmural pressure. Vu is volume at p=0; V−Vu is stressed volume. Listed Vs includes stiffness/compliance scaling: original Vs×0.65/1.42 for Ao/SA/Art, original Vs/1.42 for PA/PArt.

Venous-type compartments join collapsed Cc, open Co and distended Cd smoothly. S(z)=ln(1+exp z), σ(z)=1/(1+exp(−z)); ΔS(p;a,d)=S((p−a)/d)−S(−a/d). Tone u changes Vu=Vu,original−G u. The table includes baseline u=0.15.

The venous inverse saturates at −20/45 mmHg; arterial strain is bounded below by ln(0.05). These are numerical boundaries, not normal ranges; saturated results must not be extrapolated physiologically. Compliance readback is floored at 10⁻⁴ mL/mmHg without replacing V(p).

p=P0(emax[(VVu)/Vs,ln0.05]1)(arterial),p=(VVu)/C(linear)p=P_0\left(e^{\max[(V-V_u)/V_s,\ln0.05]}-1\right)\quad\text{(arterial)},\qquad p=(V-V_u)/C\quad\text{(linear)}
V(p)=Vu+Ccp+(CoCc)doΔS(p;po,do)(CoCd)dsΔS(p;ps,ds)V(p)=V_u+C_cp+(C_o-C_c)d_o\Delta S(p;p_o,d_o)-(C_o-C_d)d_s\Delta S(p;p_s,d_s)
dVdp=Cc+(CoCc)σ(ppodo)(CoCd)σ(ppsds)\frac{dV}{dp}=C_c+(C_o-C_c)\sigma\left(\frac{p-p_o}{d_o}\right)-(C_o-C_d)\sigma\left(\frac{p-p_s}{d_s}\right)
Effective arterial and linear coefficients
CompartmentVuV_u (mL)P0P_0 (mmHg)VsV_s (mL)C (mL/mmHg)
Proximal aorta (Ao)05068.661971831
Systemic arteries (SA)050183.098591549
Systemic resistance-vessel side (Art)04554.9295774648
Systemic capillary bed (Cap)015
Proximal pulmonary artery (PA)02042.2535211268
Pulmonary resistance-vessel side (PArt)02063.3802816901
Effective venous-type coefficients
CompartmentVuV_u (mL)Cc/Co/CdC_c / C_o / C_d (mL/mmHg)po/psp_o / p_s (mmHg)do/dsd_o / d_s (mmHg)Vu,元 / G (mL)
Systemic veins (SV)1538.40915 / 130 / 35-2 / 161.5 / 41653.909 / 770
Vena cava (VC)150.0915 / 45 / 12-1 / 121 / 3169.591 / 130
Pulmonary capillary bed (PCap)1051 / 2 / 10 / 141 / 3105 / 0
Pulmonary venular side (PVen)1601.2 / 3 / 1.2-1 / 141 / 3160 / 0
Pulmonary vein / LA inlet (PVein)2151.5 / 4 / 1.5-1 / 141 / 3215 / 0
Main-circuit links and effective non-valve resistance
Link: positive directionR (mmHg·s/mL)Multiplier groupAdditional law
Left atrium (LA) → Left ventricle (LV)See valve tableValve opening / direction
Left ventricle (LV) → Proximal aorta (Ao)See valve tableValve opening / direction
Right atrium (RA) → Right ventricle (RV)See valve tableValve opening / direction
Right ventricle (RV) → Proximal pulmonary artery (PA)See valve tableValve opening / direction
Proximal aorta (Ao) → Systemic arteries (SA)0.05581056Systemic ×1.2
Systemic arteries (SA) → Systemic resistance-vessel side (Art)0.089296896Systemic ×1.2
Systemic resistance-vessel side (Art) → Systemic capillary bed (Cap)0.72553728Systemic ×1.2
Systemic capillary bed (Cap) → Systemic veins (SV)0.15
Systemic veins (SV) → Vena cava (VC)0.05
Vena cava (VC) → Right atrium (RA)0.04waterfall + χ (Pth)
Proximal pulmonary artery (PA) → Pulmonary resistance-vessel side (PArt)0.00625Pulmonary ×0.625
Pulmonary resistance-vessel side (PArt) → Pulmonary capillary bed (PCap)0.025Pulmonary ×0.625
Pulmonary capillary bed (PCap) → Pulmonary venular side (PVen)0.03waterfall + χ (Palv)
Pulmonary venular side (PVen) → Pulmonary vein / LA inlet (PVein)0.01
Pulmonary vein / LA inlet (PVein) → Left atrium (LA)0.03025

Vascular coefficients and resistance partition are lumped construction/calibration values, not direct measurements at specified human vascular sites. Do not apply baseline multipliers a second time to the listed R.

Coronary and external-pressure coupling

Coronary inflow and venous return contribute to global volume balance. Myocardial compression modulates coronary flow, while pericardial and thoracic pressures load cavities and vessels externally.

Equations and assumptions

Coronary territories and layers include resistance/storage, intramyocardial pressure, collapse and autoregulation, some with provisional adult priors. Baseline has no mechanical support, valve regurgitation or respiratory oscillation. Fixed-control preload testing holds HR and tone fixed; it is not a bedside fluid challenge with autonomic and whole-body responses.

AoArtC1C1C2C2CVRAR1RmR2SubepicardialSubendocardial
This branch is repeated for LAD/LCx/RCA, sharing only CV. Art/CV receive common cardiac external pressure; C1/C2 receive territory/layer-specific intramyocardial pressure.

Coronary storage and resistance

Each territory has one large-arterial Art reservoir, splitting into epi/endo paths with proximal C1 and distal C2 storage, then merging into common venous CV and RA. C1/C2 name compartments, not numerical compliances. Sixteen coronary volumes contribute to TBV without duplication.

All coronary storage uses v=V/Vref, m=4, n=2; Cref is tangent compliance at Vref. Collapse resistance uses a different hydraulic reference Vh. With a=0.67, x=min(1,max(0,V/Vh)), compute f below.

Branch flow is pressure difference divided by effective resistance. R1 uses tone θ and C1 collapse; Rm the geometric mean of C1/C2 collapse; R2 C2 collapse. Structural multipliers are 1 and focal stenosis additions 0 at baseline.

p=P(vmvn),P=VrefCref(m+n),V>0p=P_*(v^m-v^{-n}),\quad P_*=\frac{V_{\mathrm{ref}}}{C_{\mathrm{ref}}(m+n)},\quad V>0
f(V)={a+(1a)x2(32x)}2f(V)=\{a+(1-a)x^2(3-2x)\}^{-2}
R1,eff=R1θf(VC1),Rm,eff=Rmf(VC1)f(VC2),R2,eff=R2f(VC2)R_{1,eff}=R_1\theta f(V_{C1}),\quad R_{\mathrm{m},eff}=R_m\sqrt{f(V_{C1})f(V_{C2})},\quad R_{2,eff}=R_2f(V_{C2})

Intramyocardial pressure and beat-wise regulation

Art/CV external pressure is Pe=Pth+Pperi; C1/C2 use PIM. Weights wL/wS/wR, depth d and shortening gain K are tabulated. Septal depth s=d for LAD/LCx, 1−d for RCA. eMVC,w updates at the previous accepted mitral closure. F is the valve-section positive part with width 0.005.

