4D display of CT LV endocardial and epicardial models morphed from PET Rb-82 perfusion studies accurately quantifies segmental myocardial thickening.

BACKGROUND: MPI-derived LV wall thickening assessments for diagnostic purposes has been part of clinical guidelines for two decades. It relies on visual evaluation of tomographic slices or regional quantification displayed in 2D polar maps. 4D displays have not entered clinical usage nor have they been validated on their potential to provide equivalent information. The purpose of this work was to validate a 4D realistic display recently designed to quantitatively represent the thickening information from gated MPI into CT-morphed endocardial and epicardial moving surfaces. METHODS: Forty patients who underwent 82Rb PET were selected based on LV perfusion quantification. CTA templates of heart anatomy were selected to represent the LV anatomy. Generic CT-derived LV endocardial and epicardial surfaces were modified to represent the end diastolic (ED) phase according to PET-derived ED LV dimensions and wall thickness. These CT myocardial surfaces were then morphed by means of thin plate spline (TPS) techniques, according to the gated PET slices count changes (WThPET) and LV wall motion (WMoPET). A geometric thickening (GeoTh) equivalent to LV WThPET was defined on epicardial and endocardial CT surfaces over the cardiac cycle and the two measures compared. WThPET and GeoTh correlations were performed on a case-by-case basis, by segment and by pooling all 17 segments. Pearson's correlation coefficients (PCC) were calculated to assess the equivalence of the two measures. RESULTS: Two cohorts of patients (normal and abnormal) were identified based on SSS. R coefficients were as follows: for all pooled segments PCCstress and PCCrest were respectively 0.91 and 0.89 (normal), and 0.9 and 0.91 (abnormal); when individual 17 segments were considered mean PCCstress = 0.92 [0.81-0.98] and mean PCCrest = 0.93 [0.83-0.98] for the abnormal perfusion group; mean PCCstress = 0.89 [0.78-0.97] and mean PCCrest = 0.89 [0.77-0.97] for the normal. When individual studies were considered, R was always > .70 with the exception of five abnormal studies. Inter-user analysis was also conducted. CONCLUSIONS: Our novel technique for the visualization of LV wall thickening by means of 4D CT endocardial and epicardial surface models accurately replicated 82Rb slice thickening results showing promise for its usage for diagnostic purposes.

Determination of [N-13]-ammonia extraction fraction in patients with coronary artery disease by calibration to invasive coronary and fractional flow reserve.

BACKGROUND: This study presents a new extraction fraction (EF) model based on physiological measures of invasive coronary flow reserve (CFR) and fractional flow reserve (FFR) in patients with suspected coronary artery disease (CAD) and normal index microcirculatory resistance (IMR). To ascertain the clinical relevance of the new EFs, flow measurements using the newly patient-determined EFs were compared to flow measurements using traditional animal-determined EFs. METHODS: 39 patients were retrospectively selected that included a total of 91 vascular territories with invasive coronary angiography physiological measures. [N-13]-ammonia dynamic rest/adenosine-stress PET imaging was conducted in all patients and absolute myocardial flow was estimated using four published compartmental models. The extraction fraction during hyperemic flow was iteratively estimated by maximizing the agreement between invasive CFR and FFR with the non-invasive analogs myocardial flow reserve (MFR) and relative flow reserve (RFR) at similar physiological states, respectively. RESULTS: Using the new patient-determined EFs, agreement between CFR vs MFR for Model 1 and 2 was moderate and poor for Model 3 and 4. All models showed moderate agreement for FFR vs RFR. When using published models of animal-determined EFs, agreement between CFR vs MFR remained moderate for Model 1 and 2, and poor for Model 3 and 4. Similarly, all models showed moderate agreement for FFR vs RFR using animal-determined EF values. None of the observed differences were statistically significant. CONCLUSIONS: Flow measurements using extraction fraction correction for [N-13]-ammonia based on calibration to invasive intracoronary angiography physiological measures in patients with CAD were not discordant from those reported in the literature. Either patient-determined or traditional animal-determined EF correction, when used with the appropriate flow model, yields moderate agreement with invasive measurements of coronary flow reserve and fractional flow reserve.

