Dataset Information

Automated quality control for segmentation of myocardial perfusion SPECT.

ABSTRACT: Left ventricular (LV) segmentation, including accurate assignment of LV contours, is essential for the quantitative assessment of myocardial perfusion SPECT (MPS). Two major types of segmentation failures are observed in clinical practices: incorrect LV shape determination and incorrect valve-plane (VP) positioning. We have developed a technique to automatically detect these failures for both nongated and gated studies.A standard Cedars-Sinai perfusion SPECT (quantitative perfusion SPECT [QPS]) algorithm was applied to derive LV contours in 318 consecutive (99m)Tc-sestamibi rest/stress MPS studies consisting of stress/rest scans with or without attenuation correction and gated stress/rest images (1,903 scans total). Two numeric parameters, shape quality control (SQC) and valve-plane quality control, were derived to categorize the respective contour segmentation failures. The results were compared with the visual classification of automatic contour adequacy by 3 experienced observers.The overall success of automatic LV segmentation in the 1,903 scans ranged from 66% on nongated images (incorrect shape, 8%; incorrect VP, 26%) to 87% on gated images (incorrect shape, 3%; incorrect VP, 10%). The overall interobserver agreement for visual classification of automatic LV segmentation was 61% for nongated scans and 80% for gated images; the agreement between gray-scale and color-scale display for these scans was 86% and 91%, respectively. To improve the reliability of visual evaluation as a reference, the cases with intra- and interobserver discrepancies were excluded, and the remaining 1,277 datasets were considered (101 with incorrect LV shape and 102 with incorrect VP position). For the SQC, the receiver-operating-characteristic area under the curve (ROC-AUC) was 1.0 +/- 0.00 for the overall dataset, with an optimal sensitivity of 100% and a specificity of 98%. The ROC-AUC was 1.0 in all specific datasets. The algorithm was also able to detect the VP position errors: VP overshooting with ROC-AUC, 0.91 +/- 0.01; sensitivity, 100%; and specificity, 70%; and VP undershooting with ROC-AUC, 0.96 +/- 0.01; sensitivity, 100%; and specificity, 70%.A new automated method for quality control of LV MPS contours has been developed and shows high accuracy for the detection of failures in LV segmentation with a variety of acquisition protocols. This technique may lead to an improvement in the objective, automated quantitative analysis of MPS.

SUBMITTER: Xu Y

PROVIDER: S-EPMC2935909 | biostudies-other | 2009 Sep

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Automated quality control for segmentation of myocardial perfusion SPECT.

Xu Yuan Y Kavanagh Paul P Fish Mathews M Gerlach James J Ramesh Amit A Lemley Mark M Hayes Sean S Berman Daniel S DS Germano Guido G Slomka Piotr J PJ

Journal of nuclear medicine : official publication, Society of Nuclear Medicine 20090818 9

<h4>Unlabelled</h4>Left ventricular (LV) segmentation, including accurate assignment of LV contours, is essential for the quantitative assessment of myocardial perfusion SPECT (MPS). Two major types of segmentation failures are observed in clinical practices: incorrect LV shape determination and incorrect valve-plane (VP) positioning. We have developed a technique to automatically detect these failures for both nongated and gated studies.<h4>Methods</h4>A standard Cedars-Sinai perfusion SPECT (q ...[more]

PMID: 19690019

Similar Datasets

Project description:BackgroundRecent studies have shown that the right ventricular (RV) quantitative analysis in myocardial perfusion imaging (MPI) SPECT can be beneficial in the diagnosis of many cardiopulmonary diseases. This study proposes a new algorithm for right ventricular 3D segmentation and quantification.MethodsThe proposed Quantitative Cardiac analysis in Nuclear Medicine imaging (QCard-NM) algorithm provides RV myocardial surface estimation and creates myocardial contour using an iterative 3D model fitting method. The founded contour is then used for quantitative RV analysis. The proposed method was assessed using various patient datasets and digital phantoms. First, the physician's manually drawn contours were compared to the QCard-NM RV segmentation using the Dice similarity coefficient (DSC). Second, using repeated MPI scans, the QCard-NM's repeatability was evaluated and compared with the QPS (quantitative perfusion SPECT) algorithm. Third, the bias of the calculated RV cavity volume was analyzed using 31 digital phantoms using the QCard-NM and QPS algorithms. Fourth, the ability of QCard-NM analysis to diagnose coronary artery diseases was assessed in 60 patients referred for both MPI and coronary angiography.ResultsThe average DSC value was 0.83 in the first dataset. In the second dataset, the coefficient of repeatability of the calculated RV volume between two repeated scans was 13.57 and 43.41 ml for the QCard-NM and QPS, respectively. In the phantom study, the mean absolute percentage errors for the calculated cavity volume were 22.6% and 42.2% for the QCard-NM and QPS, respectively. RV quantitative analysis using QCard-NM in detecting patients with severe left coronary system stenosis and/or three-vessel diseases achieved a fair performance with the area under the ROC curve of 0.77.ConclusionA novel model-based iterative method for RV segmentation task in non-gated MPI SPECT is proposed. The precision, accuracy, and consistency of the proposed method are demonstrated by various validation techniques. We believe this preliminary study could lead to developing a framework for improving the diagnosis of cardiopulmonary diseases using RV quantitative analysis in MPI SPECT.

Project description:BackgroundThe role of myocardial perfusion imaging (MPI) in patients with suspected coronary artery disease (CAD) presenting with syncope is controversial. We aimed to determine diagnostic yield of MPI for evaluation of syncope in patients without known CAD, as a function of pre-test patient risk and test modality (PET vs SPECT).MethodsBetween 1/2010 and 12/2016, 1324 consecutive patients presenting with syncope without known CAD underwent MPI with PET (n = 640) or SPECT (n = 684). Rates of abnormal MPI (summed difference score (SDS) > 2 or left ventricular ejection fraction (LVEF) reserve ≤ 0 for PET and SDS > 2 or post-stress LVEF ≤ 45% for SPECT) were determined among patients stratified by pre-test risk. In patients who were referred for coronary angiography, diagnostic yield of obstructive CAD was calculated in the overall cohort as well as in a propensity-matched cohort compared to patients without syncope.ResultsAbnormal MPI was noted in 36.5% (201/551) of patients who had PET compared with 13.0% (87/671) who had SPECT (P < 0.001), which is largely related to higher comorbidity burden and greater pre-test CAD risk in the PET population. Among patients who had an abnormal MPI, 8.5% (47/551) with PET and 0.7% (5/671) with SPECT were found to have obstructive CAD if referred for coronary angiography. Patients at intermediate-high pre-test risk had a higher proportion of abnormal MPIs and obstructive CAD as compared to those at low risk in both the PET and SPECT cohorts. The rate of abnormal testing and diagnostic yield of PET MPI was similar and proportionate to pre-test likelihood among matched patients with and without syncope.ConclusionsAmong patients referred for PET MPI with syncope at an intermediate-high pre-test CAD risk, 1 in 3 had an abnormal MPI and 1 in 10 had obstructive CAD. The value of MPI was related to pre-test risk as opposed to the presence of syncope, and MPI testing with PET or SPECT in the low-risk population was low value.

Dataset Information

Automated quality control for segmentation of myocardial perfusion SPECT.

Publications

Automated quality control for segmentation of myocardial perfusion SPECT.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets