CardioNexus: Mapping Every Heartbeat with CMR and ECG.

Valvular and flow phenotypes combine LVOT cine geometry, 4-chamber annular measurements, and phase-contrast aortic flow. This page is source-audited for the current CardiacNexus output names, but it treats valve disease labels as interpretation context only; no row is a stand-alone aortic stenosis, aortic regurgitation, mitral disease, or tricuspid disease diagnosis.

Modality: LVOT cine CMR, cine long-axis CMR, and phase-contrast flow CMR
UKB source: Data Fields 20212, 20208, and 20213
Pipeline step: LVOT aortic-root diameter extraction, combined valve landmark extraction, phase-contrast velocity and flow analysis
Outputs: LVOT.csv, ventricular_atrial_feature.csv, aortic_flow.csv, timeseries/aorta.npz, LVOT and flow QC plots
Maturity: Source-audited phenotype page

Clinical question

This page answers three separate questions: how large the aortic root is, how large the mitral and tricuspid annular spans are in 4-chamber cine, and how blood crosses the aortic phase-contrast plane. These are related to valve disease, aortopathy, and flow remodeling, but their acquisition planes and formulas differ enough that they should not be collapsed into one valve label.

Anatomical and physiological definition

The LVOT cine route measures aortic valve annulus, sinuses of Valsalva, and sinotubular junction diameters. The combined long-axis route measures mitral and tricuspid annular minimum diameters on 4-chamber segmentation at ventricular ED and ES.

Phase-contrast flow converts phase images into through-plane velocity, integrates velocity over the segmented aortic lumen, and summarizes area, velocity, gradient, forward/backward flow, valve area, VTI, systolic flow reversal, and flow displacement. Mean gradient is derived from velocity with a simplified Bernoulli relationship, while aortic valve area uses forward flow divided by VTI [2] [3].

Source acquisition and UKB field

LVOT cine measurements use UK Biobank Data Field 20212. Mitral and tricuspid annular measurements use the long-axis 4-chamber segmentation from Data Field 20208 plus ventricular timing from the short-axis timeseries. Aortic phase-contrast flow uses Data Field 20213 and relies on VENC metadata from the DICOM CSA header.

What exactly CardiacNexus measures

LVOT aortic-root diameters

eval_LVOT.py segments the LVOT cine frame stack and computes distances between landmarks for the aortic valve annulus, aortic sinuses, and sinotubular junction. The emitted rows are median diameters across accepted frames, with BSA-indexed versions when BSA is available.

Mitral and tricuspid annular diameters

eval_ventricular_atrial_feature.py calls evaluate_valve_length on the 4-chamber segmentation at ED and ES. It writes tricuspid and mitral annular minimum diameters at ED and ES, and writes the ED TA/MA ratio only when the ratio is not implausibly high.

Phase-contrast aortic flow

eval_phase_contrast.py computes aortic area from the phase-contrast segmentation, converts phase to velocity using VENC, integrates flow, estimates peak systole and end systole from the flow curve, and writes velocity, gradient, flow-volume, VTI, valve-area, flow-displacement, and systolic reversal summaries.

Regurgitant fraction schematic from the CardiacNexus manuscript image cache citing Gomes et al. 2022 — Regurgitant-fraction schematic from the manuscript image cache, cited there to Gomes et al. 2022, shown to orient forward flow, regurgitant volume, and net flow concepts. CardiacNexus output formulas, QC skips, and method boundaries are defined in the page text and tables.Source: manuscript image cache citing Gomes et al. 2022; permission and exact source-panel review pending for draft use.

Aortic valve area measurement context from the manuscript image cache — Aortic valve area context figure from the manuscript image cache, shown to orient AVA method language. CardiacNexus reports an aortic-flow-derived AVA row; guideline severity thresholds remain contextual and are not automated disease labels.Source: manuscript image cache; permission and exact source-panel review pending for draft use.

Flow displacement schematic from the manuscript image cache — Flow-displacement schematic from the manuscript image cache, cited there to Sigovan et al. 2011, shown to orient velocity-weighted flow-center displacement. Plane prescription, VENC, segmentation leakage, and timing remain QC-dependent.Source: manuscript image cache citing Sigovan et al. 2011; permission and exact source-panel review pending for draft use.

Aortic valve area, VTI, and flow displacement

The current aortic valve area row is derived from forward flow divided by VTI. Flow displacement compares the velocity-weighted flow center with the lumen center and normalizes by radius. These rows are method-sensitive and should be interpreted with plane prescription, segmentation, and velocity-aliasing context.

Output columns and units

CardiacNexus writes this page's current output families into LVOT.csv, ventricular_atrial_feature.csv, and aortic_flow.csv.

