AGMN: Association Graph-based Graph Matching Network for Coronary Artery Semantic Labeling on Invasive Coronary Angiograms.

Semantic labeling of coronary arterial segments in invasive coronary angiography (ICA) is important for automated assessment and report generation of coronary artery stenosis in computer-aided coronary artery disease (CAD) diagnosis. However, separating and identifying individual coronary arterial segments is challenging because morphological similarities of different branches on the coronary arterial tree and human-to-human variabilities exist. Inspired by the training procedure of interventional cardiologists for interpreting the structure of coronary arteries, we propose an association graph-based graph matching network (AGMN) for coronary arterial semantic labeling. We first extract the vascular tree from invasive coronary angiography (ICA) and convert it into multiple individual graphs. Then, an association graph is constructed from two individual graphs where each vertex represents the relationship between two arterial segments. Thus, we convert the arterial segment labeling task into a vertex classification task; ultimately, the semantic artery labeling becomes equivalent to identifying the artery-to-artery correspondence on graphs. More specifically, the AGMN extracts the vertex features by the embedding module using the association graph, aggregates the features from adjacent vertices and edges by graph convolution network, and decodes the features to generate the semantic mappings between arteries. By learning the mapping of arterial branches between two individual graphs, the unlabeled arterial segments are classified by the labeled segments to achieve semantic labeling. A dataset containing 263 ICAs was employed to train and validate the proposed model, and a five-fold cross-validation scheme was performed. Our AGMN model achieved an average accuracy of 0.8264, an average precision of 0.8276, an average recall of 0.8264, and an average F1-score of 0.8262, which significantly outperformed existing coronary artery semantic labeling methods. In conclusion, we have developed and validated a new algorithm with high accuracy, interpretability, and robustness for coronary artery semantic labeling on ICAs.

Clinical impacts of scar reduction on gated myocardial perfusion SPECT after cardiac resynchronization therapy.

BACKGROUND: It had not been reported that myocardial scar shown on gated myocardial perfusion SPECT (GMPS) might reduce after cardiac resynchronization therapy (CRT). In this study, we aim to investigate the clinical impact and characteristic of scar reduction (SR) after CRT. METHODS AND RESULTS: Sixty-one heart failure patients following standard indication for CRT received twice GMPS as pre- and post-CRT evaluations. The patients with an absolute reduction of scar ≥ 10% after CRT were classified as the SR group while the rest were classified as the non-SR group. The SR group (N = 22, 36%) showed more improvement on LV function (∆LVEF: 18.1 ± 12.4 vs 9.4 ± 9.9 %, P = 0.007, ∆ESV: - 91.6 ± 52.6 vs - 38.1 ± 46.5 mL, P < 0.001) and dyssynchrony (ΔPSD: - 26.19 ± 18.42 vs - 5.8 ± 23.0°, P < 0.001, Δ BW: - 128.7 ± 82.8 vs - 25.2 ± 109.0°, P < 0.001) than non-SR group (N = 39, 64%). Multivariate logistic regression analysis showed baseline QRSd (95% CI 1.019-1.100, P = 0.006) and pre-CRT Reduced Wall Thickening (RWT) (95% CI 1.016-1.173, P = 0.028) were independent predictors for the development of SR. CONCLUSION: More than one third of patients showed SR after CRT who had more post-CRT improvement on LV function and dyssynchrony than those without SR. Wider QRSd and higher RWT before CRT were related to the development of SR after CRT.

Comparison of ejection fraction calculation between CT and SPECT at high heart rate: A dynamic cardiac phantom study.

The purpose of this study is to compare the ejection fraction (EF) calculation of CT and SPECT at high heart rate. A dynamic cardiac phantom with programmable end-systolic volume (ESV), end-diastolic volume (EDV), and heart rate was used to compare CT, which has high spatial resolution (< 1 mm) and modest temporal resolution of 175 msec, and SPECT, which has high temporal resolution of 16 bins per cardiac cycle but poor spatial resolution (> 1 cm) in EF, ESV, and EDV at the heart rates ≤ 100 bpm for EF = 30 (disease state) and EF = 60 (healthy state). EF calculations for SPECT were accurate in 2% for 40 to 100 bpm for both EF = 30 and EF = 60, and were not heart rate dependent although both ESV and EDV could be underestimated by 18-20%. EF calculations for CT were accurate in 2.2% for 40 and 60 bpm. Inaccuracy in EF calculations, ESV and EDV estimates increased when the heart rate or EF increased. SPECT was accurate for EF calculation for the heart rates ≤ 100 bpm and CT was accurate for the heart rates of ≤ 60 bpm. CT was less accurate for the high heart rates of 80 and 100 bpm, or high EF = 60.

