Improved quantification and normal limits for myocardial perfusion stress-rest change

Quantitative clinical nuclear cardiology, part 2: Evolving/emerging applications

Quantitative analysis has been applied extensively to image processing and interpretation in nuclear cardiology to improve disease diagnosis and risk stratification. This is Part 2 of a two-part continuing medical education article, which will review the potential clinical role for emerging quantitative analysis tools. The article will describe advanced methods for quantifying dyssynchrony, ventricular function and perfusion, and hybrid imaging analysis. This article discusses evolving methods to measure myocardial blood flow with positron emission tomography and single-photon emission computed tomography. Novel quantitative assessments of myocardial viability, microcalcification and in patients with cardiac sarcoidosis and cardiac amyloidosis will also be described. Lastly, we will review the potential role for artificial intelligence to improve image analysis, disease diagnosis, and risk prediction. The potential clinical role for all these novel techniques will be highlighted as well as methods to optimize their implementation.

Quantitative clinical nuclear cardiology, part 2: Evolving/emerging applications

Quantitative analysis has been applied extensively to image processing and interpretation in nuclear cardiology to improve disease diagnosis and risk stratification. This is Part 2 of a two-part continuing medical education article, which will review the potential clinical role for emerging quantitative analysis tools. The article will describe advanced methods for quantifying dyssynchrony, ventricular function and perfusion, and hybrid imaging analysis. This article discusses evolving methods to measure myocardial blood flow with positron emission tomography and single-photon emission computed tomography. Novel quantitative assessments of myocardial viability, microcalcification and in patients with cardiac sarcoidosis and cardiac amyloidosis will also be described. Lastly, we will review the potential role for artificial intelligence to improve image analysis, disease diagnosis, and risk prediction. The potential clinical role for all these novel techniques will be highlighted as well as methods to optimize their implementation. (J Nucl Cardiol 2020).

Automated SPECT analysis compared with expert visual scoring for the detection of FFR-defined coronary artery disease

Purpose

Traditionally, interpretation of myocardial perfusion imaging (MPI) is based on visual assessment. Computer-based automated analysis might be a simple alternative obviating the need for extensive reading experience. Therefore, the aim of the present study was to compare the diagnostic performance of automated analysis with that of expert visual reading for the detection of obstructive coronary artery disease (CAD).

Methods

206 Patients (64% men, age 58.2 ± 8.7 years) with suspected CAD were included prospectively. All patients underwent ^99mTc-tetrofosmin single-photon emission computed tomography (SPECT) and invasive coronary angiography with fractional flow reserve (FFR) measurements. Non-corrected (NC) and attenuation-corrected (AC) SPECT images were analyzed both visually as well as automatically by commercially available SPECT software. Automated analysis comprised a segmental summed stress score (SSS), summed difference score (SDS), stress total perfusion deficit (S-TPD), and ischemic total perfusion deficit (I-TPD), representing the extent and severity of hypoperfused myocardium. Subsequently, software was optimized with an institutional normal database and thresholds. Diagnostic performances of automated and visual analysis were compared taking FFR as a reference.

Results

Sensitivity did not differ significantly between visual reading and most automated scoring parameters, except for SDS, which was significantly higher than visual assessment (p < 0.001). Specificity, however, was significantly higher for visual reading than for any of the automated scores (p < 0.001 for all). Diagnostic accuracy was significantly higher for visual scoring (77.2%) than for all NC images scores (p < 0.05), but not compared with SSS AC and S-TPD AC (69.8% and 71.2%, p = 0.063 and p = 0.134). After optimization of the automated software, diagnostic accuracies were similar for visual (73.8%) and automated analysis. Among the automated parameters, S-TPD AC showed the highest accuracy (73.5%).

Conclusion

Automated analysis of myocardial perfusion SPECT can be as accurate as visual interpretation by an expert reader in detecting significant CAD defined by FFR.

Electronic supplementary material

The online version of this article (10.1007/s00259-018-3951-1) contains supplementary material, which is available to authorized users.

The value of local normal limits in quantitative Thallium-201 CZT MPI SPECT

BACKGROUND

Considering the distinctive characteristics of CZT detectors, automatic quantification of ischemia using normal limits included with software package may deliver suboptimal results for CAD detection. The present study aims to evaluate the benefits of creating normal limits specific to a local population and laboratory protocol.

