Machine Learning Algorithms at Myocardial Perfusion Imaging - a Preliminary Study

Introduction: A major focus of data mining process - especially machine learning researches - is to automatically learn to recognize complex patterns and help to take the adequate decisions strictly based on the acquired data. Since imaging techniques like MPI – Myocardial Perfusion Imaging on Nuclear Cardiology, can implicate a huge part of the daily workflow and generate gigabytes of data, there could be advantages on Computerized Analysis of data over Human Analysis: shorter time, homogeneity and consistency, automatic recording of analysis results, relatively inexpensive, etc.Objectives: The aim of this study relates with the evaluation of the efficacy of this methodology on the evaluation of MPI Stress studies and the process of decision taking concerning the continuation – or not – of the evaluation of each patient. It has been pursued has an objective to automatically classify a patient test in one of three groups: “Positive”, “Negative” and “Indeterminate”. “Positive” would directly follow to the Rest test part of the exam, the “Negative” would be directly exempted from continuation and only the “Indeterminate” group would deserve the clinician analysis, so allowing economy of clinician’s effort, increasing workflow fluidity at the technologist’s level and probably sparing time to patients. Methods: WEKA v3.6.2 open source software was used to make a comparative analysis of three WEKA algorithms (“OneR”, “J48” and “Naïve Bayes”) - on a retrospective study using the comparison with correspondent clinical results as reference, signed by nuclear cardiologist experts - on “SPECT Heart Dataset”, available on University of California – Irvine, at the Machine Learning Repository. For evaluation purposes, criteria as “Precision”, “Incorrectly Classified Instances” and “Receiver Operating Characteristics (ROC) Areas” were considered. Results: The interpretation of the data suggests that the Naïve Bayes algorithm has the best performance among the three previously selected algorithms. Conclusions: It is believed - and apparently supported by the findings - that machine learning algorithms could significantly assist, at an intermediary level, on the analysis of scintigraphic data obtained on MPI, namely after Stress acquisition, so eventually increasing efficiency of the entire system and potentially easing both roles of Technologists and Nuclear Cardiologists. In the actual continuation of this study, it is planned to use more patient information and significantly increase the population under study, in order to allow improving system accuracy.

Radiopharmaceuticals for myocardial perfusion imaging: SPECT & PET

Myocardial perfusion imaging (MPI) is used on a daily basis to access coronary blood flow in patients that are suspected or have known Coronary Artery Disease (CAD). A Single Photon Emission Computed Tomography (SPECT) or and Positron Emission Tomography (PET) scan are used to access regional blood flow quantification either at rest or stress, the imaging acquisition is connected to an Electrocardiogram (ECG) and it is able to determine and quantify other myocardial parameters like myocardial wall thickness and wall motion. PET is not used so broadly due to its high procedure cost, the proximity with cyclotron, where are produced the majority of radiopharmaceuticals used in PET, due to their shor thalf-life. This work is intended to carry out a review of the tests relating to radiopharmaceuticals that are used in clinical practice in SPECT or PET for assessment of myocardial perfusion, also focusing very promising radiopharmaceuticals that are under investigation or in clinical trials with great potential for conventional nuclear medicine or PET, proceeding to a comparative analysis of both techniques and respective radiopharmaceuticals used.

Detection of the viable myocardium. A perfusion scintigraphic study, before and after coronary bypass surgery in myocardial infarction patients

OBJECTIVE: To compare single-photon-emission computed tomography (SPECT) imaging scans using 201Tl and 99mTc-MIBI in detection of viable myocardium, in regions compromised by infarction. METHODS: Thirty-two (59.3±9.8 years old and 87% male) myocardial infarction patients were studied. All had Q waves on the ECG and left ventricle ejection fraction of <50%. They underwent coronary and left ventricle angiographies and SPECT before (including 201Tl reinjection) and after coronary artery bypass surgery (CABG). Improvement in perfusion observed after surgery was considered the gold standard for myocardial viability. RESULTS: Among 102 studied regions of the heart, there were 40 (39.2%) areas of transient perfusion defects in the conventional protocol with 201Tl and 52 (51.0%) after reinjection. Therefore, 12/62 (19.4%) more viable regions were identified by reinjection. Using 99mTc-MIBI, only 14 (13.7%) regions with transient defects were identified, all of which were seen also in 201Tl protocols. After surgery, 49 of a total of 93 regions analyzed (52.7%) were viable. Sensitivity, specificity, accuracy, positive and negative prediction values were, respectively, 201Tl SPECT scans - 65.3%, 90.9%, 77.4%, 88.9% and 70.2%, reinjection protocol with 201Tl scans - 81.6%, 81.8%, 81.7%, 83.3% and 80.0%; 99mTc-MIBI SPECT scans - 20.4%, 90.9%, 53.8%, 71.4% and 50.6%. Logistic regression demonstrated that the reinjection protocol with 201Tl was the best predictor of viability (P<0.001). CONCLUSION: Our data suggest the election of 201Tl for viability studies, especially when using the reinjection protocol.

