Left Ventricular Late Gadolinium Enhancement Assessment for MR

Purpose

Automate late gadolinium enhancement assessment

Tag(s)

 

Panel

 Cardiac

TOUCH-AI ID

 TAI-04180008

Originator

 Akos Szemes

Panel Chair

 Carlo De Cecco

Panel Reviewers

 Cardiac Panel

License

 Creative Commons 4.0

Status  Public Commenting
                               

Clinical Implementation


Value Proposition

Late gadolinium enhancement (LGE) can be observed in several myocardial pathologies, including different stages of myocardial infarction (MI), myocarditis, and certain cardiomyopathies. The clinical assessment of LGE is primarily qualitative, with descriptions of the location and number of affected segments, or semiquantitative, such as LGE transmurality. Although latter techniques allow for fast interpretation, the accurate quantification of MI has become necessary to systematically analyze the association of MI size with short- and long-term effects, associated complications, and general prognoses. It has been shown that the size of MI correlates with ejection fraction, predicts segmental functional outcome after revascularization, prognosticates recovery of left ventricular function, and shows association with the occurrence of spontaneous ventricular arrhythmias. Automated LGE quantification would save time and eliminate subjectivity from the analysis.

Narrative(s)

All patients undergoing contrast-enhanced cardiovascular magnetic resonance imaging (CMR) examinations should have automated quantification of LGE if data are available, which should then be automatically populated into the radiology report or a report supplement.

Workflow Description

A patient receives contrast-enhanced MRI of the heart, with LGE acquisition included. An algorithm retrieves input from the imaging data set, relevant EMR data (including age, sex, and body surface area [BSA; m2]), and user-selected quantification methods (standard deviation [SD] based: 2 or 5 SD and full width at half maximum). The algorithm automatically segments the borders of the LV endocardium and LV epicardium on LGE data sets and identifies the base and apex of the heart; the LV myocardium (the area between the epicardium and endocardium) automatically segmented to identify normal and enhanced myocardium based on the quantification method selected. The algorithm then detects affected myocardium segments based on the 17-segment American Heart Association model and determines transmurality in these segments (<25%, 25-50%, 51-75%, and >75%). With this information, the algorithm calculates absolute myocardial mass (LVM; mL and g), total LGE volume (mL), LGE percentage (%LVM), segmental LGE volume (mL), LGE percentage (%), and BSA-indexed LVM (g/m2). The measurements are returned and automatically populated into the report. Segmental assessment is visualized on a color-coded polar map also indicating coronary artery territories. The algorithm flags abnormal LVM values.

Considerations for Dataset Development


Procedures(s): Contrast enhanced cardiac MRI with LGE acquisition

View(s): {short-axis LGE images, long-axis 4-chamber LGE images, long-axis 2-chamber LGE images, long-axis left ventricular outflow tract LGE images}

 

Sex at Birth: {Male, Female}

 

Age: [0,90]

 

Body Surface Area: varied

 

Comorbidities: {Congenital heart disease, ventricular septal defects}

 

Intervention: {Congenital heart disease repair, mitral valve replacements, coronary artery bypass graft, alcohol ablation in hypertrophic cardiomyopathy, etc}

 

Contrast: contrast-enhanced MRI

Technical Specifications


Inputs

DICOM Study

Procedure

Contrast-enhanced cardiac MRI with LGE acquisition

Views

CMR: short-axis LGE images, long-axis 4-chamber LGE images, long-axis 2-chamber LGE images, long-axis LVOT LGE images

Data Type

DICOM

Modality

CMR

Body Region

Chest

Anatomic Focus

Heart

Primary Outputs

 

Number of segments with LGE

RadElement ID

RDE225

Definition

Number of segments with LGE based on the AHA model

Data Type

Numeric

Value Set

 

Units

 

 

List of Segments with LGE

RadElement ID

RDE226

Definition

List of segments with LGE (displayed on polar map)

Data Type

Segmentation value according to AHA segmentation

Value Set

1-basal anterior

2-basal anteroseptal

3-basal inferoseptal

4-basal inferior

5-basal inferolateral

6-basal anterolateral

7-mid anterior

8-mid anteroseptal

9-mid inferoseptal

10-mid inferior

11-mid inferolateral

12-mid anterolateral

13-apical anterior

14-apical septal

15-apical inferior

16-apical lateral

17-apex

Units

N/A

 

Transmurality of LGE

RadElement ID

RDE227

Definition

Transmurality of LGE in each segment (displayed on polar map)

Data Type

Numeric

Value Set

 

Units

Percentage

 

Total LGE Volume

RadElement ID

RDE228

Definition

Volume of LGE in the LV (mL)

Data Type

Numeric

Value Set

 

Units

mL

 

Total Percentage of LGE

RadElement ID

RDE229

Definition

Total percentage of LGE in the LV (%)

Data Type

Numeric

Value Set

[0,100]

Units

Percentage

 

Segmental Volume of LGE

RadElement ID

RDE230

Definition

Volume of LGE based on the AHA model (displayed on polar map)

Data Type

Numeric

Value Set

 

Units

mL

 

Segmental Percentage of LGE

RadElement ID

RDE231

Definition

Percentage of LGE based on the AHA model (displayed on polar map)

Data Type

Numeric

Value Set

[0,100]

Units

Percentage

Secondary Outputs

Myocardial Mass (mL)

RadElement ID

RDE232

Definition

Myocardial mass (ml)

Data Type

Numeric

Value Set

 

Units

mL

 

Myocardial Mass (g)

RadElement ID

RDE233

Definition

Myocardial mass (ml)

Data Type

Numeric

Value Set

 

Units

mL

 

Indexed Myocardial Mass

RadElement ID

RDE234

Definition

Myocardial Mass Indexed to body surface area (g/m2)

Data Type

Numeric

Value Set

 

Units

mL

Public Commenting


Use cases are meant to be a primary vehicle for distributing clinical information to the developer community. They pinpoint precise scenarios within radiology workflows where potential automation could add noticeable value and establish standards for interpreting and passing corresponding common data elements. Implementing effective standards requires the perspective from all stakeholders. So to that end, we encourage your feedback on use cases.

To submit comments, please email DSIUseCases@acr.org with the use case title(s) and relevant comments by January 1, 2019. If more convenient, you may also download this use case and comment directly on the PDF. Just attach the PDF copy on the email.