METHODS AND RESULTS

Two groups were selected from patients who had undergone a CZT MPI. Normal limits were generated with the QPS application based on the population with low likelihood of CAD. Using the vendor-supplied and the population-specific normal limits i-TPD and vessel-specific SDS results obtained for patients who had subsequently undergone coronary angiography were compared with coronary angiography data. A weak correlation was observed for low i-TPD (stress TPD minus rest TPD) and SDS values. Both databases gave similar values for the area under the ROC curve concerning i-TPD (0.75 to 0.74) and SDS results (0.72 to 0.75 for the LAD, 0.62 to 0.64 for the LCx, and 0.63 to 0.67 for the RCA). Sensitivity (60%), specificity (78%), and predictive positive (84%) and negative (52%) values were also similar with a diagnostic and prognostic threshold value.

CONCLUSION

The use of a population-specific created database did not influence the diagnostic value of thallium-201 MPI QPS results using a CZT camera.

Novel SPECT Technologies and Approaches in Cardiac Imaging

Recent novel approaches in myocardial perfusion single photon emission CT (SPECT) have been facilitated by new dedicated high-efficiency hardware with solid-state detectors and optimized collimators. New protocols include very low-dose (1 mSv) stress-only, two-position imaging to mitigate attenuation artifacts, and simultaneous dual-isotope imaging. Attenuation correction can be performed by specialized low-dose systems or by previously obtained CT coronary calcium scans. Hybrid protocols using CT angiography have been proposed. Image quality improvements have been demonstrated by novel reconstructions and motion correction. Fast SPECT acquisition facilitates dynamic flow and early function measurements. Image processing algorithms have become automated with virtually unsupervised extraction of quantitative imaging variables. This automation facilitates integration with clinical variables derived by machine learning to predict patient outcome or diagnosis. In this review, we describe new imaging protocols made possible by the new hardware developments. We also discuss several novel software approaches for the quantification and interpretation of myocardial perfusion SPECT scans.

Automated Quantitative Nuclear Cardiology Methods

Quantitative analysis of SPECT and PET has become a major part of nuclear cardiology practice. Current software tools can automatically segment the left ventricle, quantify function, establish myocardial perfusion maps, and estimate global and local measures of stress/rest perfusion, all with minimal user input. State-of-the-art automated techniques have been shown to offer high diagnostic accuracy for detecting coronary artery disease, as well as predict prognostic outcomes. This article briefly reviews these techniques, highlights several challenges, and discusses the latest developments.

Validation of an automated method to quantify stress-induced ischemia and infarction in rest-stress myocardial perfusion SPECT

BACKGROUND

Myocardial perfusion SPECT (MPS) is one of the frequently used methods for quantification of perfusion defects in patients with known or suspected coronary artery disease. This article describes open access software for automated quantification in MPS of stress-induced ischemia and infarction and provides phantom and in vivo validation.

METHODS AND RESULTS

A total of 492 patients with known or suspected coronary artery disease underwent both stress and rest MPS. The proposed perfusion analysis algorithm (Segment) was trained in 140 patients and validated in the remaining 352 patients using visual scoring in MPS by an expert reader as reference standard. Furthermore, validation was performed with simulated perfusion defects in an anthropomorphic computer model. Total perfusion deficit (TPD, range 0-100), including both extent and severity of the perfusion defect, was used as the global measurement of the perfusion defects. Mean bias ± SD between TPD by Segment and the simulated TPD was 3.6 ± 3.8 (R(2) = 0.92). Mean bias ± SD between TPD by Segment and the visual scoring in the patients was 1.2 ± 2.9 (R (2)= 0.64) for stress-induced ischemia and -0.3 ± 3.1 (R(2) = 0.86) for infarction.

CONCLUSION

The proposed algorithm can detect and quantify perfusion defects in MPS with good agreement to expert readers and to simulated values in a computer phantom.

Improved accuracy of myocardial perfusion SPECT for detection of coronary artery disease by machine learning in a large population

OBJECTIVE

We aimed to improve the diagnostic accuracy of myocardial perfusion SPECT (MPS) by integrating clinical data and quantitative image features with machine learning (ML) algorithms.

METHODS

1,181 rest (201)Tl/stress (99m)Tc-sestamibi dual-isotope MPS studies [713 consecutive cases with correlating invasive coronary angiography (ICA) and suspected coronary artery disease (CAD) and 468 with low likelihood (LLk) of CAD <5%] were considered. Cases with stenosis <70% by ICA and LLk of CAD were considered normal. Total stress perfusion deficit (TPD) for supine/prone data, stress/rest perfusion change, and transient ischemic dilatation were derived by automated perfusion quantification software and were combined with age, sex, and post-electrocardiogram CAD probability by a boosted ensemble ML algorithm (LogitBoost). The diagnostic accuracy of the model for prediction of obstructive CAD ≥70% was compared to standard prone/supine quantification and to visual analysis by two experienced readers utilizing all imaging, quantitative, and clinical data. Tenfold stratified cross-validation was performed.