Diagnostic performance of an artificial intelligence-driven cardiac-structured reporting system for myocardial perfusion SPECT imaging.

OBJECTIVES: To describe and validate an artificial intelligence (AI)-driven structured reporting system by direct comparison of automatically generated reports to results from actual clinical reports generated by nuclear cardiology experts. BACKGROUND: Quantitative parameters extracted from myocardial perfusion imaging (MPI) studies are used by our AI reporting system to generate automatically a guideline-compliant structured report (sR). METHOD: A new nonparametric approach generates distribution functions of rest and stress, perfusion, and thickening, for each of 17 left ventricle segments that are then transformed to certainty factors (CFs) that a segment is hypoperfused, ischemic. These CFs are then input to our set of heuristic rules used to reach diagnostic findings and impressions propagated into a sR referred as an AI-driven structured report (AIsR). The diagnostic accuracy of the AIsR for detecting coronary artery disease (CAD) and ischemia was tested in 1,000 patients who had undergone rest/stress SPECT MPI. RESULTS: At the high-specificity (SP) level, in a subset of 100 patients, there were no statistical differences in the agreements between the AIsr, and nine experts' impressions of CAD (P = .33) or ischemia (P = .37). This high-SP level also yielded the highest accuracy across global and regional results in the 1,000 patients. These accuracies were statistically significantly better than the other two levels [sensitivity (SN)/SP tradeoff, high SN] across all comparisons. CONCLUSIONS: This AI reporting system automatically generates a structured natural language report with a diagnostic performance comparable to those of experts.

Development and validation of an automatic method to detect the latest contracting viable left ventricular segments to assist guide CRT therapy from gated SPECT myocardial perfusion imaging.

OBJECTIVES: The purpose of this study is to use ECG-gated SPECT MPI to detect the latest contracting viable left ventricular (LV) segments to help guide the LV probe placement used in CRT therapy and to validate segment selection against the visual integration method by experts. METHODS: For each patient, the resting ECG-gated SPECT MPI short-axis images were sampled in 3D to generate a polar map of the perfusion distribution used to determine LV myocardial viability, and to measure LV synchronicity using our phase analysis tool. In the visual integration method, two experts visually interpreted the LV viability and mechanical dyssynchrony from the short-axis images and polar maps of viability and phase, to determine the latest contracting viable segments using the 17-segment model. In the automatic method, the apical segments, septal segments, and segments with more than 50% scar were excluded as these are not candidates for CRT LV probe placement. Amongst the remaining viable segments, the segments, whose phase angles were within 10° of the latest phase angle (the most delayed contracting segment), were identified for potential CRT LV probe placement and ranked based on the phase angles of the segments. Both methods were tested in 36 pre-CRT patients who underwent ECG-gated SPECT MPI. The accuracy was determined as the percent agreement between the visual integration and automatic methods. The automatic method was performed by a second independent operator to evaluate the inter-operator processing reproducibility. RESULTS: In all the 36 patients, the LV lead positions of the 1st choices recommended by the automatic and visual integration methods were in the same segments in 35 patients, which achieved an agreement rate of 97.2%. In the inter-operator reproducibility test, the LV lead positions of the 1st choices recommended by the two operators were in the same segments in 25 patients, and were in the adjacent segments in 7 patients, which achieved an overall agreement of 88.8%. CONCLUSIONS: An automatic method has been developed to detect the latest contracting viable LV segments to help guide the LV probe placement used in CRT therapy. The retrospective clinical study with 36 patients suggests that this method has high agreement against the visual integration method by experts and good inter-operator reproducibility. Consequently, this method is promising to be a clinical tool to recommend the CRT LV lead positions.

Diagnostic performance of the quantification of myocardium at risk from MPI SPECT/CTA 2G fusion for detecting obstructive coronary disease: A multicenter trial.

