Limits...
Validation and Development of a New Automatic Algorithm for Time-Resolved Segmentation of the Left Ventricle in Magnetic Resonance Imaging.

Tufvesson J, Hedström E, Steding-Ehrenborg K, Carlsson M, Arheden H, Heiberg E - Biomed Res Int (2015)

Bottom Line: Ninety subjects imaged with a cine balanced steady state free precession sequence were included in the study (training set n = 40, test set n = 50).Manual delineation was reference standard and second observer analysis was performed in a subset (n = 25).The mean differences between automatic segmentation and manual delineation were EDV -11 mL, ESV 1 mL, EF -3%, and LVM 4 g in the test set.

View Article: PubMed Central - PubMed

Affiliation: Department of Clinical Physiology, Lund University Hospital, Lund University, 221 85 Lund, Sweden ; Department of Numerical Analysis, Centre for Mathematical Sciences, Faculty of Engineering, Lund University, 221 00 Lund, Sweden.

ABSTRACT

Introduction: Manual delineation of the left ventricle is clinical standard for quantification of cardiovascular magnetic resonance images despite being time consuming and observer dependent. Previous automatic methods generally do not account for one major contributor to stroke volume, the long-axis motion. Therefore, the aim of this study was to develop and validate an automatic algorithm for time-resolved segmentation covering the whole left ventricle, including basal slices affected by long-axis motion.

Methods: Ninety subjects imaged with a cine balanced steady state free precession sequence were included in the study (training set n = 40, test set n = 50). Manual delineation was reference standard and second observer analysis was performed in a subset (n = 25). The automatic algorithm uses deformable model with expectation-maximization, followed by automatic removal of papillary muscles and detection of the outflow tract.

Results: The mean differences between automatic segmentation and manual delineation were EDV -11 mL, ESV 1 mL, EF -3%, and LVM 4 g in the test set.

Conclusions: The automatic LV segmentation algorithm reached accuracy comparable to interobserver for manual delineation, thereby bringing automatic segmentation one step closer to clinical routine. The algorithm and all images with manual delineations are available for benchmarking.

No MeSH data available.


Related in: MedlinePlus

Detection of outflow tract (Step 7). The endocardial (red) and epicardial (green) segmentations are shown prior to the detection of outflow tract in Step 7 (a) and after adjustment of segmentation for presence of outflow tract (b) in the same basal slice in end-diastole.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4491381&req=5

fig3: Detection of outflow tract (Step 7). The endocardial (red) and epicardial (green) segmentations are shown prior to the detection of outflow tract in Step 7 (a) and after adjustment of segmentation for presence of outflow tract (b) in the same basal slice in end-diastole.

Mentions: Detection of Outflow Tract (Step 7). The deformable model gives endocardial and epicardial segmentation in all selected slices and time frames. Thereafter, long-axis motion and outflow tract are detected and in the basal slices the segmentation is adjusted accordingly. The detection of the long-axis motion is based on detecting sectors in the basal slices for which the intensities between the endocardial and epicardial segmentation are not typical for myocardium and sectors with a mean wall thickness of less than 2 millimeters. Basal slices were for detection of outflow tract defined as the most basal 40% of the ventricular length in end-diastole and all slices were divided into 24 sectors circumferentially. The intensities in basal slices are compared to intensities in all slices. Sectors with a mean intensity 2 SD above the mean are marked as sectors to remove. Sectors can only be marked as sectors to remove if the sectors are also removed in a more basal slice. Sectors to be removed are smoothed over time and circumferentially in each slice and a morphological opening is performed to get a cohesive region to remove. To remove the marked sectors a straight line is drawn for both endocardium and epicardium. Thereby a D-shaped segmentation is obtained after adjustment for presence of outflow tract. Figure 3 shows the segmentation in a basal slice before and after the detection of outflow tract.


Validation and Development of a New Automatic Algorithm for Time-Resolved Segmentation of the Left Ventricle in Magnetic Resonance Imaging.

Tufvesson J, Hedström E, Steding-Ehrenborg K, Carlsson M, Arheden H, Heiberg E - Biomed Res Int (2015)

Detection of outflow tract (Step 7). The endocardial (red) and epicardial (green) segmentations are shown prior to the detection of outflow tract in Step 7 (a) and after adjustment of segmentation for presence of outflow tract (b) in the same basal slice in end-diastole.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4491381&req=5

fig3: Detection of outflow tract (Step 7). The endocardial (red) and epicardial (green) segmentations are shown prior to the detection of outflow tract in Step 7 (a) and after adjustment of segmentation for presence of outflow tract (b) in the same basal slice in end-diastole.
Mentions: Detection of Outflow Tract (Step 7). The deformable model gives endocardial and epicardial segmentation in all selected slices and time frames. Thereafter, long-axis motion and outflow tract are detected and in the basal slices the segmentation is adjusted accordingly. The detection of the long-axis motion is based on detecting sectors in the basal slices for which the intensities between the endocardial and epicardial segmentation are not typical for myocardium and sectors with a mean wall thickness of less than 2 millimeters. Basal slices were for detection of outflow tract defined as the most basal 40% of the ventricular length in end-diastole and all slices were divided into 24 sectors circumferentially. The intensities in basal slices are compared to intensities in all slices. Sectors with a mean intensity 2 SD above the mean are marked as sectors to remove. Sectors can only be marked as sectors to remove if the sectors are also removed in a more basal slice. Sectors to be removed are smoothed over time and circumferentially in each slice and a morphological opening is performed to get a cohesive region to remove. To remove the marked sectors a straight line is drawn for both endocardium and epicardium. Thereby a D-shaped segmentation is obtained after adjustment for presence of outflow tract. Figure 3 shows the segmentation in a basal slice before and after the detection of outflow tract.

Bottom Line: Ninety subjects imaged with a cine balanced steady state free precession sequence were included in the study (training set n = 40, test set n = 50).Manual delineation was reference standard and second observer analysis was performed in a subset (n = 25).The mean differences between automatic segmentation and manual delineation were EDV -11 mL, ESV 1 mL, EF -3%, and LVM 4 g in the test set.

View Article: PubMed Central - PubMed

Affiliation: Department of Clinical Physiology, Lund University Hospital, Lund University, 221 85 Lund, Sweden ; Department of Numerical Analysis, Centre for Mathematical Sciences, Faculty of Engineering, Lund University, 221 00 Lund, Sweden.

ABSTRACT

Introduction: Manual delineation of the left ventricle is clinical standard for quantification of cardiovascular magnetic resonance images despite being time consuming and observer dependent. Previous automatic methods generally do not account for one major contributor to stroke volume, the long-axis motion. Therefore, the aim of this study was to develop and validate an automatic algorithm for time-resolved segmentation covering the whole left ventricle, including basal slices affected by long-axis motion.

Methods: Ninety subjects imaged with a cine balanced steady state free precession sequence were included in the study (training set n = 40, test set n = 50). Manual delineation was reference standard and second observer analysis was performed in a subset (n = 25). The automatic algorithm uses deformable model with expectation-maximization, followed by automatic removal of papillary muscles and detection of the outflow tract.

Results: The mean differences between automatic segmentation and manual delineation were EDV -11 mL, ESV 1 mL, EF -3%, and LVM 4 g in the test set.

Conclusions: The automatic LV segmentation algorithm reached accuracy comparable to interobserver for manual delineation, thereby bringing automatic segmentation one step closer to clinical routine. The algorithm and all images with manual delineations are available for benchmarking.

No MeSH data available.


Related in: MedlinePlus