Detection of hypoplastic left heart syndrome anatomy from cardiovascular magnetic resonance images using machine learning

OBJECTIVE

The prospect of being able to gain relevant information from cardiovascular magnetic resonance (CMR) image analysis automatically opens up new potential to assist the evaluating physician. For machine-learning-based classification of complex congenital heart disease, only few studies have used CMR.

MATERIALS AND METHODS

This study presents a tailor-made neural network architecture for detection of 7 distinctive anatomic landmarks in CMR images of patients with hypoplastic left heart syndrome (HLHS) in Fontan circulation or healthy controls and demonstrates the potential of the spatial arrangement of the landmarks to identify HLHS. The method was applied to the axial SSFP CMR scans of 46 patients with HLHS and 33 healthy controls.

RESULTS

The displacement between predicted and annotated landmark had a standard deviation of 8-17 mm and was larger than the interobserver variability by a factor of 1.1-2.0. A high overall classification accuracy of 98.7% was achieved.

DISCUSSION

Decoupling the identification of clinically meaningful anatomic landmarks from the actual classification improved transparency of classification results. Information from such automated analysis could be used to quickly jump to anatomic positions and guide the physician more efficiently through the analysis depending on the detected condition, which may ultimately improve work flow and save analysis time.

Reference Values for Pediatric Atrial Volumes Assessed by Steady-State Free-Precession Magnetic Resonance Imaging Using Monoplane and Biplane Area-Length Methods

BACKGROUND

Measurement of atrial volumes by MRI is becoming increasingly important in pediatric cardiac disorders. However, MRI normal values for atrial volumes in children are lacking.

PURPOSE

To establish pediatric reference values for atrial volumes.

STUDY TYPE

Retrospective.

SUBJECTS

A total of 155 healthy children from two large institutions (103 male, age 13.9 ± 2.8 years, range 4-18 years).

FIELD STRENGTH/SEQUENCE

A 1.5 T; balanced steady-state free precession (bSSFP) sequence.

ASSESSMENT

The monoplane and biplane area-length methods were used to measure minimal and maximal left and right atrial volumes (LAmin , LAmax , RAmin , and RAmax ) from four-chamber (4ch) and two-chamber (2ch) MR cine images. Centile charts and tables for atrial volumes were created.

STATISTICAL TESTS

Descriptive statistics, lambda-mu-sigma (LMS)-method of Cole and Green, univariable and multivariable linear regression models. A P value < 0.05 was considered to be statistically significant.

RESULTS

In the multivariable linear model, body surface area was significantly associated with all atrial volumes and sex was significantly associated with RA volumes, LA volumes measured in the 2ch-view as well as biplane LAmax. Average atrial volumes measured: monoplane 4ch: LAmin 13.1 ± 4.8 mL/m2 , LAmax 33.4 ± 8.8 mL/m2 , RAmin 18.5 ± 6.8 mL/m2 , RAmax 33.2 ± 9.6 mL/m2 ; monoplane 2ch: LAmin 12.7 ± 4.9 mL/m2 , LAmax 30.5 ± 9.5 mL/m2 ; biplane: LAmin 12.3 ± 4.5 mL/m2 , LAmax 30.9 ± 8.7 mL/m2 .

DATA CONCLUSION

Pediatric MRI reference values for atrial volumes have been provided.

TECHNICAL EFFICACY

2 EVIDENCE LEVEL: 4.

Serial Assessment of Right Ventricular Deformation in Patients With Hypoplastic Left Heart Syndrome: A Cardiovascular Magnetic Resonance Feature Tracking Study

Background As right ventricular dysfunction is a major cause of adverse outcome in patients with hypoplastic left heart syndrome, the aim was to assess right ventricular function and deformation after Fontan completion by performing 2-dimensional cardiovascular magnetic resonance feature tracking in serial cardiovascular magnetic resonance studies. Methods and Results Cardiovascular magnetic resonance examinations of 108 patients with hypoplastic left heart syndrome (female: 31) were analyzed. Short-axis cine images were used for right ventricular volumetry. Two-dimensional cardiovascular magnetic resonance feature tracking was performed using long-axis and short-axis cine images to measure myocardial global longitudinal, circumferential, and radial strain. All patients had at least 2 cardiovascular magnetic resonance examinations after Fontan completion and 41 patients had 3 examinations. Global strain values and right ventricular ejection fraction decreased from the first to the third examination with a significant decline in global longitudinal strain from the first examination to the second examination (median, first, and third quartile: -18.8%, [-20.5;-16.5] versus -16.9%, [-19.3;-14.7]) and from the first to the third examination in 41 patients (-18.6%, [-20.9;-15.7] versus -15.8%, [-18.7;-12.6]; P-values <0.004). Right ventricular ejection fraction decreased significantly from the first to the third examination (55.4%, [49.8;59.3] versus 50.2%, [45.0;55.9]; P<0.002) and from the second to the third examination (53.8%, [47.2;58.7] versus 50.2%, [45.0;55.9]; P<0.0002). Conclusions Serial assessment of cardiovascular magnetic resonance studies in patients with hypoplastic left heart syndrome after Fontan completion demonstrates a significant reduction in global strain values and right ventricular ejection fraction at follow-up. The significant reduction in global longitudinal strain between the first 2 examinations with non-significant changes in right ventricular ejection fraction suggest that global longitudinal strain measured by 2-dimensional cardiovascular magnetic resonance feature tracking might be a superior technique for the detection of changes in myocardial function.

Myocardial deformation in patients with a single left ventricle using 2D cardiovascular magnetic resonance feature tracking: a case-control study

Ventricular dysfunction is a well-known complication in single ventricle patients in Fontan circulation. As studies exclusively examining patients with a single left ventricle (SLV) are sparse, we assessed left ventricular (LV) function in SLV patients by using 2D-cardiovascular magnetic resonance (CMR) feature tracking (2D-CMR-FT) and 2D-speckle tracking echocardiography (2D-STE). 54 SLV patients (11.4, 3.1–38.1 years) and 35 age-matched controls (12.3, 6.3–25.8 years) were included. LV global longitudinal, circumferential and radial strain (GLS, GCS, GRS) and strain rate (GLSR, GCSR, GRSR) were measured using 2D-CMR-FT. LV volumes, ejection fraction (LVEF) and mass were determined from short axis images. 2D-STE was applied in patients to measure peak systolic GLS and GLSR. In a subgroup analysis, we compared double inlet left ventricle (DILV) with tricuspid atresia (TA) patients. The population consisted of 19 DILV patients, 24 TA patients and 11 patients with diverse diagnoses. 52 patients were in NYHA class I and 2 patients were in class II. Most SLV patients had a normal systolic function but median LVEF in patients was lower compared to controls (55.6% vs. 61.2%, p = 0.0001). 2D-CMR-FT demonstrated reduced GLS, GCS and GCSR values in patients compared to controls. LVEF correlated with GS values in patients (p < 0.05). There was no significant difference between GLS values from 2D-CMR-FT and 2D-STE in the patient group. LVEF, LV volumes, GS and GSR (from 2D-CMR-FT) were not significantly different between DILV and TA patients. Although most SLV patients had a preserved EF derived by CMR, our results suggest that, LV deformation and function may behave differently in SLV patients compared to healthy subjects.