Function-based dementia severity assessment for vascular cognitive impairment.

BACKGROUND/PURPOSES: Unimpaired activities of daily living (ADL) is essential for the diagnosis of normal cognition and mild cognitive impairment. However, diagnosis according to this concept is difficult to apply to patients comorbid with motor dysfunction. We aim to use a novel ADL questionnaire for operationally diagnosing unimpaired ADL in vascular cognitive impairment with no dementia (VCIND). METHODS AND PARTICIPANTS: This was a retrospective cohort study with both cross-sectional and long-term follow-up analysis. Patients with cerebrovascular disease with normal cognition (CVDNC), VCIND, and vascular dementia (VaD) were analyzed. Cutoff scores for differentiating different stages of cognitive impairment were compared between the new History-based Artificial Intelligent ADL questionnaire (HAI-ADL) and other tools. RESULTS: A total of 596 individuals were analyzed, including 40 CVDNC, 167 VCIND, 218 mild, 119 moderate, and 52 severe-dementia patients. The cutoff scores for determining unimpaired ADL in VCIND were 8.5, 3.5, 5, 100, and 60 in HAI-ADL, CDR-SB, IADL, BI, and CASI, respectively. HAI-ADL had the highest correlations with CDR-SB and the CDR staging system compared to other tools. Four models of progression rates from CVDNC/VCIND to VaD revealed it was much higher in the group with HAI-ADL > 8.5 compared to those with HAI-ADL≦8.5 with odds ratios of 3.75, 3.66, 3.31, and 2.77, respectively. CONCLUSION: Our study showed that HAI-ADL provides an operational determinates unimpaired ADL which is necessary for the diagnosis of VCIND. The predictive value for progression to dementia was proved by a long-term follow-up analysis of the research cohort.

Simplified quantification of 13N-ammonia PET myocardial blood flow: A comparative study with the standard compartment model to facilitate clinical use.

BACKGROUND: Short imaging protocol to quantify myocardial blood flow (MBF) and myocardial flow reserve (MFR) may enhance the clinical application of 13N-ammonia cardiac PET. We assessed the flow quantitation of 13N-ammonia PET implementing simple retention model and two-compartment model. METHODS: Fourteen healthy volunteers (HVT) and twenty-three clinical patients received 13N-ammonia PET/CT. The simple retention model used the first 7-minute image to quantify MBF. Global and regional MBF and MFR of the two models were compared. RESULTS: Global and regional MBF and MFR of these two models were highly correlated with mildly inferior correlation in RCA territory (global R2: rest MBF = 0.79, stress MBF = 0.65, MFR = 0.77; regional R2: rest MBF ≥ 0.72, stress MBF ≥ 0.52, MFR ≥ 0.68). There were significant differences for MFR (4.04 ± 0.72, 3.66 ± 0.48, p = .02) and rest MBF (0.69 ± 0.12, 0.78 ± 0.12, p = .02) between the two models in the HVT group. CONCLUSIONS: 13N-ammonia global and regional MBF and MFR from the simple retention model demonstrate strong correlations with that from the two-compartment model. Significant differences of MFR and rest MBF are noted in the HVT group, with a proposed normal reference value for the 13N-ammonia short simple retention protocol.

The Utilization of Stress Tests Prior to Percutaneous Coronary Intervention for Stable Coronary Artery Disease in Taiwan.