The shortening reference updates when accepted MV flow changes from >1 to ≤1 mL/s, using that endpoint's strains for subsequent steps. This is not an imaging leaflet-contact time or an interpolated metric closure time.

Tone is held within a cycle and updated at completion. Q̄m is signed cycle-mean flow through Rm; Qtarget is the resting target. ℓ=ln θ, T cycle length, τa=25 s, θmin=4/45 and θmax=2. Baseline demand=1 and hyperemia=0 give the update below. It does not autonomously alter contractility from oxygen demand.

PIM=Pe+(wLd+wSs)Ptm,L+[wRd+wS(1s)]Ptm,R+KwwwF0.005(1eeweMVC,w)P_{\mathrm{IM}}=P_e+(w_Ld+w_Ss)P_{\mathrm{tm},L}+[w_Rd+w_S(1-s)]P_{\mathrm{tm},R}+K\sum_w w_wF_{0.005}(1-e^{e_w-e_{\mathrm{MVC},w}})
Qˉm,k=1Tn in cycle kΔtnQm,n+1\bar Q_{\mathrm{m},k}=\frac{1}{T}\sum_{n\text{ in cycle }k}\Delta t_n Q_{m,n+1}
k+1=cliplnθmin,lnθmax[k+Tτalnmax(Qˉm,kQtarget,0.05)]\ell_{k+1}=\operatorname{clip}_{\ln\theta_{\mathrm{min}},\ln\theta_{\mathrm{max}}}\left[\ell_k+\frac{T}{\tau_a}\ln\max\left(\frac{\bar Q_{\mathrm{m},k}}{Q_{\mathrm{target}}},0.05\right)\right]
Coronary compartment PV constants
CompartmentVrefV_{\mathrm{ref}} (mL)CrefC_{\mathrm{ref}} (mL/mmHg)PP_* (mmHg)VhV_h (mL)
LAD large arterial storage (Art)1.433910583980.00153221919053155.9731785161.53221919053
LAD epi C10.481074522860.003095967391325.89791073660.496293624668
LAD epi C20.2764290256570.09287902173910.4960377856430.496293624668
LAD endo C10.481074522860.003095967391325.89791073660.496293624668
LAD endo C20.2764290256570.09287902173910.4960377856430.496293624668
LCx large arterial storage (Art)0.9432187491580.00100788562723155.9731785161.00788562723
LCx epi C10.3164482602850.0020365108695725.89791073660.32645930444
LCx epi C20.1818335415920.0610953260870.4960377856430.32645930444
LCx endo C10.3164482602850.0020365108695725.89791073660.32645930444
LCx endo C20.1818335415920.0610953260870.4960377856430.32645930444
RCA large arterial storage (Art)1.010844310020.00108014758225155.9731785161.08014758225
RCA epi C10.3391365190840.0021825217391325.89791073660.349865320892
RCA epi C20.1948703851070.06547565217390.4960377856430.349865320892
RCA endo C10.3391365190840.0021825217391325.89791073660.349865320892
RCA endo C20.1948703851070.06547565217390.4960377856430.349865320892
Common coronary venous storage (CV)4.129948287330.0980374927.02103552954.9018746
All coronary links and reference resistance (before tone / collapse)
Link: positive directionR (mmHg·s/mL)
Proximal aorta (Ao) → LAD large arterial storage (Art)20.749926765
LAD large arterial storage (Art) → LAD epi C158.0014068184
LAD epi C1 → LAD epi C235.7263939068
LAD epi C2 → Common coronary venous storage (CV)11.9087979689
LAD large arterial storage (Art) → LAD endo C142.2821588661
LAD endo C1 → LAD endo C232.1859404566
LAD endo C2 → Common coronary venous storage (CV)10.7286468189
Proximal aorta (Ao) → LCx large arterial storage (Art)31.1248901474
LCx large arterial storage (Art) → LCx epi C194.8761406353
LCx epi C1 → LCx epi C253.5895908603
LCx epi C2 → Common coronary venous storage (CV)17.8631969534
LCx large arterial storage (Art) → LCx endo C164.2296097205
LCx endo C1 → LCx endo C248.2789106849
LCx endo C2 → Common coronary venous storage (CV)16.0929702283
Proximal aorta (Ao) → RCA large arterial storage (Art)29.049897471
RCA large arterial storage (Art) → RCA epi C189.1455943902
RCA epi C1 → RCA epi C250.0169514696
RCA epi C2 → Common coronary venous storage (CV)16.6723171565
RCA large arterial storage (Art) → RCA endo C174.0663831102
RCA endo C1 → RCA endo C245.0603166393
RCA endo C2 → Common coronary venous storage (CV)15.0201055464
Common coronary venous storage (CV) → Right atrium (RA)2.44019138756
Coronary coupling weights and shortening gain
TerritorywL/wS/wRw_L / w_S / w_RK (mmHg)Rest Qtarget: epi / endo (mL/s)
LAD0.75 / 0.25 / 081.15464997120.485355450237 / 0.538744549763
LCx1 / 0 / 079.65412002480.323570300158 / 0.359163033175
RCA0 / 0.2 / 0.884.20784419340.346682464455 / 0.384817535545

Layer depths are 0.25 (epi) and 0.75 (endo). Volume allocation starts from porcine morphometry and C1/C2 from canine effective compliance, with selected adjustments from beating-boundary checks. This is not a uniquely identified human circuit.

Kassab et al. 1994 · Spaan et al. 2000

Oxygen supply and consumption

Beat-mean flow, hemoglobin, inspired oxygen and prescribed consumption determine systemic oxygen balance separately from myocardial contraction.

Equations and assumptions

Alveolar gas, oxygen dissociation and shunt mixing determine arterial content. Fick balance gives required mixed-venous content; negative required content is infeasible. Local diffusion and full metabolic adaptation are not resolved.

D˙O2=10COCaO2,V˙O2=10CO(CaO2CvO2)\dot D_{\mathrm{O_2}}=10\,\mathrm{CO}\,C_{\mathrm{aO}_2},\qquad \dot V_{\mathrm{O_2}}=10\,\mathrm{CO}(C_{\mathrm{aO}_2}-C_{\mathrm{vO}_2})

CO\mathrm{\mathrm{CO}}
Beat-mean systemic flow (L/min)
CaO2,  CvO2C_{\mathrm{aO}_2},\;C_{\mathrm{vO}_2}
Arterial and mixed-venous oxygen content (mL O₂/dL)
D˙O2,  V˙O2\dot D_{\mathrm{O}_2},\;\dot V_{\mathrm{O}_2}
Oxygen delivery and prescribed consumption (mL O₂/min); 10 converts L to dL

Beat-mean oxygen balance

Oxygen transport is an algebraic beat-mean readout without feedback to blood volume or tension. P is O₂ pressure (mmHg), S saturation, Hb g/dL, C mL O₂/dL. Alveolar and end-capillary PO₂ are equated. PB is barometric pressure and R respiratory exchange ratio.

CO is L/min, VO₂ mL/min and s an oxygen-mixing shunt fraction, not an added hydraulic connection. End-capillary Cc, venous Cv and arterial Ca satisfy Fick balance and content mixing. Nonpositive flow or negative required content makes the evaluation unavailable.

PAO2=FIO2(PB47)PaCO2/R>0,S(P)=P2.726.82.7+P2.7P_{\mathrm{AO}_2}=F_{\mathrm{IO}_2}(P_B-47)-P_{\mathrm{aCO}_2}/R>0,\quad S(P)=\frac{P^{2.7}}{26.8^{2.7}+P^{2.7}}
C(P)=1.34HbS(P)+0.0031P,D=VO210CO,Ca=Ccs1sD,Cv=CaDC(P)=1.34\,Hb\,S(P)+0.0031P,\quad D=\frac{VO_2}{10\mathrm{CO}},\quad C_a=C_c-\frac{s}{1-s}D,\quad C_v=C_a-D

Analysis: relationships beyond one loop

A PV loop is the simulated beat trajectory. ESPVR, EDPVR, Starling/Guyton curves and PVA/PE are derived through separate loading protocols.

Equations and assumptions

Analyses branch without mutating the live simulation. ESPVR uses the reduced-preload limb through baseline; EDPVR and Starling retain high-volume conditions too. Estimates depend on the finite loading range, settlement and event definitions; linearity and load independence are not guaranteed. Unavailable analysis stays unassessed, not an exact-output placeholder.

Coupled equations and saved initial state

Blood occupies 31 compartments, constrained by fixed TBV (30 independent volume degrees of freedom). Each wall has six Land states, one viscous strain and two calcium states; four valves have opening states. Six coronary tones and event/load/previous-MVC memories are retained. Pressures, flows and internal geometry follow simultaneous algebraic constraints.

Vn+1Vn=ΔtNQn+1,zn+1zn=Δtf(zn+1,Can+1,λn+1,λ˙n+1),0=g(Vn+1,zn+1,Pn+1,Qn+1,vS,n+1,yn+1).\begin{aligned}V_{n+1}-V_n&=\Delta t\,NQ_{n+1},\\z_{n+1}-z_n&=\Delta t\,f(z_{n+1},Ca_{n+1},\lambda_{n+1},\dot\lambda_{n+1}),\\0&=g(V_{n+1},z_{n+1},P_{n+1},Q_{n+1},v_{S,n+1},y_{n+1}).\end{aligned}

z collects Land/viscous/opening states; f is the documented evolution and g the constitutive, hydraulic and force-balance constraints. The coupled system uses backward Euler, with exact inter-event calcium propagation. Candidate volume, geometry, stress, pressure and flow must satisfy continuity together; vascular updates do not keep stale cavity pressures.

Nominal step is 2 ms, split at activation, calcium and control-window boundaries. Coronary tone updates from completed-cycle flow integrals. Invalid populations, nonfinite states, volume imbalance or failed nonlinear solves are not accepted. Another integrator may approximate the same continuous equations, but finite-step peaks/events and saved baseline parity require separate checks.

Saved setting's initial conditions

The settled launch state below is at t₀=133.716 s。If shifting time to zero, shift event and control times equally. Keeping volumes but zeroing material/Ca states is not equivalent. Display uses 12 significant digits; CSV retains full stored precision.

Initial blood volume (mL)
CompartmentV(t0)V(t_0)
Proximal aorta (Ao)60.7974182576
Systemic resistance-vessel side (Art)49.5877122232
Systemic capillary bed (Cap)313.053642764
Left atrium (LA)57.3167860333
Left ventricle (LV)213.427061319
Proximal pulmonary artery (PA)27.1194113675
Pulmonary resistance-vessel side (PArt)40.731619674
Pulmonary capillary bed (PCap)136.094800018
Pulmonary vein / LA inlet (PVein)278.462837908
Pulmonary venular side (PVen)207.053540772
Right atrium (RA)30.7552157612
Right ventricle (RV)125.003400925
Systemic arteries (SA)161.089364255
Systemic veins (SV)2782.872251
Vena cava (VC)439.462971682
Common coronary venous storage (CV)4.55465316263
LAD large arterial storage (Art)1.50693981091
LAD endo C10.539500512265
LAD epi C10.539906216416
LAD endo C20.1749080481
LAD epi C20.347495239095
LCx large arterial storage (Art)0.991648299263
LCx endo C10.355230192399
LCx epi C10.353524530672
LCx endo C20.118505343692
LCx epi C20.274266149514
RCA large arterial storage (Art)1.06060659051
RCA endo C10.381837588606
RCA epi C10.379841461835
RCA endo C20.29507311099
RCA epi C20.298029779845
Initial Land states (dimensionless)
WallcbWSζw\zeta_wζs\zeta_s
Left atrium (LA)0.6301187521560.05235958079670.3393865592460.226111867685-0.0376362413446-0.132825098168
LV free wall (LVFW)0.09578461064890.9954527313420.001721036182420.0008736379585770.008682860454240.198264518171
Ventricular septum (SEP)0.085709384250.9965356681940.00130340133530.0006802255121460.00909110690660.19738001841
RV free wall (RVFW)0.06073906095210.9984349120270.0005927842450040.0003429024325990.001897679431410.126710084624
Right atrium (RA)0.2360555658020.8558921994580.04982386256240.04582559269010.01283291894680.0397999847835
Initial strain / calcium states (dimensionless)
Walle(t0)e(t_0)α(t0)\alpha(t_0)xr(t0)x_r(t_0)xd(t0)x_d(t_0)
Left atrium (LA)0.2242224238720.3286694496540.00005904900549980.7071118549
LV free wall (LVFW)0.1487231871060.1097409319740.9863558500240.988542892627
Ventricular septum (SEP)0.1301361206950.08479446835760.9863558500240.988542892627
RV free wall (RVFW)0.06168667131130.0111844128260.9863558500240.988542892627
Right atrium (RA)-0.01415478820430.04276069044070.00005904900549980.7071118549
Initial opening / geometry
QuantityValue
AV ξ8.3941454043e-26
MV ξ0.588170415645
PV ξ5.4145877253e-35
TV ξ0.0286221508103
vSv_S (m³)0.0000564792328315
y (m)0.0369639529204
Coronary tone and partial-window integrals
Territory / layerθ(t0)\theta(t_0)Qmdt\int Q_m\,dt (mL)
LAD epi0.7924425384890.000873481607701
LAD endo0.6939309027830.000876155822331
LCx epi0.8276932387720.000539011765354
LCx endo0.6910375090450.000591086208632
RCA epi0.7877185420870.000591683288816
RCA endo0.7922384387070.0006995759526
Retained timing and discrete memory
QuantityValue
Next atrial activation (s)134.439428571
Last ventricular activation (s)133.702285714
L(t0)L(t_0)1.52456187266
Coronary window start (s)133.714285714
Elapsed coronary window (s)0.00171428571431
eMVC,LVFWe_{\mathrm{MVC},\mathrm{LVFW}}0.151058531376
eMVC,RVFWe_{\mathrm{MVC},\mathrm{RVFW}}0.062000493637
eMVC,SEPe_{\mathrm{MVC},\mathrm{SEP}}0.132741561282