BNP was Associated with Ischemic Myocardial Scintigraphy and Death in Patients at Chest Pain Unit

Background:Recent studies have suggested that B-type Natriuretic Peptide (BNP) is an important predictor of ischemia and death in patients with suspected acute coronary syndrome. Increased levels of BNP are seen after episodes of myocardial ischemia and may be related to future adverse events.Objectives:To determine the prognostic value of BNP for major cardiac events and to evaluate its association with ischemic myocardial perfusion scintigraphy (MPS).Methods:This study included retrospectively 125 patients admitted to the chest pain unit between 2002 and 2006, who had their BNP levels measured on admission and underwent CPM for risk stratification. BNP values were compared with the results of the MPS. The chi-square test was used for qualitative variables and the Student t test, for quantitative variables. Survival curves were adjusted using the Kaplan-Meier method and analyzed by using Cox regression. The significance level was 5%.Results:The mean age was 63.9 ± 13.8 years, and the male sex represented 51.2% of the sample. Ischemia was found in 44% of the MPS. The mean BNP level was higher in patients with ischemia compared to patients with non-ischemic MPS (188.3 ± 208.7 versus 131.8 ± 88.6; p = 0.003). A BNP level greater than 80 pg/mL was the strongest predictor of ischemia on MPS (sensitivity = 60%, specificity = 70%, accuracy = 66%, PPV = 61%, NPV = 70%), and could predict medium-term mortality (RR = 7.29, 95% CI: 0.90-58.6; p = 0.045) independently of the presence of ischemia.Conclusions:BNP levels are associated with ischemic MPS findings and adverse prognosis in patients presenting with acute chest pain to the emergency room, thus, providing important prognostic information for an unfavorable clinical outcome.

The Benefits of Prone SPECT Myocardial Perfusion Imaging in Reducing Both Artifact Defects and Patient Radiation Exposure

AbstractBackground:Prone imaging has been demonstrated to minimize diaphragmatic and breast tissue attenuation.Objectives:To determine the role of prone imaging on the reduction of unnecessary rest perfusion studies and coronary angiographies performed, thus decreasing investigation time and radiation exposure.Methods:We examined 139 patients, 120 with an inferior wall and 19 with an anterior wall perfusion defect that might represented attenuation artifact. Post-stress images were acquired in both the supine and prone position. Coronary angiography was used as the “gold standard” for evaluating coronary artery patency. The study was terminated and rest imaging was obviated in the presence of complete improvement of the defect in the prone position. Quantitative interpretation was performed. Results were compared with clinical data and coronary angiographic findings.Results:Prone acquisition correctly revealed defect improvement in 89 patients (89/120) with inferior wall and 12 patients (12/19) with anterior wall attenuation artifact. Quantitative analysis demonstrated statistically significant difference in the mean summed stress scores (SSS) of supine and mean SSS of prone studies in patients with disappearing inferior wall defect in the prone position and patent right coronary artery (true negative results). The mean difference between SSS in supine and in prone position was higher with disappearing than with remaining defects.Conclusion:Technetium-99m (Tc-99m) tetrofosmin myocardial perfusion imaging with the patient in the prone position overcomes soft tissue attenuation; moreover it provides an inexpensive, accurate approach to limit the number of unnecessary rest perfusion studies and coronary angiographies performed.