BACKGROUND: Ischemia shown in non-invasive tests is considered to be a fundamental requirement for treating patients with stable coronary artery disease (CAD) with a percutaneous coronary intervention (PCI). In a nationwide cohort, we investigated the utilization of stress tests, including myocardial perfusion imaging (MPI), treadmill exercise test (TET) and stress echocardiography (SE) prior to elective PCI. METHODS: This retrospective study used the Longitudinal Health Insurance Database 2000 (LHID2000) of the National Health Insurance program in Taiwan. The LHID2000 is comprised of one million randomly sampled beneficiaries. We enrolled patients receiving elective PCI for stable CAD from 2000 to 2013. Stress tests performed within 90 days prior to PCI and patient characteristics correlated with the utilization of stress tests were investigated. RESULTS: During the investigation period, 3,163 patients received elective PCI for stable CAD and 1,847 (58.4%) patients had at least one stress test within 90 days prior to PCI. Among them, 1,461 (79.1%) had MPI, 1,228 had TET (66.4%) and only 1 had SE (0.05%). Age < 80 years, regional hospital and hyperlipidemia were independently associated with an increased likelihood of receiving stress tests. On the other hand, Charlson-comorbidity index score ≥ 1, prior catheterization and heart failure were independently associated with a decreased likelihood of receiving stress tests. CONCLUSIONS: In the setting of stable CAD, almost 60% of our patients received stress tests within 90 days prior to elective PCI, and MPI was the most commonly used test.

Impact of initial myocardial perfusion imaging versus invasive coronary angiography on outcomes in coronary artery disease: a nationwide cohort study.

PURPOSE: In patients with stable coronary artery disease (CAD), two main options exist to guide management: initial invasive coronary angiography (CAG), or selective CAG after risk stratification using myocardial perfusion imaging (MPI). This study compared clinical outcomes between these two strategies in a large, real-world population. METHODS: The initial cohort comprised 1,000,000 randomly selected patients who had been entered in the National Health Insurance Research Database of Taiwan between 2000 and 2011. Patients with acute coronary syndromes, prior myocardial infarction (MI) or coronary revascularization, and prior treadmill testing or stress echocardiography were excluded. The remaining patients with suspected or known CAD were divided into those in whom initial CAG had been performed and those in whom initial MPI had been performed, and were followed until the end of 2011 for all-cause mortality, MI, and revascularization. A Cox proportional hazards model was used to estimate the risk of events after adjusting for covariates. RESULTS: The MPI and CAG groups each comprised 4,495 patients after frequency matching, with a similar Charlson comorbidity index (CCI). The MPI group had a significantly and dramatically lower incidence of revascularization (729 vs. 2,380, p < 0.001), MI (268 vs. 1,044, p < 0.001), and all-cause mortality (522 vs. 784, p < 0.001) than the CAG group. Multivariable analysis adjusting for age, gender, CCI, and comorbidities showed that in the MPI group fewer patients had revascularization (HR 0.24, 95% CI 0.22-0.26) and MI (HR 0.23, 95% CI 0.20-0.26), and the rate of all-cause mortality was lower (HR 0.58, 95% CI 0.52-0.64). CONCLUSIONS: In patients with suspected stable CAD, compared with initial invasive CAG, a selective strategy guided by MPI was associated with lower rates of revascularization and MI and improved survival.

An Informant-Based Simple Questionnaire for Language Assessment in Neurodegenerative Disorders.

BACKGROUND/AIMS: Language dysfunction is a crucial feature of brain disorders. This study investigated language dysfunction in patients with dementia with or without parkinsonism by using an informant-based simple questionnaire. METHODS: Language dysfunction in normal controls (NCs), and patients with Parkinson disease (PD), Parkinson disease dementia (PDD), dementia with Lewy bodies (DLB), or Alzheimer disease (AD) were analyzed and compared. RESULTS: A total of 1,662 individuals were studied: 285 NCs, 157 PD patients, 161 PDD patients, 248 DLB patients, and 811 AD patients. Patients with PD displayed higher frequency of language dysfunction in several language domains than NC. Patients with PDD and DLB showed higher frequency of language dysfunction in most of the language domains than those with AD. A composite score of our simple questionnaire was comparable with the score for the language domain of the Cognitive Abilities Screening Instrument (CASI) in different stages of dementia due to Lewy body diseases or not. CONCLUSION: Our study showed that the informant-based simple questionnaire is a practical screening tool and is comparable with the language subscale of CASI. This tool can be applied in clinical practice and in the registration platform for rapid language dysfunction screening.

SPECT myocardial blood flow quantitation toward clinical use: a comparative study with 13N-Ammonia PET myocardial blood flow quantitation.