There are no pending calcium/conduction events at this instant. Next calcium deposits follow the next atrial activation by 12 ms (atria) or 132 ms (ventricles). Coronary windows are T=60/70 s from original time zero. Recompute several cycles to verify consistency; this table does not establish uniqueness or stability in every scenario.

Save parameter and initial-state tables (CSV)

Settings

BSA 1.9 m²; density 1053 kg/m³. LV free-wall/septal reference area and tissue volume are fixed at 1.15 and 1.25 times reference. RV free wall and atria are unchanged. This is neither evolving remodeling nor measured patient anatomy. Actual tissue occupies the unchanged pericardial bag; capacity, stiffness and fluid are not retuned.

HR
70 bpm
TBV
4935 mL
BSA
1.9 m²
LV tissue mass
135.4 g

LV free-wall/septal active tension is 0.35/0.35 and systemic resistance 1.2 times reference. One is a model reference, not an absolute unit of normal contractility. This tissue mass does not establish pathological hypertrophy. Calcium, crossbridge kinetics and passive coefficients are unchanged.

Complete inputs and material provenance
Saved primitive inputs
ParameterValueUnit
Heart rate (HR)70bpm
Total blood volume (TBV)4,935mL
Systemic vascular resistance (SVR)1.21
Common ventricular active tensionSee individual walls1
Venous tone0.151
PEEP0cmH2O
Pulmonary vascular resistance (PVR)0.6251
Arterial stiffness1.421
LA active tension11
LV free wall active tension0.351
Ventricular septum active tension0.351
RV free wall active tension11
RA active tension11
LA passive stiffness11
LV free wall passive stiffness1.041
Ventricular septum passive stiffness1.041
RV free wall passive stiffness1.041
RA passive stiffness11
Mitral valve (MV) maximum EOA5.5cm2
Mitral valve (MV) reverse EROA0cm2
Aortic valve (AoV) maximum EOA3.5cm2
Aortic valve (AoV) reverse EROA0cm2
Tricuspid valve (TV) maximum EOA8cm2
Tricuspid valve (TV) reverse EROA0cm2
Pulmonary valve (PV) maximum EOA4cm2
Pulmonary valve (PV) reverse EROA0cm2
Hemoglobin (Hb)15g/dL
Inspired oxygen fraction (FiO₂)0.211
Arterial PCO₂ (PaCO₂)40mmHg
Respiratory exchange ratio (RER)0.81
Barometric pressure760mmHg
True shunt fraction0.021
Target oxygen consumption (VO₂)250mL O2/min
pericardium.reference-capacity-scale11
pericardium.pressure-scale11
pericardium.exponential-stiffness-scale11
pericardium.prescribed-fluid-volume-ml0mL
coronary.focal-diameter-loss-fraction.LAD01
coronary.focal-diameter-loss-fraction.LCx01
coronary.focal-diameter-loss-fraction.RCA01
coronary.structural-r1-resistance-scale.LAD.subepicardial11
coronary.structural-r1-resistance-scale.LAD.subendocardial11
coronary.structural-r1-resistance-scale.LCx.subepicardial11
coronary.structural-r1-resistance-scale.LCx.subendocardial11
coronary.structural-r1-resistance-scale.RCA.subepicardial11
coronary.structural-r1-resistance-scale.RCA.subendocardial11
coronary.structural-rm-resistance-scale.LAD.subepicardial11
coronary.structural-rm-resistance-scale.LAD.subendocardial11
coronary.structural-rm-resistance-scale.LCx.subepicardial11
coronary.structural-rm-resistance-scale.LCx.subendocardial11
coronary.structural-rm-resistance-scale.RCA.subepicardial11
coronary.structural-rm-resistance-scale.RCA.subendocardial11
LV contractility0.351

Grouped contractility sets the individual wall inputs; it does not multiply them again. Cases retain primitive inputs, fixed anatomy and state, not only knob positions. Switching geometry loads the target case's own state.

Land et al. 2017
Material calibration before case tension scaling
ParameterSourceModel reference
kTRPN0.1 1/ms100 1/s
nTRPN2 dimensionless2 dimensionless
CaT50Ref0.805 uM0.6 uM
ku1 1/ms1000 1/s
nTm5 dimensionless5 dimensionless
TRPN500.35 dimensionless0.35 dimensionless
kuw0.182 1/ms104 1/s
kws0.012 1/ms4.8 1/s
rw0.5 dimensionless0.5 dimensionless
rs0.25 dimensionless0.25 dimensionless
gammaS0.0085 1/ms8.5 1/s
gammaW0.615 1/ms615 1/s
phi2.23 dimensionless2.23 dimensionless
Aeff25 dimensionless26.5 dimensionless
beta02.3 dimensionless2.3 dimensionless
beta1-2.4 uM-1.2 uM
Tref120 kPa238816.54628141236 Pa
temperatureK37 degC310.15 K
Fixed pericardial and coronary construction

Added tissue occupies the same pericardial bag, without automatic enlargement. Coronary reference storage/resistance and absolute oxygen demand are not rescaled to current mass. Adequate perfusion or oxygen supply in the enlarged heart is not established.

Case assessment

Rest comparisons and launch states were calculated in this exact model. PVA, passive curves, control diagnostics and slow-tail observations refer to prior experiments with the same construction, retaining their original model identity and provenance.