BACKGROUND: The effective non-invasive identification of coronary artery disease (CAD) and its proper referral for invasive treatment are still unresolved issues. We evaluated our quantification of myocardium at risk (MAR) from our second generation 3D MPI/CTA fusion framework for the detection and localization of obstructive coronary disease. METHODS: Studies from 48 patients who had rest/stress MPI, CTA, and ICA were analyzed from 3 different institutions. From the CTA, a 3D biventricular surface of the myocardium with superimposed coronaries was extracted and fused to the perfusion distribution. Significant lesions were identified from CTA readings and positioned on the fused display. Three estimates of MAR were computed on the 3D LV surface on the basis of the MPI alone (MARp), the CTA alone (MARa), and the fused information (MARf). The extents of areas at risk were used to generate ROC curves using ICA anatomical findings as reference standard. RESULTS: Areas under the ROC curve (AUC) for CAD detection using MARf was 0.88 (CI = 0.75-0.95) and for MARp and MARa were, respectively 0.82 (CI = 0.69-0.92) and 0.75 (CI = 0.60-0.86) using the ≥70% stenosis criterion. AUCs for CAD localization (all vessels) using MARf showed significantly higher performance than either MARa or MARp or both. CONCLUSIONS: Using ICA as the reference standard, MAR as the quantitative parameter, and AUC to measure diagnostic performance, MPI-CTA fusion imaging provided incremental diagnostic information compared to MPI or CTA alone for the diagnosis and localization of CAD.

99mTc-MAG3: Image Wisely.

Purpose To determine if commonly administered doses of technetium 99m (99mTc) mertiatide (MAG3) in the range of 300-370 MBq (approximately 8-10 mCi) contribute to image interpretation and justify the resulting radiation exposure. Materials and Methods The respective institutional review boards approved this HIPAA-compliant study and waived informed consent. Baseline and furosemide 99mTc-MAG3 imaging examinations in 50 patients suspected of having renal obstruction and 48 patients suspected of having renovascular hypertension (RVH) were randomly selected from archived databases and were independently scored by three experienced readers without access to 2-second flow images. Readers were blinded to their original scores, and then they rescored each examination with access to high-activity 2-second flow images. Relative renal function was determined after a low activity (62.9 MBq ± 40.7) baseline acquisition for RVH and a high activity (303.4 MBq ± 48.1) acquisition after administration of enalaprilat. Data were analyzed by using random effects analysis of variance and mean and standard error of the mean for the difference between sets of scores and the difference between relative function measurements. Results There was no significant difference in the scores without flow images compared with blinded scores with high-activity flow images for patients suspected of having obstruction (P = .80) or RVH (P = .24). Moreover, there was no significant difference in the relative uptake measurements after administration of low and high activities (P > .99). Conclusion Administered doses of 99mTc-MAG3 in the range of 300-370 MBq (approximately 8-10 mCi) do not affect the relative function measurements or contribute to interpretation of images in patients suspected of having RVH or obstruction compared with administration of lower doses; unnecessary radiation exposure can be avoided by administering doses in the range of 37-185 MBq as recommended incurrent guidelines. © RSNA, 2017.

Assessment of 123I-mIBG and 99mTc-tetrofosmin single-photon emission computed tomographic images for the prediction of arrhythmic events in patients with ischemic heart failure: Intermediate severity innervation defects are associated with higher arrhythmic risk.