OBJECTIVES: The aim of this study was to evaluate the accuracy of myocardial blood flow (MBF) quantitation of 99mTc-Sestamibi (MIBI) single photon emission computed tomography (SPECT) compared with 13N-Ammonia (NH3) position emission tomography (PET) on the same cohorts. BACKGROUND: Recent advances of SPECT technologies have been applied to develop MBF quantitation as a promising tool to diagnose coronary artery disease (CAD) for areas where PET MBF quantitation is not available. However, whether the SPECT approach can achieve the same level of accuracy as the PET approach for clinical use still needs further investigations. METHODS: Twelve healthy volunteers (HVT) and 16 clinical patients with CAD received both MIBI SPECT and NH3 PET flow scans. Dynamic SPECT images acquired with high temporary resolution were fully corrected for physical factors and processed to quantify K1 using the standard compartmental modeling. Human MIBI tracer extraction fraction (EF) was determined by comparing MIBI K1 and NH3 flow on the HVT group and then used to convert flow values from K1 for all subjects. MIBI and NH3 flow values were systematically compared to validate the SPECT approach. RESULTS: The human MIBI EF was determined as [1.0-0.816*exp(-0.267/MBF)]. Global and regional MBF and myocardial flow reserve (MFR) of MIBI SPECT and NH3 PET were highly correlated for all subjects (global R2: MBF = 0.92, MFR = 0.78; regional R2: MBF ≥ 0.88, MFR ≥ 0.71). No significant differences for rest flow, stress flow, and MFR between these two approaches were observed (All p ≥ 0.088). Bland-Altman plots overall revealed small bias between MIBI SPECT and NH3 PET (global: ΔMBF = -0.03Lml/min/g, ΔMFR = 0.07; regional: ΔMBF = -0.07 - 0.06 , ΔMFR = -0.02 - 0.22). CONCLUSIONS: Quantitation with SPECT technologies can be accurate to measure myocardial blood flow as PET quantitation while comprehensive imaging factors of SPECT to derive the variability between these two approaches were fully addressed and corrected.

Impact of cardiac reverse remodeling after cardiac resynchronization therapy assessed by myocardial perfusion imaging on ventricular arrhythmia.

BACKGROUND: Although cardiac resynchronization therapy (CRT) has been a useful treatment of heart failure, patients with CRT are still in risk of sudden cardiac death due to ventricular arrhythmia. The aim of this study was to investigate the impact of cardiac reverse remodeling after CRT on the prevalence of ventricular tachycardia or fibrillation (VT/VF). METHODS AND RESULTS: Forty-one heart failure patients (26 men, age 66 ± 10 years), who were implanted with CRT for at least 12 months, were enrolled. All patients received myocardial perfusion imaging (MPI) under CRT pacing to evaluate left ventricle (LV) function, dyssynchrony, and scar. VT/VF episodes during the follow-up period after MPI were recorded by the CRT devices. Sixteen patients (N = 16/41, 39%) were found to have VT/VF. Multivariate Cox regression analysis and receiver operating characteristic curve analysis showed that five risk factors were significant predictors of VT/VF, including increased left ventricle ejection fraction (LVEF) by ≤7% after CRT, low LVEF after CRT (≤30%), change of intrinsic QRS duration (iQRSd) by ≤7 ms, wide iQRSd after CRT (≥121 ms), and high systolic dyssynchrony after CRT (phase standard deviation ≥45.6°). For those patients with all of the 5 risk factors, 85.7% or more developed VT/VF. CONCLUSIONS: The characteristics of cardiac reverse remodeling after CRT as assessed by MPI are associated with the prevalence of ventricular arrhythmia.

Prediction of Both Electrical and Mechanical Reverse Remodeling on Acute Electrocardiogram Changes After Cardiac Resynchronization Therapy.