2 ms · launch grid

Both grids meet required conditions and case preferences. This qualifies a selected operating point, not a diagnosis or clinical validation. Population statistics and model selection intervals are distinct.

Saved baseline and case comparison
MetricComparator baseline2 ms, fixedThis case2 msUnit
LVEF55.828.5%
LV EDVI75.6113.4mL/m²
LV ESVI33.581.1mL/m²
CI2.952.26L/min/m²
SVI42.232.3mL/m²
Ao node mean93.681.0mmHg
mean LAP8.416.6mmHg
mean RAP3.14.3mmHg
PA node mean17.923.9mmHg
RVEF56.949.0%
RV EDVI74.165.8mL/m²
RV ESVI31.933.5mL/m²
LV end-filling P10.925.5mmHg
LV end-filling Ptm10.922.9mmHg
LV +dP/dt2,5641,113mmHg/s
LV −dP/dt-1,539-824mmHg/s
ICT89.148.0ms
ET258.0280.0ms
IRT92.0128.0ms
Tei0.700.63
E/A (flow)0.940.86
τ Weiss32.051.3ms
τ Glantz52.4ms

The comparator baseline remains its saved 2 ms record; selecting the 1 ms case does not recompute it. Unrounded values are included in the JSON export.

This is not a load-matched comparison. Reduced EF can coexist with forward output; a smaller Tei than baseline does not demonstrate better contractility. Mean LA, end-filling cavity and transmural pressures differ. E/A uses volumetric flow, not Doppler velocity.

Intervals and rationale

LVEF · Required28.46 / 20–40 % · Met

Reduced EF defines the chosen phenotype. The 20% lower bound selects severity; it is not a diagnostic lower limit.

Method: native-valve-closure-volumes

LV EDVI · Required113.4 / 90–180 mL/m² · Met

Selects a chronic LV-dilated volume range, not a requirement for all HFrEF or an imaging diagnostic boundary.

Method: native-valve-closure-volumes

CI · Required2.258 / 1.8–3.5 L/min/m² · Met

Allows lower or retained output; reduced EF is not equated with extreme low output.

Method: exact-beat-aortic-valve-net-flow

mean LAP · Required16.61 / 3–25 mmHg · Met

Retains the broad screen against extreme filling. Elevated pressure is a case preference, not a required condition.

Method: exact-beat-mean-left-atrial-pressure

mean RAP · Required4.261 / 0–12 mmHg · Met

Selects loading for an LV-dominant case; it proves neither normal RVEF nor absence of RV involvement.

Method: exact-beat-mean-right-atrial-pressure

Ao node mean · Required81.04 / 60–110 mmHg · Met
LVEF · Case preference28.46 / 25–35 % · Met

Meet EF jointly with the other primary preferences, without ranking EF above all other objectives.

Method: native-valve-closure-volumes

LV EDVI · Case preference113.4 / 110–150 mL/m² · Met

Selects dilation and a large residual volume. ESVI is not added as a duplicate objective.

Method: native-valve-closure-volumes

CI · Case preference2.258 / 2.2–2.8 L/min/m² · Met

Selects one case with retained resting net output, not a disease-specific normal CI interval.

Method: exact-beat-aortic-valve-net-flow

Ao node mean · Case preference81.04 / 80–100 mmHg · Met

A resting arterial-pressure preference, strongly dependent on output and vascular load. Missing it does not invalidate HFrEF; the lower bound was not lowered to admit a known 77-mmHg candidate.

Method: exact-beat-mean-aortic-node-pressure

mean LAP · Case preference16.61 / 12–20 mmHg · Met

Intentionally selects higher filling than the mean PAWP of 10–12 mmHg in treated cohorts. Lavine supports the existence of such cases, not a frequency weight. Pressure alone does not diagnose edema or clinical congestion; native LV end-filling cavity and transmural pressures are also reported. Higher is not rewarded beyond the interval.

Method: exact-beat-mean-left-atrial-pressure

τ Weiss · Case preference51.31 / 40–75 ms · Met

A broad design interval informed by 54±14 ms, not a normal/abnormal boundary. Use only quality-qualified Weiss fits and retain the window and baseline comparison. With calcium/crossbridge rates fixed, tau changes can arise from geometry, tension, load and regression window; intrinsic relaxation change is not identified. No substitution by Glantz or extra reward for longer tau.

Method: main-wire-lv-relaxation-tau-v1:Weiss-zero-asymptote-time-weighted-log-linear

Relaxation, pressure rates and filling
P(t)=P+Ae(tt0)/τ,Tei=ICT+IRTETP(t)=P_{\infty}+A e^{-(t-t_0)/\tau},\qquad \mathrm{Tei}=\frac{ICT+IRT}{ET}

Tau is fitted to cavity-pressure decay. Weiss fixes the asymptote at zero; Glantz estimates it. Glantz does not pass fit quality here and is shown as unavailable. Weiss is prolonged despite unchanged calcium/crossbridge kinetics; effects of loading, geometry and the observation window are not identified intrinsic relaxation disease.

Weiss: 51.31 ms · R² 0.9985 · window 25.98 ms · 15 samples.

The window runs from the midpoint of the minimum-dP/dt interval to the next end-filling LV pressure plus 5 mmHg, before mitral opening. Time-weighted linear regressions use log-pressure versus time for Weiss and dP/dt versus pressure for Glantz. Required support is 6 samples, 15 ms and 10 mmHg decay; minimum R² is 0.97/0.95, and reconstructed-pressure RMSE must be ≤5% of the decay. These are measurement-usability design rules, not disease-normality limits.

Glantz R² describes the dP/dt-versus-pressure regression. A small reconstructed-pressure error does not override that fit-quality requirement.

dP/dt uses accepted-step cavity-pressure differences divided by elapsed time, not an analytic derivative or a filtered sensor. ICT/ET/IRT follow valve events/forward flow, not interchangeable clinical acquisition methods.

Flow E/A is 0.8579, A-dominant as in baseline (0.9379). Lavine's high-filling-pressure DCM group had a greater early filling fraction and a lower atrial fraction, a different direction. The flow-versus-Doppler distinction alone does not resolve this mismatch. A-dominance is a model inflow trait shared with baseline, not an established HFrEF finding reproduced here.

End-filling LV cavity pressure exceeds mean LA pressure by 8.847 mmHg. Kato's difference of cohort means was 4 mmHg, not a distribution of individual differences. Mean LA is not substituted for PCWP or end-filling LV pressure; this discrepancy remains an inflow-mechanism limitation.