RESULTS

The diagnostic accuracy of ML (87.3% ± 2.1%) was similar to Expert 1 (86.0% ± 2.1%), but superior to combined supine/prone TPD (82.8% ± 2.2%) and Expert 2 (82.1% ± 2.2%) (P < .01). The receiver operator characteristic areas under curve for ML algorithm (0.94 ± 0.01) were higher than those for TPD and both visual readers (P < .001). The sensitivity of ML algorithm (78.9% ± 4.2%) was similar to TPD (75.6% ± 4.4%) and Expert 1 (76.3% ± 4.3%), but higher than that of Expert 2 (71.1% ± 4.6%), (P < .01). The specificity of ML algorithm (92.1% ± 2.2%) was similar to Expert 1 (91.4% ± 2.2%) and Expert 2 (88.3% ± 2.5%), but higher than TPD (86.8% ± 2.6%), (P < .01).

CONCLUSION

ML significantly improves diagnostic performance of MPS by computational integration of quantitative perfusion and clinical data to the level rivaling expert analysis.

Quantitative analysis of perfusion studies: strengths and pitfalls

Tools for automated quantification of myocardial perfusion are available to nuclear cardiology practitioners and researchers. These methods have demonstrated superior reproducibility with comparable diagnostic and prognostic performance, when compared with segmental visual scoring by expert observers. A particularly useful application of the quantitative analysis can be in the detection of subtle changes or in precise determination of ischemia. Some challenges remain in the routine application of perfusion quantification. Multiple quantitative parameters may need to be reconciled by the expert reader for the final diagnosis. Computer analysis may be sensitive to imaging artifacts, resulting in false positive scans. Perfusion quantification may require site specific normal limits and some degree of manual interaction. New software improvements have been proposed to address some of these challenges.

Tumor dosimetry and response for 153Sm-ethylenediamine tetramethylene phosphonic acid therapy of high-risk osteosarcoma

(153)Sm-ethylenediamine tetramethylene phosphonic acid ((153)Sm-EDTMP) therapy for osteosarcoma is being investigated. In this study, we analyzed the influence of (153)Sm-EDTMP administered activity (AA), osteosarcoma tumor density, mass, and the shape of the tumor on absorbed dose (AD). We also studied the biologic implication of the nonuniform tumor AD distribution using radiobiologic modeling and examined the relationship between tumor AD and response.

METHODS

Nineteen tumors in 6 patients with recurrent, refractory osteosarcoma enrolled in a phase I or II clinical trial of (153)Sm-EDTMP were analyzed using the 3-dimensional radiobiologic dosimetry (3D-RD) software package. Patients received a low dose of (153)Sm-EDTMP (37.0-51.8 MBq/kg), followed on hematologic recovery by a second, high dose (222 MBq/kg). Treatment response was evaluated using either CT or MRI after each therapy. SPECT/CT of the tumor regions were obtained at 4 and 48 h or 72 h after (153)Sm-EDTMP therapy for 3D-RD analysis. Mean tumor AD was also calculated using the OLINDA/EXM unit-density sphere model and was compared with the 3D-RD estimates.

RESULTS

On average, a 5-fold increase in the AA led to a 4-fold increase in the mean tumor AD over the high- versus low-dose-treated patients. The range of mean tumor AD and equivalent uniform dose (EUD) for low-dose therapy were 1.48-14.6 and 0.98-3.90 Gy, respectively. Corresponding values for high-dose therapy were 2.93-59.3 and 1.89-12.3 Gy, respectively. Mean tumor AD estimates obtained from OLINDA/EXM were within 5% of the mean AD values obtained using 3D-RD. On an individual tumor basis, both mean AD and EUD were positively related to percentage tumor volume reduction (P = 0.031 and 0.023, respectively).

CONCLUSION

The variations in tumor density, mass, and shape seen in these tumors did not affect the mean tumor AD estimation significantly. The tumor EUD was approximately 2- and 3-fold lower than the mean AD for low- and high-dose therapy, respectively. A dose-response relationship was observed for transient tumor volume shrinkage.