Quantification of Myocardial Perfusion Using 13N-ammonia PET: a Cross-Comparison Study on Analysis Software Packages - Carimas, PMOD and FlowQuant

Aim. Several software packages (SWP) and models have been released for quantification of myocardial perfusion (MP). Although they all are validated against something, the question remains how well their values agree. The present analysis focused on cross-comparison of three SWP for MP quantification of 13N-ammonia PET studies. Materials & Methods. 48 rest and stress MP 13N-ammonia PET studies of hypertrophic cardiomyopathy (HCM) patients (Sciagrà et al., 2009) were analysed with three SW packages - Carimas, PMOD, and FlowQuant - by three observers blinded to the results of each other. All SWP implement the one-tissue-compartment model (1TCM, DeGrado et al. 1996), and first two - the two-tissue-compartment model (2TCM, Hutchins et al. 1990) as well. Linear mixed model for the repeated measures was fitted to the data. Where appropriate we used Bland-Altman plots as well. The reproducibility was assessed on global, regional and segmental levels. Intraclass correlation coefficients (ICC), differences between the SWPs and between models were obtained. ICC≥0.75 indicated excellent reproducibility, 0.4≤ICC<0.75 indicated fair to good reproducibility, ICC<0.4 - poor reproducibility (Rosner, 2010). Results. When 1TCM MP values were compared, the SW agreement on global and regional levels was excellent, except for Carimas vs. PMOD at RCA: ICC=0.715 and for PMOD vs. FlowQuant at LCX:ICC=0.745 which were good. In segmental analysis in five segments: 7,12,13, 16, and 17 the agreement between all SWP was excellent; in the remaining 12 segments the agreement varied between the compared SWP. Carimas showed excellent agreement with FlowQuant in 13 segments and good in four - 1, 5, 6, 11: 0.687≤ICCs≤0.73; Carimas had excellent agreement with PMOD in 11 segments, good in five_4, 9, 10, 14, 15: 0.682≤ICCs≤0.737, and poor in segment 3: ICC=0.341. PMOD had excellent agreement with FlowQuant in eight segments and substantial-to-good in nine_1, 2, 3, 5, 6,8-11: 0.585≤ICCs≤0.738. Agreement between Carimas and PMOD for 2TCM was good at a global level: ICC=0.745, excellent at LCX (0.780) and RCA (0.774), good at LAD (0.662); agreement was excellent for ten segments, fair-to-substantial for segments 2, 3, 8, 14, 15 (0.431≤ICCs≤0.681), poor for segments 4 (0.384) and 17 (0.278). Conclusions. The three SWP used by different operators to analyse 13N-ammonia PET MP studies provide results that agree well at a global level, regional levels, and mostly well even at a segmental level. Agreement is better for 1TCM. Poor agreement at segments 4 and 17 for 2TCM needs further clarification.

Cardiac Imaging in the Diagnosis of Coronary Artery Disease: The Value of Quantitative Imaging of Myocardial Perfusion and Deformation

Atherosclerosis is a chronic and progressive disease of the vasculature. Increasing coronary atherosclerosis can lead to obstructive coronary artery disease (CAD) or myocardial infarction. Computed tomography angiography (CTA) allows noninvasive assessment of coronary anatomy and quantitation of atherosclerotic burden. Myocardial blood flow (MBF) can be accurately measured in absolute terms (mL/g/min) by positron emission tomography (PET) with [15O] H O as a radiotracer. We studied the coronary microvascular dysfunction as a risk factor for future coronary calcification in healthy young men by measuring the coronary flow reserve (CFR) which is the ratio between resting and hyperemic MBF. Impaired vasodilator function was not linked with accelerated atherosclerosis 11 years later. Currently, there is a global interest in quantitative PET perfusion imaging. We established optimal thresholds of [15O] H O PET perfusion for diagnosis of CAD (hyperemic MBF of 2.3 mL/g/min and CFR of 2.5) in the first multicenter study of this type (Turku, Amsterdam and Uppsala). In myocardial bridging a segment of the coronary artery travels inside the myocardium and can be seen as intramural course (CTA) or systolic compression (invasive coronary angiography). Myocardial bridging is frequently linked with proximal atherosclerotic plaques. We used quantitative [15O] H O PET perfusion to evaluate the hemodynamic effects of myocardial bridging. Myocardial bridging was not associated with decreased absolute MBF or increased atherosclerotic burden. Speckle tracking allows quantitative echocardiographic imaging of myocardial deformation. Speckle tracking during dobutamine stress echocardiography was feasible and comparable to subjective wall motion analysis in the diagnosis of CAD. In addition, it correctly risk stratified patients with multivessel disease and extensive ischemia.