Waveform and settlement

LVP/RVP each have one significant peak. Peak position and PV-loop roundness are not forced into a universal normal interval. This does not rule out small fluctuations or ringing under every control combination.

LVP: 1 peaks · variation 1.217 · peak 71.53% ET.
RVP: 1 peaks · variation 1.237.

Three consecutive periodic-boundary passes were followed by a separate audit beat. 156 cycles · 0.0009783.

An additional 60 beats from the earlier saved state changed coronary tone 0.693→0.6484. CI changed 0.0004584 L/min/m² and mean Ao-node pressure -0.01326 mmHg. Small hemodynamic changes do not establish complete stationarity.

This extension was run on lab-003 and checked against the then-current lab-004 numerical state, solver predictor, construction and initial observations. It is not this model's launch record. It is retained as historical supporting evidence for the same construction, not a fresh execution.

PV analysis and passive accommodation

These loading analyses use a historical 2 ms launch with the same construction, not a rerun from this model's launch. Selecting the 1 ms record does not change the grid of these analyses. Pressure is transmural.

LV · J/beatbaselineHFrEF
SW1.0800.627
PE0.3820.907
PVA1.4621.534

SW falls while PE contributes more. Model load-family ESPVR/EDPVR are not catheter Ees or an isolated passive material law. PE includes a low-volume tangent extension. 2.56 mL.

A separate fully passive, viscously relaxed comparison at fixed RV volume 140 mL found 150.8→184.3 mL LV volume at 10 mmHg LV transmural pressure. It is neither RV-pressure-matched nor a dynamic EDPVR.

The PVA-derived oxygen estimate uses current LV mass, but remains illustrative: a canine relation and an uncalibrated contractility-dependent intercept do not validate disease metabolism, ischemia or mechanical efficiency.

Controls and qualification scope

The 53 controls are inherited, not a certification of all combinations. Single edits to LV tension 0.25/0.75/1.33, TBV4200/7000 mL and HR60 passed interaction, continuation and checkpoint tests. The altered states need not retain HFrEF targets.

Known limit: TBV7000 mL cold initialization failed at 2 ms. Changing from a saved state and independent 1 ms cold initialization succeeded. No automatic time-step fallback is implemented. Standard case launch uses its qualified 2 ms saved state.

Case assessment

Rest comparisons and launch states were calculated in this exact model. PVA, passive curves, control diagnostics and slow-tail observations refer to prior experiments with the same construction, retaining their original model identity and provenance.

1 ms · independent fine grid

Both grids meet required conditions and case preferences. This qualifies a selected operating point, not a diagnosis or clinical validation. Population statistics and model selection intervals are distinct.

Saved baseline and case comparison
MetricComparator baseline2 ms, fixedThis case1 msUnit
LVEF55.828.5%
LV EDVI75.6113.4mL/m²
LV ESVI33.581.1mL/m²
CI2.952.26L/min/m²
SVI42.232.3mL/m²
Ao node mean93.681.1mmHg
mean LAP8.416.6mmHg
mean RAP3.14.3mmHg
PA node mean17.923.9mmHg
RVEF56.949.1%
RV EDVI74.165.8mL/m²
RV ESVI31.933.5mL/m²
LV end-filling P10.925.2mmHg
LV end-filling Ptm10.922.7mmHg
LV +dP/dt2,5641,126mmHg/s
LV −dP/dt-1,539-828mmHg/s
ICT89.148.0ms
ET258.0279.0ms
IRT92.0130.0ms
Tei0.700.64
E/A (flow)0.940.86
τ Weiss32.050.8ms
τ Glantz52.4ms

The comparator baseline remains its saved 2 ms record; selecting the 1 ms case does not recompute it. Unrounded values are included in the JSON export.

This is not a load-matched comparison. Reduced EF can coexist with forward output; a smaller Tei than baseline does not demonstrate better contractility. Mean LA, end-filling cavity and transmural pressures differ. E/A uses volumetric flow, not Doppler velocity.

Intervals and rationale

LVEF · Required28.48 / 20–40 % · Met

Reduced EF defines the chosen phenotype. The 20% lower bound selects severity; it is not a diagnostic lower limit.

Method: native-valve-closure-volumes

LV EDVI · Required113.4 / 90–180 mL/m² · Met

Selects a chronic LV-dilated volume range, not a requirement for all HFrEF or an imaging diagnostic boundary.

Method: native-valve-closure-volumes

CI · Required2.261 / 1.8–3.5 L/min/m² · Met

Allows lower or retained output; reduced EF is not equated with extreme low output.

Method: exact-beat-aortic-valve-net-flow

mean LAP · Required16.6 / 3–25 mmHg · Met

Retains the broad screen against extreme filling. Elevated pressure is a case preference, not a required condition.

Method: exact-beat-mean-left-atrial-pressure

mean RAP · Required4.253 / 0–12 mmHg · Met

Selects loading for an LV-dominant case; it proves neither normal RVEF nor absence of RV involvement.

Method: exact-beat-mean-right-atrial-pressure

Ao node mean · Required81.13 / 60–110 mmHg · Met
LVEF · Case preference28.48 / 25–35 % · Met

Meet EF jointly with the other primary preferences, without ranking EF above all other objectives.

Method: native-valve-closure-volumes

LV EDVI · Case preference113.4 / 110–150 mL/m² · Met

Selects dilation and a large residual volume. ESVI is not added as a duplicate objective.

Method: native-valve-closure-volumes

CI · Case preference2.261 / 2.2–2.8 L/min/m² · Met

Selects one case with retained resting net output, not a disease-specific normal CI interval.

Method: exact-beat-aortic-valve-net-flow

Ao node mean · Case preference81.13 / 80–100 mmHg · Met

A resting arterial-pressure preference, strongly dependent on output and vascular load. Missing it does not invalidate HFrEF; the lower bound was not lowered to admit a known 77-mmHg candidate.

Method: exact-beat-mean-aortic-node-pressure

mean LAP · Case preference16.6 / 12–20 mmHg · Met

Intentionally selects higher filling than the mean PAWP of 10–12 mmHg in treated cohorts. Lavine supports the existence of such cases, not a frequency weight. Pressure alone does not diagnose edema or clinical congestion; native LV end-filling cavity and transmural pressures are also reported. Higher is not rewarded beyond the interval.

Method: exact-beat-mean-left-atrial-pressure

τ Weiss · Case preference50.81 / 40–75 ms · Met

A broad design interval informed by 54±14 ms, not a normal/abnormal boundary. Use only quality-qualified Weiss fits and retain the window and baseline comparison. With calcium/crossbridge rates fixed, tau changes can arise from geometry, tension, load and regression window; intrinsic relaxation change is not identified. No substitution by Glantz or extra reward for longer tau.