BACKGROUND: The development of both electrical reverse remodeling and mechanical reverse remodeling (ERR+MRR) after cardiac resynchronization therapy (CRT) implantation could reduce the incidence of lethal arrhythmia, hence the prediction of ERR+MRR is clinically important.Methods and Results:Eighty-three patients (54 male; 67±12 years old) with CRT >6 months were enrolled. ERR was defined as baseline intrinsic QRS duration (iQRSd) shortening ≥10 ms in lead II on ECG after CRT, and MRR as improvement in LVEF ≥25% on echocardiography after CRT. Acute ECG changes were measured by comparing the pre-implant and immediate post-implant ECG. Ventricular arrhythmia episodes, including ventricular tachycardia and ventricular fibrillation, detected by the implanted device were recorded. Patients were classified as ERR only (n=12), MRR only (n=23), ERR+MRR (n=26), or non-responder (ERR- & MRR-, n=22). On multivariate regression analysis, difference between baseline intrinsic QRS and paced QRS duration (∆QRSd) >35 ms was a significant predictor of ERR+MRR (sensitivity, 68%; specificity, 64%; AUC, 0.7; P=0.003), and paced QTc >443 ms was a negative predictor of ERR+MRR (sensitivity, 78%; specificity, 60%; AUC, 0.7; P=0.002). On Cox proportional hazard modeling, ERR+MRR may reduce risk of ventricular arrhythma around 70% compared with non-responder (HR, 0.29; 95% CI: 0.13-0.65). CONCLUSIONS: Acute ECG changes after CRT were useful predictors of ERR+MRR. ERR+MRR was also a protective factor for ventricular arrhythmia.

New Trends in Radionuclide Myocardial Perfusion Imaging.

UNLABELLED: Radionuclide myocardial perfusion imaging (MPI) with single photon emission computed tomography (SPECT) has been widely used clinically as one of the major functional imaging modalities for patients with coronary artery disease (CAD) for decades. Ample evidence has supported the use of MPI as a useful and important tool in the diagnosis, risk stratification and treatment planning for CAD. Although popular in the United States, MPI has become the most frequently used imaging modality among all nuclear medicine tests in Taiwan. However, it should be acknowledged that MPI SPECT does have its limitations. These include false-positive results due to certain artifacts, false-negative due to balanced ischemia, complexity and adverse reaction arising from current pharmacological stressors, time consuming nature of the imaging procedure, no blood flow quantitation and relatively high radiation exposure. The purpose of this article was to review the recent trends in nuclear cardiology, including the utilization of positron emission tomography (PET) for MPI, new stressor, new SPECT camera with higher resolution and higher sensitivity, dynamic SPECT protocol for blood flow quantitation, new software of phase analysis for evaluation of LV dyssynchrony, and measures utilized for reducing radiation exposure of MPI. KEY WORDS: Coronary artery disease • Myocardial flow reserve • Myocardial perfusion imaging • Phase analysis • PET • SPECT.

Impact of right-ventricular apical pacing on the optimal left-ventricular lead positions measured by phase analysis of SPECT myocardial perfusion imaging.

PURPOSE: The use of SPECT phase analysis to optimize left-ventricular (LV) lead positions for cardiac resynchronization therapy (CRT) was performed at baseline, but CRT works as simultaneous right ventricular (RV) and LV pacing. The aim of this study was to assess the impact of RV apical (RVA) pacing on optimal LV lead positions measured by SPECT phase analysis. METHODS: This study prospectively enrolled 46 patients. Two SPECT myocardial perfusion scans were acquired under sinus rhythm with complete left bundle branch block and RVA pacing, respectively, following a single injection of (99m)Tc-sestamibi. LV dyssynchrony parameters and optimal LV lead positions were measured by the phase analysis technique and then compared between the two scans. RESULTS: The LV dyssynchrony parameters were significantly larger with RVA pacing than with sinus rhythm (p ~0.01). In 39 of the 46 patients, the optimal LV lead positions were the same between RVA pacing and sinus rhythm (kappa = 0.861). In 6 of the remaining 7 patients, the optimal LV lead positions were along the same radial direction, but RVA pacing shifted the optimal LV lead positions toward the base. CONCLUSION: The optimal LV lead positions measured by SPECT phase analysis were consistent, no matter whether the SPECT images were acquired under sinus rhythm or RVA pacing. In some patients, RVA pacing shifted the optimal LV lead positions toward the base. This study supports the use of baseline SPECT myocardial perfusion imaging to optimize LV lead positions to increase CRT efficacy.

Quantitation of myocardial blood flow and myocardial flow reserve with 99mTc-sestamibi dynamic SPECT/CT to enhance detection of coronary artery disease.