Quantitative coronary arterial stenosis assessment by multidetector CT and invasive coronary angiography for identifying patients with myocardial perfusion abnormalities

Semi-quantitative stenosis assessment by coronary CT angiography only modestly predicts stress-induced myocardial perfusion abnormalities. The performance of quantitative CT angiography (QCTA) for identifying patients with myocardial perfusion defects remains unclear. CorE-64 is a multicenter, international study to assess the accuracy of 64-slice QCTA for detecting a parts per thousand yen50% coronary arterial stenoses by quantitative coronary angiography (QCA). Patients referred for cardiac catheterization with suspected or known coronary artery disease were enrolled. Area under the receiver-operating-characteristic curve (AUC) was used to evaluate the diagnostic accuracy of the most severe coronary artery stenosis in a subset of 63 patients assessed by QCTA and QCA for detecting myocardial perfusion abnormalities on exercise or pharmacologic stress SPECT. Diagnostic accuracy of QCTA for identifying patients with myocardial perfusion abnormalities by SPECT revealed an AUC of 0.71, compared to 0.72 by QCA (P = .75). AUC did not improve after excluding studies with fixed myocardial perfusion abnormalities and total coronary arterial occlusions. Optimal stenosis threshold for QCTA was 43% yielding a sensitivity of 0.81 and specificity of 0.50, respectively, compared to 0.75 and 0.69 by QCA at a threshold of 59%. Sensitivity and specificity of QCTA to identify patients with both obstructive lesions and myocardial perfusion defects were 0.94 and 0.77, respectively. Coronary artery stenosis assessment by QCTA or QCA only modestly predicts the presence and the absence of myocardial perfusion abnormalities by SPECT. Confounding variables affecting the relationship between coronary anatomy and myocardial perfusion likely account for some of the observed discrepancies between coronary angiography and SPECT results.

Functionally relevant coronary artery disease: comparison of 64-section CT angiography with myocardial perfusion SPECT

PURPOSE: To prospectively determine the accuracy of 64-section computed tomographic (CT) angiography for the depiction of coronary artery disease (CAD) that induces perfusion defects at myocardial perfusion imaging with single photon emission computed tomography (SPECT), by using myocardial perfusion imaging as the reference standard. MATERIALS AND METHODS: All patients gave written informed consent after the study details, including radiation exposure, were explained. The study protocol was approved by the local institutional review board. In patients referred for elective conventional coronary angiography, an additional 64-section CT angiography study and a myocardial perfusion imaging study (1-day adenosine stress-rest protocol) with technetium 99m-tetrofosmin SPECT were performed before conventional angiography. Coronary artery diameter narrowing of 50% or greater at CT angiography was defined as stenosis and was compared with the myocardial perfusion imaging findings. Quantitative coronary angiography served as a reference standard for CT angiography. RESULTS: A total of 1093 coronary segments in 310 coronary arteries in 78 patients (mean age, 65 years +/- 9 [standard deviation]; 35 women) were analyzed. CT angiography revealed stenoses in 137 segments (13%) corresponding to 91 arteries (29%) in 46 patients (59%). SPECT revealed 14 reversible, 13 fixed, and six partially reversible defects in 31 patients (40%). Sensitivity, specificity, and negative and positive predictive values, respectively, of CT angiography in the detection of reversible myocardial perfusion imaging defects were 95%, 53%, 94%, and 58% on a per-patient basis and 95%, 75%, 96%, and 72% on a per-artery basis. Agreement between CT and conventional angiography was very good (96% and kappa = 0.92 for patient-based analysis, 93% and kappa = 0.84 for vessel-based analysis). CONCLUSION: Sixty-four-section CT angiography can help rule out hemodynamically relevant CAD in patients with intermediate to high pretest likelihood, although an abnormal CT angiography study is a poor predictor of ischemia.

Partial volume correction incorporating Rb-82 positron range for quantitative myocardial perfusion PET based on systolic-diastolic activity ratios and phantom measurements.