LVOT and annular structure

Display family	Exact output column	Unit	Status	Schema note
LVOT diameter	`LVOT: Aortic Valve Annulus Diameter [mm]`	mm	current	median accepted-frame distance
LVOT diameter	`LVOT: Aortic Sinuses Diameter [mm]`	mm	current	median accepted-frame distance
LVOT diameter	`LVOT: Sinotubular Junction Diameter [mm]`	mm	current	median accepted-frame distance
LVOT indexed diameter	`LVOT: Aortic Valve Annulus Diameter/BSA [mm/m^2]`	mm/m^2	conditional current output	requires BSA
LVOT indexed diameter	`LVOT: Aortic Sinuses Diameter/BSA [mm/m^2]`	mm/m^2	conditional current output	requires BSA
LVOT indexed diameter	`LVOT: Sinotubular Junction Diameter/BSA [mm/m^2]`	mm/m^2	conditional current output	requires BSA
Annular diameter	`Valve: Tricuspid_diameter_ED [cm]`	cm	current	4-chamber ED segmentation
Annular diameter	`Valve: Mitral_diameter_ED [cm]`	cm	current	4-chamber ED segmentation
Annular ratio	`Valve: TA/MA Ratio_ED`	unitless	conditional current output	skipped if implausibly high
Annular diameter	`Valve: Tricuspid_diameter_ES [cm]`	cm	current	4-chamber ES segmentation
Annular diameter	`Valve: Mitral_diameter_ES [cm]`	cm	current	4-chamber ES segmentation

Phase-contrast aortic flow

Display family	Exact output column	Unit	Status	Schema note
Aortic area	`Aortic Flow: Maximum Area [mm^2]`	mm^2	current	maximum segmented lumen area
Aortic area	`Aortic Flow: Minimum Area [mm^2]`	mm^2	current	minimum segmented lumen area
Aortic timing	`Aortic Flow: T_ES [frame]`	frame	current	end-systolic frame from flow curve
Aortic timing	`Aortic Flow: T_peak_systole [frame]`	frame	current	peak-systolic frame
Aortic timing	`Aortic Flow: T_ES [ms]`	ms	current	end-systolic time
Aortic timing	`Aortic Flow: T_peak_systole [ms]`	ms	current	peak-systolic time
Aortic velocity	`Aortic Flow: Peak Velocity [cm/s]`	cm/s	current	maximum velocity before ES
Aortic gradient	`Aortic Flow: Mean Gradient [mmHg]`	mmHg	current	simplified Bernoulli over systolic frames
Aortic flow	`Aortic Flow: Forward Flow [mL]`	mL	current	positive integrated flow
Aortic flow	`Aortic Flow: Backward Flow [mL]`	mL	current	reverse integrated flow
Aortic flow	`Aortic Flow: Regurgitant Fraction [%]`	%	current	backward/forward flow; >100% skipped
Aortic valve area	`Aortic Flow: Aortic Valve Area [cm^2]`	cm^2	conditional current output	skipped when VTI exceeds QC threshold
Aortic VTI	`Aortic Flow: Velocity Time Integral [cm]`	cm	conditional current output	written when VTI passes QC threshold
Flow displacement	`Aortic Flow: Flow Displacement Systolic Average [%]`	%	current	systolic velocity-center displacement
Flow displacement	`Aortic Flow: Flow Displacement Late Systolic Average [%]`	%	current	late-systolic displacement
Flow displacement	`Aortic Flow: Flow Displacement Diastolic Average [%]`	%	current	diastolic displacement
Systolic reversal	`Aortic Flow: Systolic Forward Flow [mL]`	mL	current	systolic positive flow
Systolic reversal	`Aortic Flow: Systolic Reverse Flow [mL]`	mL	current	systolic reverse flow
Systolic reversal	`Aortic Flow: Systolic Flow Reversal Ratio [%]`	%	conditional current output	skipped when >50%

Output reconciliation

Evidence layer	Result
Implementation source	LVOT, phase-contrast flow, and 4-chamber annular rows checked against `eval_LVOT.py`, `eval_phase_contrast.py`, and `eval_ventricular_atrial_feature.py`
Output inventory	`docs/data/output_column_inventory.yml` records `LVOT`, `aortic_flow`, and `ventricular_atrial_feature` artifact families
Phenotype dictionary	`docs/data/phenotype_dictionary.yml` links promoted valvular/flow rows to this page
Page output table	promoted dictionary-backed rows and broader phase-contrast output families are listed above

Required upstream inputs

LVOT cine image and segmentation for aortic-root diameters;
long-axis 4-chamber image and segmentation for annular diameters;
ventricular and atrial aggregate CSVs, plus timeseries/ventricle.npz and timeseries/atrium.npz, for combined 4-chamber outputs;
phase-contrast magnitude/phase images, aortic segmentation, and DICOM VENC metadata for aortic flow;
BSA table only for indexed LVOT diameter outputs.