Method: main-wire-lv-relaxation-tau-v1:Weiss-zero-asymptote-time-weighted-log-linear

Relaxation, pressure rates and filling
P(t)=P+Ae(tt0)/τ,Tei=ICT+IRTETP(t)=P_{\infty}+A e^{-(t-t_0)/\tau},\qquad \mathrm{Tei}=\frac{ICT+IRT}{ET}

Tau is fitted to cavity-pressure decay. Weiss fixes the asymptote at zero; Glantz estimates it. Glantz does not pass fit quality here and is shown as unavailable. Weiss is prolonged despite unchanged calcium/crossbridge kinetics; effects of loading, geometry and the observation window are not identified intrinsic relaxation disease.

Weiss: 50.81 ms · R² 0.9986 · window 25.78 ms · 28 samples.

The window runs from the midpoint of the minimum-dP/dt interval to the next end-filling LV pressure plus 5 mmHg, before mitral opening. Time-weighted linear regressions use log-pressure versus time for Weiss and dP/dt versus pressure for Glantz. Required support is 6 samples, 15 ms and 10 mmHg decay; minimum R² is 0.97/0.95, and reconstructed-pressure RMSE must be ≤5% of the decay. These are measurement-usability design rules, not disease-normality limits.

Glantz R² describes the dP/dt-versus-pressure regression. A small reconstructed-pressure error does not override that fit-quality requirement.

dP/dt uses accepted-step cavity-pressure differences divided by elapsed time, not an analytic derivative or a filtered sensor. ICT/ET/IRT follow valve events/forward flow, not interchangeable clinical acquisition methods.

Flow E/A is 0.8582, A-dominant as in baseline (0.9379). Lavine's high-filling-pressure DCM group had a greater early filling fraction and a lower atrial fraction, a different direction. The flow-versus-Doppler distinction alone does not resolve this mismatch. A-dominance is a model inflow trait shared with baseline, not an established HFrEF finding reproduced here.

End-filling LV cavity pressure exceeds mean LA pressure by 8.65 mmHg. Kato's difference of cohort means was 4 mmHg, not a distribution of individual differences. Mean LA is not substituted for PCWP or end-filling LV pressure; this discrepancy remains an inflow-mechanism limitation.

Waveform and settlement

LVP/RVP each have one significant peak. Peak position and PV-loop roundness are not forced into a universal normal interval. This does not rule out small fluctuations or ringing under every control combination.

LVP: 1 peaks · variation 1.214 · peak 71.38% ET.
RVP: 1 peaks · variation 1.231.

Three consecutive periodic-boundary passes were followed by a separate audit beat. 155 cycles · 0.0009788.

An additional 60 beats from the earlier saved state changed coronary tone 0.6971→0.6525. CI changed 0.0004628 L/min/m² and mean Ao-node pressure -0.01312 mmHg. Small hemodynamic changes do not establish complete stationarity.

This extension was run on lab-003 and checked against the then-current lab-004 numerical state, solver predictor, construction and initial observations. It is not this model's launch record. It is retained as historical supporting evidence for the same construction, not a fresh execution.

PV analysis and passive accommodation

These loading analyses use a historical 2 ms launch with the same construction, not a rerun from this model's launch. Selecting the 1 ms record does not change the grid of these analyses. Pressure is transmural.

LV · J/beatbaselineHFrEF
SW1.0800.627
PE0.3820.907
PVA1.4621.534

SW falls while PE contributes more. Model load-family ESPVR/EDPVR are not catheter Ees or an isolated passive material law. PE includes a low-volume tangent extension. 2.56 mL.

A separate fully passive, viscously relaxed comparison at fixed RV volume 140 mL found 150.8→184.3 mL LV volume at 10 mmHg LV transmural pressure. It is neither RV-pressure-matched nor a dynamic EDPVR.

The PVA-derived oxygen estimate uses current LV mass, but remains illustrative: a canine relation and an uncalibrated contractility-dependent intercept do not validate disease metabolism, ischemia or mechanical efficiency.

Controls and qualification scope

The 53 controls are inherited, not a certification of all combinations. Single edits to LV tension 0.25/0.75/1.33, TBV4200/7000 mL and HR60 passed interaction, continuation and checkpoint tests. The altered states need not retain HFrEF targets.

Known limit: TBV7000 mL cold initialization failed at 2 ms. Changing from a saved state and independent 1 ms cold initialization succeeded. No automatic time-step fallback is implemented. Standard case launch uses its qualified 2 ms saved state.

Evidence and reproducibility

Source populations, methods and limitations
2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure

EF classification in the 2022 heart-failure guideline.

The guideline's EF-based classification is used as a construction convention.

Neither a claim about the latest definition nor a clinical diagnosis from low EF alone.

Cardiac output and cardiac index measured with cardiovascular magnetic resonance in healthy subjects, elite athletes and patients with congestive heart failure

157 patients with CHF and EF ≤40%, age 60±13 years. Moderate or greater MR, AR and aortic velocity >3 m/s were among the exclusions.

Supine CMR; whole-cycle phase-contrast flow in the ascending aorta at the pulmonary-artery bifurcation.

CI and EF were weakly related. Flow measured beyond the coronary origins is not identical to model AV net output. Low CI is not a necessary HFrEF finding.

Beneficial Effects of Heart Rate Reduction on Cardiac Mechanics and Energetics in Patients With Left Ventricular Dysfunction

14 patients with LV dysfunction: 10 with previous infarction and four with DCM. EF 34±12%, range 23–53%; not all had EF ≤40%.

Control HR 83±10; conductance volume and Millar cavity pressure. Linear ESPVR during IVC occlusion; loop area for work; Raff–Glantz relaxation.

Diuretics and vasodilators were withheld for at least 24 hours. EDP differs between Tables 1 and 2; Ees indexing units also differ between text and table. These values cannot directly define model thresholds or validate transmural work and a nonlinear common-isochrone ESPVR.

Comparison of haemodynamic response to muscle reflex in heart failure with reduced vs. preserved ejection fraction

10 treated HFrEF patients with improved symptoms; EF 23±6%, HR 72±17. PAWP ≥25 mmHg, resting SBP ≥160 mmHg, AF, severe valve disease and recent AMI were excluded.

Right-heart catheterization and 333-Hz microchip/conductance PV, calibrated to MRI volume. Senzaki single-beat Ees; MAP estimated from LV end-systolic and estimated diastolic pressures.

Small, selected, treated cohort. Do not index individual volumes using the cohort's mean BSA. PAWP is not measured mean LA pressure. Nonzero-asymptote tau is not Weiss tau and requires method checking before comparison with Glantz.

Association between systolic ejection time and outcomes in heart failure by ejection fraction

171 ambulatory HFrEF patients; median EF 30%, interquartile range 25–35%.