BACKGROUND: Quantitative myocardial PET perfusion imaging requires partial volume corrections. METHODS: Patients underwent ECG-gated, rest-dipyridamole, myocardial perfusion PET using Rb-82 decay corrected in Bq/cc for diastolic, systolic, and combined whole cycle ungated images. Diastolic partial volume correction relative to systole was determined from the systolic/diastolic activity ratio, systolic partial volume correction from phantom dimensions comparable to systolic LV wall thicknesses and whole heart cycle partial volume correction for ungated images from fractional systolic-diastolic duration for systolic and diastolic partial volume corrections. RESULTS: For 264 PET perfusion images from 159 patients (105 rest-stress image pairs, 54 individual rest or stress images), average resting diastolic partial volume correction relative to systole was 1.14 ± 0.04, independent of heart rate and within ±1.8% of stress images (1.16 ± 0.04). Diastolic partial volume corrections combined with those for phantom dimensions comparable to systolic LV wall thickness gave an average whole heart cycle partial volume correction for ungated images of 1.23 for Rb-82 compared to 1.14 if positron range were negligible as for F-18. CONCLUSION: Quantitative myocardial PET perfusion imaging requires partial volume correction, herein demonstrated clinically from systolic/diastolic absolute activity ratios combined with phantom data accounting for Rb-82 positron range.

Relationship between myocardial perfusion and dysfunction in diabetic cardiomyopathy: A study of quantitative contrast echocardiography and strain rate imaging

Objective: To use quantitative myocardial contrast echocardiography (MCE) and strain rate imaging (SRI) to assess the role of microvascular disease in subclinical diabetic cardiomyopathy. Methods: Stress MCE and SRI were performed in 48 patients (22 with type II diabetes mellitus (DM) and 26 controls), all with normal left ventricular systolic function and no obstructive coronary disease by quantitative coronary angiography. Real-time MCE was acquired in three apical views at rest and after combined dipyridamole-exercise stress. Myocardial blood flow (MBF) was quantified in the 10 mid- and apical cardiac segments at rest and after stress. Resting peak systolic strain rate (SR) and peak systolic strain (epsilon) were calculated in the same 10 myocardial segments. Results: The DM and control groups were matched for age, sex and other risk factors, including hypertension. The DM group had higher body mass index and left ventricular mass index. Quantitative SRI analysis was possible in all patients and quantitative MCE in 46 (96%). The mean e, SR and MBF reserve were all significantly lower in the DM group than in controls, with diabetes the only independent predictor of each parameter. No correlation was seen between MBF and SR (r = -0.01, p = 0.54) or between MBF and epsilon ( r = -0.20, p = 0.20). Conclusions: Quantitative MCE shows that patients with diabetes but no evidence of obstructive coronary artery disease have impaired MBF reserve, but abnormal transmural flow and subclinical longitudinal myocardial dysfunction are not related.

Early Increase in Myocardial Perfusion After Stem Cell Therapy in Patients Undergoing Incomplete Coronary Artery Bypass Surgery

Incomplete revascularization is associated with worse long-term outcomes. Autologous bone marrow cells (BMC) have recently been tested in patients with severe coronary artery disease. We tested the hypothesis that intramyocardial injection of autologous BMC increases myocardial perfusion in patients undergoing incomplete coronary artery bypass grafting (CABG). Twenty-one patients (19 men), 59 +/- 7 years old, with limiting angina and multivessel coronary artery disease (CAD), not amenable to complete CABG were enrolled. BMC were obtained prior to surgery, and the lymphomonocytic fraction separated by density gradient centrifugation. During surgery, 5 mL containing 2.1 +/- 1.3 x 10(8) BMC (CD34+ = 0.8 +/- 0.3%) were injected in the ischemic non-revascularized myocardium. Myocardial perfusion was assessed by magnetic resonance imaging (MRI) at baseline and 1 month after surgery. The increase in myocardial perfusion was compared between patients with < 50% (group A, n = 11) with that of patients with > 50% (group B, n = 10) of target vessels (stenosis a parts per thousand yenaEuro parts per thousand 70%) successfully bypassed. Injected myocardial segments included the inferior (n = 12), anterior (n = 7), and lateral (n = 2) walls. The number of treated vessels (2.3 +/- 0.8) was significantly smaller than the number of target vessels (4.2 +/- 1.0; P < 0.0001). One month after surgery, cardiac MRI showed a similar reduction (%) in the ischemic score of patients in group A (72.5 +/- 3.2), compared to patients in group B (78.1 +/- 3.2; P = .80). Intramyocardial injection of autologous BMC may help increase myocardial perfusion in patients undergoing incomplete CABG, even in those with fewer target vessels successfully treated. This strategy may be an adjunctive therapy for patients suffering from a more advanced (diffuse) CAD not amenable for complete direct revascularization.