PURPOSE: Conventional dual-head single photon emission computed tomography (SPECT)/CT systems capable of fast dynamic SPECT (DySPECT) imaging have a potential for flow quantitation. This study introduced a new method to quantify myocardial blood flow (MBF) and myocardial flow reserve (MFR) with DySPECT scan and evaluated the diagnostic performance of detecting coronary artery disease (CAD) compared with perfusion using invasive coronary angiography (CAG) as the reference standard. METHODS: This study included 21 patients with suspected or known CAD who had received DySPECT, ECG-gated SPECT (GSPECT), and CAG (13 with ≥ 50% stenosis in any vessel; non-CAD group: 8 with patent arteries or < 50% stenosis). DySPECT and GSPECT scans were performed on a widely used dual-head SPECT/CT scanner. The DySPECT imaging protocol utilized 12-min multiple back-and-forth gantry rotations during injections of (99m)Tc-sestamibi (MIBI) tracer at rest or dipyridamole-stress stages. DySPECT images were reconstructed with full physical corrections and converted to the physical unit of becquerels per milliliter. Stress MBF (SMBF), rest MBF (RMBF), and MFR were quantified by a one-tissue compartment flow model using time-activity curves derived from DySPECT images. Perfusion images were processed for GSPECT scan and interpreted to obtain summed stress score (SSS) and summed difference score (SDS). Receiver-operating characteristic (ROC) analyses were conducted to evaluate the diagnostic performance of flow and perfusion. RESULTS: Using the criteria of ≥ 50% stenosis as positive CAD, areas under the ROC curve (AUCs) of flow assessment were overall significantly greater than those of perfusion. For patient-based analysis, AUCs for MFR, SMBF, SSS, and SDS were 0.91 ± 0.07, 0.86 ± 0.09, 0.64 ± 0.12, and 0.59 ± 0.13. For vessel-based analysis, AUCs for MFR, SMBF, SSS, and SDS were 0.81 ± 0.05, 0.76 ± 0.06, 0.62 ± 0.07, and 0.56 ± 0.08, respectively. CONCLUSION: The preliminary data suggest that MBF quantitation with a conventional SPECT/CT system and the flow quantitation method is a clinically effective approach to enhance CAD detection.

Relation of early post-stress left ventricular dyssynchrony and the extent of angiographic coronary artery disease.

BACKGROUND: Previous studies showed different dyssynchrony patterns between ischemic and normal myocardium at early post-stress using Tl-201 gated SPECT myocardial perfusion imaging (MPI). The aim of this study was to assess the relation of stress-induced dyssynchrony and the extent of angiographic coronary artery disease (CAD). METHODS AND RESULTS: The MPI images of 144 patients were retrospectively analyzed. With ≥70% stenosis as the criteria of CAD, 57 had no CAD, 32 had 1-vessel disease, 36 had 2-vessel disease, and 19 had 3-vessel disease, respectively. LV global and territorial dyssynchrony parameters were measured by the phase analysis from stress/rest Tl-201 SPECT MPI and compared between stress and rest among the patient groups. The patients with multi-vessel CAD had significantly more global dyssynchrony than the patients without ≥70% stenosis at stress. The patients with multi-vessel CAD showed significantly more global and territorial dyssynchrony on stress images than on rest. More patients with 3-vessel CAD were correctly classified as multi-vessel disease, when combining both visual interpretation and dyssynchrony assessment. CONCLUSION: The patients with multi-vessel CAD had significantly more global and territorial dyssynchrony at early post-stress than at rest. Such quantitative measures of myocardial stunning may assist in the diagnosis of multi-vessel CAD.

Feasibility and operator variability of myocardial blood flow and reserve measurements with 99mTc-sestamibi quantitative dynamic SPECT/CT imaging.