Reference ranges with cohort and method context

Feature	Source	Cohort	Reference value	Status	Note
Aortic root diameters	CMR aortic-root reference context [1]	CMR reference cohorts	method-dependent	Verified context source	Useful context, but CardiacNexus uses automated LVOT segmentation and median accepted-frame distances
Phase-contrast flow method	Phase-contrast CMR review [2]	method review	not a normal range	Verified context source	Used for velocity/flow method context
Valvular function at UKB scale	Kany et al. [3]	UKB-scale flow cohort	source-specific values	Verified context source	Do not display as direct thresholds without table-row extraction and method adjudication
Dynamic aortic flow volumes	Gomes et al. [4]	UKB-scale flow/genetics cohort	source-specific values	Verified context source	Regurgitant-fraction context, not a diagnostic cutoff in this page
Flow displacement	Sigovan et al. and Zhao et al. [5] [6]	4D-flow / aortic-flow cohorts	method-dependent	Verified context source	Current CardiacNexus row is 2D phase-contrast-derived and needs method labels
Valve guideline thresholds	AHA/ACC valve guideline context [7]	clinical guideline	severe AS/AR thresholds are clinical-context dependent	Guideline context	Page cites only as clinical context, not as an automated classifier

Source-located registry status: reference_range_sources.yml maps aortic-root and valve-flow context to CMR root, phase-contrast, UKB flow, regurgitant fraction, and flow-displacement sources. Guideline thresholds are not promoted as CardiacNexus thresholds because plane placement, VENC, phase-offset behavior, valve morphology, and clinical report context are outside the automated output contract.

Disease interpretation

Aortic-root dilation can contribute to aortic regurgitation and aortopathy interpretation, including bicuspid-valve and connective-tissue disease contexts. Peak velocity, mean gradient, VTI, and valve area are clinically relevant to aortic stenosis, while forward/backward flow and regurgitant fraction are relevant to aortic regurgitation [3] [7].

Mitral and tricuspid annular rows describe 4-chamber geometry and are useful in remodeling context, especially when read with atrial and ventricular size. Flow displacement and systolic reversal rows are eccentric-flow and hemodynamic context signals, not stand-alone disease labels [5] [6].

QC caveats and maturity boundary

LVOT diameter outputs depend on aortic-root segmentation quality and accepted-frame selection. Annular diameters depend on 4-chamber segmentation, ED/ES timing, and landmark geometry. Phase-contrast outputs depend on VENC, plane prescription, phase-to-velocity conversion, segmentation of the aortic lumen, aliasing, background phase behavior, and reliable flow-curve timing.

The current implementation skips or warns on implausible regurgitant fraction, high VTI, very high systolic flow reversal ratio, and unstable annular ratios. These are cohort phenotyping outputs and require review of QC plots and acquisition context before clinical interpretation.

Implementation provenance

Feature family	Formula or computational route	Exact output columns	Source code file and function	Upstream dependencies	Conditional behavior	QC artifacts	Schema debt
LVOT aortic-root diameters	LVOT segmentation landmarks, median accepted-frame distances, optional BSA normalization	`LVOT:* Diameter [mm]`, `LVOT:* Diameter/BSA [mm/m^2]`	`eval_LVOT.py`	`lvot.nii.gz`, `seg_lvot.nii.gz`, BSA table for indexed rows	indexed rows require BSA; frames can fail LVOT QC	`visualization/aorta/LVOT_*.png`	artifact key is uppercase `LVOT` in the aggregate registry
Annular diameters	4-chamber ED/ES valve-length helper on segmentation-derived landmarks	`Valve: Tricuspid_diameter_ED [cm]`, `Valve: Mitral_diameter_ED [cm]`, `Valve: TA/MA Ratio_ED`, `Valve: Tricuspid_diameter_ES [cm]`, `Valve: Mitral_diameter_ES [cm]`	`eval_ventricular_atrial_feature.py`; `evaluate_valve_length`	`la_4ch.nii.gz`, `seg4_la_4ch.nii.gz`, `ventricle.npz`, ventricular and atrial aggregate CSVs	TA/MA ratio skipped when implausibly high; subject skipped if upstream aggregates/timeseries missing	`visualization/combined/valve_ED.png`, `visualization/combined/valve_ES.png`	combined artifact also emits AV/IPVT rows that belong mainly to cross-chamber interpretation
Aortic velocity and flow	phase image to velocity using VENC, lumen integration, peak/ES timing from flow curve	velocity, gradient, forward/backward flow, regurgitant fraction, timing rows	`eval_phase_contrast.py`	phase-contrast images, segmentation, DICOM CSA VENC metadata	subject skipped if VENC/inputs fail; regurgitant fraction >100% skipped	phase overlays, velocity and flow time-series plots, `timeseries/aorta.npz`	plane and velocity correction context is not encoded in the output labels
Aortic valve area and VTI	systolic VTI from peak-velocity curve; valve area from forward flow divided by VTI	`Aortic Flow: Aortic Valve Area [cm^2]`, `Aortic Flow: Velocity Time Integral [cm]`	`eval_phase_contrast.py`	same phase-contrast flow inputs	VTI rows skipped if VTI exceeds QC threshold	velocity/VTI plots	continuity-equation interpretation requires method context
Flow displacement and reversal	velocity-center displacement normalized by lumen radius; systolic forward/reverse flow summaries	`Aortic Flow: Flow Displacement`, `Aortic Flow: Systolic` rows	`eval_phase_contrast.py`	segmented lumen and phase-derived velocity field	systolic reversal ratio skipped when >50%	flow-displacement and rotation-angle plots	displacement rows are sensitive to noisy contours and 2D plane choice