Myocardial Perfusion Imaging Is a Strong Predictor of Death in Women

OBJECTIVES We sought to assess the prognostic value and risk classification improvement using contemporary single-photon emission computed tomography myocardial perfusion imaging (SPECT-MPI) to predict all-cause mortality. BACKGROUND Myocardial perfusion is a strong estimator of prognosis. Evidence published to date has not established the added prognostic value of SPECT-MPI nor defined an approach to detect improve classification of risk in women from a developing nation. METHODS A total of 2,225 women referred for SPECT-MPI were followed by a mean period of 3.7 +/- 1.4 years. SPECT-MPI results were classified as abnormal on the presence of any perfusion defect. Abnormal scans were further classified as with mild/moderate reversible, severe reversible, partial reversible, or fixed perfusion defects. Risk estimates for incident mortality were categorized as <1%/year, 1% to 2%/year, and >2%/year using Cox proportional hazard models. Risk-adjusted models incorporated clinical risk factors, left ventricular ejection fraction (LVEF), and perfusion variables. RESULTS All-cause death occurred in 139 patients. SPECT-MPI significantly risk stratified the population; patients with abnormal scans had significantly higher death rates compared with patients with normal scans, 13.1% versus 4.0%, respectively (p < 0.001). Cox analysis demonstrated that after adjusting for clinical risk factors and LVEF, SPECT-MPI improved the model discrimination (integrated discrimination index = 0.009; p = 0.02), added significant incremental prognostic information (global chi-square increased from 87.7 to 127.1; p < 0.0001), and improved risk prediction (net reclassification improvement = 0.12; p = 0.005). CONCLUSIONS SPECT-MPI added significant incremental prognostic information to clinical and left ventricular functional variables while enhancing the ability to classify this Brazilian female population into low-and high-risk categories of all-cause mortality. (J Am Coll Cardiol Img 2011;4:880-8) (C) 2011 by the American College of Cardiology Foundation

Analysing the usefulness of motion correction software in myocardial perfusion imaging

Introduction Myocardial Perfusion Imaging (MPI) is a very important tool in the assessment of Coronary Artery Disease ( CAD ) patient s and worldwide data demonstrate an increasingly wider use and clinical acceptance. Nevertheless, it is a complex process and it is quite vulnerable concerning the amount and type of possible artefacts, some of them affecting seriously the overall quality and the clinical utility of the obtained data. One of the most in convenient artefacts , but relatively frequent ( 20% of the cases ) , is relate d with patient motion during image acquisition . Mostly, in those situations, specific data is evaluated and a decisi on is made between A) accept the results as they are , consider ing that t he “noise” so introduced does not affect too seriously the final clinical information, or B) to repeat the acquisition process . Another possib ility could be to use the “ Motion Correcti on Software” provided within the software package included in any actual gamma camera. The aim of this study is to compare the quality of the final images , obtained after the application of motion correction software and after the repetition of image acqui sition. Material and Methods Thirty cases of MPI affected by Motion Artefacts and repeated , were used. A group of three, independent (blinded for the differences of origin) expert Nuclear Medicine Clinicians had been invited to evaluate the 30 sets of thre e images - one set for each patient - being ( A) original image , motion uncorrected , (B) original image, motion corrected, and (C) second acquisition image, without motion . The results so obtained were statistically analysed . Results and Conclusion Results obtained demonstrate that the use of the Motion Correction Software is useful essentiall y if the amplitude of movement is not too important (with this specific quantification found hard to define precisely , due to discrepancies between clinicians and other factors , namely between one to another brand); when that is not the case and the amplitude of movement is too important , the n the percentage of agreement between clinicians is much higher and the repetition of the examination is unanimously considered ind ispensable.