PURPOSE: Myocardial blood flow (MBF) quantification with dynamic SPECT could lead to widespread utilization of MBF imaging in clinical practice with little cost increase over current standard SPECT myocardial perfusion imaging. This work evaluates the feasibility and operator-dependent variability of MBF and flow reserve measurements with (99m)Tc-sestamibi (MIBI) dynamic SPECT imaging using a standard dual-head SPECT camera. METHODS: Twenty-eight patients underwent dipyridamole-stress and rest imaging with dynamic SPECT/CT acquisition. Quantitative images were iteratively reconstructed with all physical corrections and then myocardial and arterial blood regions of interest (ROI) were defined semi-automatically. A compartmental model was fitted to these ROI-sampled time-activity-curves, and flow-dependent MIBI extraction correction was applied to derive regional MBF values. Myocardial flow reserve (MFR) was estimated as stress/rest MBF ratio. MBF and MFR in low and high risk populations were evaluated for ability to detect disease. Images were each processed twice (≥7 days apart) by one expert and one novice operator to evaluate intra- and inter-operator variability of MBF and MFR measurement in the three coronary artery vascular territories. RESULTS: Mean rest flow, stress flow, and MFR values were 0.83, 1.82 mL·minute(-1)·g(-1), and 2.45, respectively. For stress/rest MFR, the inter-operator reproducibility was r(2) = 0.86 with RPC = 1.1. Stress MBF and MFR were significantly reduced (P < .05) in high risk (n = 9) vs low risk populations (n = 19), indicating ability to detect disease. For expert and novice operators very good intra-operator correlations of r(2) = 0.98 and 0.95 (n = 168, P < .001) were observed for combined rest and stress regional flow values. Bland-Altman reproducibility coefficients (RPC) were 0.25 and 0.47 mL·minute(-1)·g(-1) for the expert and novice operators, respectively (P < .001). Inter-operator correlation was r(2) = 0.91 and Bland-Altman RPC = 0.58 mL·minute(-1)·g(-1) (n = 336). CONCLUSIONS: MBF and reserve measurements using (99m)Tc-sestamibi on a traditional, two-headed camera with fast rotation and with quantitative dynamic SPECT appears to be feasible, warranting further investigation.

A new method of modeling the multi-stage decision-making process of CRT using machine learning with uncertainty quantification.

Aims: Current machine learning-based (ML) models usually attempt to utilize all available patient data to predict patient outcomes while ignoring the associated cost and time for data acquisition. The purpose of this study is to create a multi-stage machine learning model to predict cardiac resynchronization therapy (CRT) response for heart failure (HF) patients. This model exploits uncertainty quantification to recommend additional collection of single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) variables if baseline clinical variables and features from electrocardiogram (ECG) are not sufficient. Methods: 218 patients who underwent rest-gated SPECT MPI were enrolled in this study. CRT response was defined as an increase in left ventricular ejection fraction (LVEF) > 5% at a 6±1 month follow-up. A multi-stage ML model was created by combining two ensemble models: Ensemble 1 was trained with clinical variables and ECG; Ensemble 2 included Ensemble 1 plus SPECT MPI features. Uncertainty quantification from Ensemble 1 allowed for multi-stage decision-making to determine if the acquisition of SPECT data for a patient is necessary. The performance of the multi-stage model was compared with that of Ensemble models 1 and 2. Results: The response rate for CRT was 55.5% (n = 121) with overall male gender 61.0% (n = 133), an average age of 62.0±11.8, and LVEF of 27.7±11.0. The multi-stage model performed similarly to Ensemble 2 (which utilized the additional SPECT data) with AUC of 0.75 vs. 0.77, accuracy of 0.71 vs. 0.69, sensitivity of 0.70 vs. 0.72, and specificity 0.72 vs. 0.65, respectively. However, the multi-stage model only required SPECT MPI data for 52.7% of the patients across all folds. Conclusions: By using rule-based logic stemming from uncertainty quantification, the multi-stage model was able to reduce the need for additional SPECT MPI data acquisition without sacrificing performance.

Classification Prediction of Alzheimer's Disease and Vascular Dementia Using Physiological Data and ECD SPECT Images.

Alzheimer's disease (AD) and vascular dementia (VaD) are the two most common forms of dementia. However, their neuropsychological and pathological features often overlap, making it difficult to distinguish between AD and VaD. In addition to clinical consultation and laboratory examinations, clinical dementia diagnosis in Taiwan will also include Tc-99m-ECD SPECT imaging examination. Through machine learning and deep learning technology, we explored the feasibility of using the above clinical practice data to distinguish AD and VaD. We used the physiological data (33 features) and Tc-99m-ECD SPECT images of 112 AD patients and 85 VaD patients in the Taiwanese Nuclear Medicine Brain Image Database to train the classification model. The results, after filtering by the number of SVM RFE 5-fold features, show that the average accuracy of physiological data in distinguishing AD/VaD is 81.22% and the AUC is 0.836; the average accuracy of training images using the Inception V3 model is 85% and the AUC is 0.95. Finally, Grad-CAM heatmap was used to visualize the areas of concern of the model and compared with the SPM analysis method to further understand the differences. This research method can quickly use machine learning and deep learning models to automatically extract image features based on a small amount of general clinical data to objectively distinguish AD and VaD.