Three echocardiographic measurements. SET from LVOT Doppler; PEP includes ECG timing; relaxation calculated as R-to-E minus PEP and SET.

PEP is not ICT; Doppler endpoints differ from flow-zero crossings. HR dependence prevents adopting the interquartile interval as an HR70 gate. Table 2 gives HFrEF relaxation 93.3 [67.3, 122] ms, while the abstract reverses the group medians. That inconsistent statistic is contextual only, not a gate or objective.

Diastolic Determinants of Excess Mortality in Heart Failure With Reduced Ejection Fraction

12,421 patients with Stage B/C LV dysfunction, including EF <50%; not an exclusively EF ≤40% cohort.

Doppler diastolic measures; filling patterns differ between E/e′ groups.

Supports not requiring restrictive filling in every HFrEF case. The 0D model does not measure e′, so E/e′ must not be invented.

Correlation between left ventricular contractility and relaxation in patients with idiopathic dilated cardiomyopathy

38 resting idiopathic DCM patients in sinus rhythm, NYHA II–III, and nine controls. Stabilized by treatment, then medications withheld for at least three days. Observed EF 17–64%; not all meet modern HFrEF classification.

High-fidelity LV pressure at 3-ms intervals. Zero-asymptote Weiss and Raff–Glantz tau from minimum dP/dt to EDP+5 mmHg. Conductance volume and four to eight IVC-occlusion beats for Ees.

Small selected cohort. Acquisition differs from time-weighted accepted-step regression and valve-event measurements. Ees and tau were not correlated; some patients retained relaxation. Inconsistent assignment of preserved TL/TD proportions in the text prevents using a finding frequency.

Prognostic value of myocardial strain and late gadolinium enhancement on cardiovascular magnetic resonance imaging in patients with idiopathic dilated cardiomyopathy with moderate to severely reduced ejection fraction

172 idiopathic DCM patients with EF <40%; HR 83±20 and MAP 84±13 mmHg.

CMR and contemporaneous clinical measurements. LV-volume-derived and RV-volume-derived output differ.

Clinical MAP and exact Ao-node cycle mean differ in site and method. Do not transfer LV-volume-derived CI to net AV flow or derive net CI from ratios of cohort means.

Functional and structural abnormalities in patients with dilated cardiomyopathy

12 DCM patients and 10 controls. Seven DCM patients had normal material stiffness and five had increased stiffness; group mean EF 37/36% and LVEDP 18/22 mmHg.

Simultaneous ventriculography and high-fidelity pressure, viscoelastic pressure–volume/stress–strain analysis, and right-heart biopsy.

Chamber pressure–volume behavior differs from tissue stiffness. Small-group abstract statistics do not directly specify model material coefficients or required intervals.

Relation between ventriculoarterial coupling and myocardial energetics in patients with idiopathic dilated cardiomyopathy

Invasive PV in 23 idiopathic DCM patients; simultaneous coronary-sinus oxygen consumption in 16.

Conductance PV, IVC-occlusion Ees and a double-thermistor coronary-sinus catheter.

Statistics are mean±SE. Inferring Ees/Ea from EF does not provide independent evidence. SW/PVA and SW/MVO2 differ; transmural model work alone establishes neither metabolic efficiency nor increased PVA. Institution and authors overlap with Kato.

Importance of the left ventricular filling pressure on diastolic filling in idiopathic dilated cardiomyopathy

33 controls; 14 DCM patients with normal and 26 with elevated filling pressure, classified by mean pulmonary capillary pressure.

Filling pressure and pulsed transmitral Doppler; mitral-regurgitation effects also examined.

Filling pressure and MR affect inflow differently. The 15-mmHg group boundary is not an identical threshold for mean LA or LVEDP. Doppler velocity and total volumetric valve flow must remain distinct.

Comparison of Doppler indexes of left ventricular diastolic function with simultaneous high fidelity left atrial and ventricular pressures in idiopathic dilated cardiomyopathy

Idiopathic DCM at baseline and during amrinone. Sample size was not stated in the available abstract.

Simultaneous LA/LV micromanometer pressures, transmitral Doppler and M-mode/2D echocardiography.

E/A can fall despite improved relaxation. Intervention mechanisms were not individually isolated. This is not validation of a numerical target or drug model.

Characteristics and prognostic value of right ventricular (dys)function in patients with non-ischaemic dilated cardiomyopathy assessed with cardiac magnetic resonance imaging

216 nonischemic DCM patients with CMR EF <50%; median EF 37% [25–44]. RV dysfunction in 38% and MR in 63%, mostly mild.

Single-center retrospective CMR cohort; preexisting RV disease and pulmonary hypertension excluded.

Includes EF >40%, treatment, AF and MR. The 38% finding is not a universal HFrEF frequency or occurrence probability. LV/RV association does not determine causal direction.

Transcription correction: Ishihara 1994 mean Ea/Ees was corrected from 3.14 to 3.24 against the primary abstract. The original record is preserved. This contextual correction changes neither selection intervals, assessment results nor model values.

Identity, export and adoption status

This saved snapshot contains the assessed values, criteria and provenance and survives retirement of authoring code. Generating it or meeting intervals is not an adoption approval.

Model
circleheart.main-wire-integrated-transaction-v3.static-anatomy.standard-73
Surface
circleheart.main-wire.surface.static-anatomy.standard-73.workbench-v1
Case
standard73-hfref-chronic-dilated-v1
sourceSha256
c42db7ce5ea71fa6fc3ea301700af1b25e0ff4c673aba120bd2b9598ee2d79cf
evidenceSha256
38bd2a289ae86f2d13bcfed72a70827e4ce55ab52056c473ec458acaaaf078df
referenceSha256
45f553dffa52e0cdc25f51f065762474d41fd692692852f079bd8e1e648bd00a
artifactSha256
f40388e5918e968ed9db093d7c9cd72467da332212c575d08164958e206f5124
artifactRevisionId
38e31e94e7b25e71d0bb99c1a7e9a27dc094985c3920d12c2fdf22afa34abce4
checkpointSha256
c97b528ac709354762a2c19f5cf4f52b32b22aa746542146cc8c99e39f102cf3

Formal registration still requires at least one unconditional approval from the two reviewers and a new exact release's own artifact, captures and Surface binding. Existing Standard72 checkpoints will not be relabeled.

  • analysis/main-wire-integrated-v3-formal-fixed-tbv-pressure-volume-relations-v1
  • analysis/main-wire-integrated-v3-guyton-starling-structural-orientation-v1
  • derivation/suga-pva-measured-load-display-exact-anatomy-mvo2-v14
Save inputs, assessment and evidence (JSON)