Source audit

Current LVOT output labels and BSA conditionals were checked against eval_LVOT.py.
Current phase-contrast area, timing, velocity, gradient, flow, valve-area, VTI, flow-displacement, and systolic reversal labels were checked against eval_phase_contrast.py.
Current annular diameter and TA/MA ratio labels were checked against eval_ventricular_atrial_feature.py.
docs/data/reference_sources.yml is present and used as the curated reference-source registry for this page.
Textbook context boundary: page-specific code, method literature, and guideline context are sufficient for draft rollout; Braunwald/Hurst background was not needed for the phenotype-specific output contract.
One regurgitant-fraction schematic is displayed from a page-local public path and registered in docs/data/figure_provenance.yml; permission and exact source-panel review remain pending for draft use.

References

Kawel-Boehm N, Hetzel SJ, Ambale-Venkatesh B, Captur G, Francois CJ, Jerosch-Herold M, Salerno M, Teague SD, Valsangiacomo-Buechel ER, Van Der Geest RJ, Bluemke DA. Reference ranges for cardiovascular magnetic resonance in adults and children: 2020 update. Journal of Cardiovascular Magnetic Resonance. 2020;22(1):87.
Nayak KS, Nielsen JF, Bernstein MA, Markl M, Gatehouse PD, Botnar RM, Saloner D, Lorenz C, Wen H, Hu BS, Epstein FH, Oshinski JN, Raman SV. Cardiovascular magnetic resonance phase contrast imaging. Journal of Cardiovascular Magnetic Resonance. 2015;17(1):71.
Kany S, Ramo JT, Hou C, Jurgens SJ, Nauffal V, Cunningham J, Lau ES, Butte AJ, Ho JE, Olgin JE, Elmariah S, Lindsay ME, Ellinor PT, Pirruccello JP. Assessment of valvular function in over 47,000 people using deep learning-based flow measurements. Cardiovascular Medicine preprint. 2023.
Gomes B, Singh A, O'Sullivan JW, Amar D, Kostur M, Haddad F, Salerno M, Parikh VN, Meder B, Ashley EA. Genetic architecture of cardiac dynamic flow volumes. Cardiovascular Medicine preprint. 2022. doi:10.1101/2022.10.05.22280733.
Sigovan M, Hope MD, Dyverfeldt P, Saloner D. Comparison of four-dimensional flow parameters for quantification of flow eccentricity in the ascending aorta. Journal of Magnetic Resonance Imaging. 2011;34(5):1226-1230.
Zhao X, Garg P, Assadi H, Tan RS, Chai P, Yeo TJ, Matthews G, Mehmood Z, Leng S, Bryant JA, Teo LLS, Ong CC, Yip JW, Tan JL, van der Geest RJ, Zhong L. Aortic flow is associated with aging and exercise capacity. European Heart Journal Open. 2023;3(4):oead079. doi:10.1093/ehjopen/oead079. PMID:37635784; PMCID:PMC10460199.
Nishimura RA, Otto CM, Bonow RO, Carabello BA, Erwin JP, Guyton RA, O'Gara PT, Ruiz CE, Skubas NJ, Sorajja P, Sundt TM, Thomas JD. 2014 AHA/ACC guideline for the management of patients with valvular heart disease. Journal of the American College of Cardiology. 2014;63(22):e57-e185.

Valvular and flow phenotypes