RATIONALE: 123I-mIBG planar image heart-to-mediastinum ratios effectively risk-stratify heart failure (HF) patients. The value of single-photon emission computed tomographic (SPECT) imaging for identifying increased risk of ventricular arrhythmias is less clear. This study sought to determine if findings from simultaneous interpretation of 123I-mIBG and 99mTc-tetrofosmin SPECT are predictive of arrhythmic events (ArEs). METHODS: 123I-mIBG SPECT images from 622 patients with ischemic HF were presented in standard displays alongside 99mTc-tetrofosmin images. Consensus interpretations using a 17-segment model produced summed scores. Cox proportional hazards analyses related findings to adjudicated ArEs over 2 years. RESULTS: 471 patients had images adequate for total 17-segment scoring. There were 48 ArEs (10.2%). Neither 123I-mIBG nor 99mTc-tetrofosmin SPECT summed scores were univariate predictors. On multivariate proportional hazards analysis, the 123I-mIBG SPECT score was independently predictive of ArEs (HR: 0.975, 95% CI 0.951-0.999, P = 0.042), but HR<1 indicated that risk decreased with increasing score. This occurred because patients with intermediately abnormal SPECT studies had a higher likelihood of ArEs compared to patients with extensive abnormalities. CONCLUSIONS: The presumption of a monotonic increase in ArE risk with increasing summed 123I-mIBG SPECT score may not be correct as ischemic HF patients with abnormalities of intermediate extent appear at highest risk.

Quantitative iodine-123-metaiodobenzylguanidine (MIBG) SPECT imaging in heart failure with left ventricular systolic dysfunction: Development and validation of automated procedures in conjunction with technetium-99m tetrofosmin myocardial perfusion SPECT.

BACKGROUND: The purpose of this study was to develop and validate new approaches to quantitative MIBG myocardial SPECT imaging in heart failure (HF) subjects. METHODS AND RESULTS: Quantitative MIBG myocardial SPECT analysis methods, alone and in conjunction with 99mTc-tetrofosmin perfusion SPECT, were adapted from previously validated techniques for the analysis of SPECT and PET perfusion imaging. To account for underestimation of MIBG defect severity in subjects with global reduction in uptake, a mixed reference database based on planar heart/mediastinum (H/M) ratio categories was used. Extent and severity of voxel-based defects and number of myocardial segments with significant dysinnervation (derived score ≥2) were determined. MIBG/99mTc-tetrofosmin mismatch was quantified using regions with preserved innervation as the reference for scaling 99mTc-tetrofosmin voxel maps. Quantification techniques were tested on studies of 619 ischemic (I) and 319 non-ischemic (NI) HF subjects. Using all analytical techniques, IHF subjects had significantly greater and more severe MIBG SPECT abnormalities compared with NIHF subjects. Innervation/perfusion mismatches were also larger in IHF subjects. Findings were consistent between voxel- and myocardial-segment-based quantitation methods. CONCLUSIONS: Multiple objective methods for quantitation of MIBG SPECT imaging studies provided internally consistent results for distinguishing the different patterns of uptake between IHF and NIHF subjects.

Diagnostic performance of low-dose rest/stress Tc-99m tetrofosmin myocardial perfusion SPECT using the 530c CZT camera: quantitative vs visual analysis.

BACKGROUND: We set out to develop normal databases and prospectively validate abnormality criteria for a low-dose Tc-99m tetrofosmin myocardial perfusion SPECT protocol using the 530c CZT camera. METHODS: All patients received 6 mCi rest/20 mCi stress doses of Tc-99m tetrofosmin. Rest and stress images were obtained over 7-9 and 5-7 minutes according to the chest size. Low-dose CT of the chest was obtained on a standalone CT scanner. Forty patients with very low likelihood (LLK) of coronary artery disease (CAD) were used to define the normal count distributions. The abnormality criteria were prospectively validated in 55 patients who had coronary angiography and in 40 patients with LLK of CAD. RESULTS: The results for quantitative non-attenuation-corrected (AC) and AC analysis and visual analysis were as follows: sensitivity of 79%, 85%, and 92% (P = NS) and specificity of 44%, 75%, and 56% (P = NS), respectively. The normalcy rates for quantitative non-AC and AC analyses and visual analysis were 95%, 98%, and 98% (P = NS). CONCLUSIONS: We have developed non-AC and AC normal databases for low-dose rest/stress Tc-99m tetrofosmin myocardial perfusion SPECT protocol using the 530c CZT camera. The per-patient diagnostic performance of quantitative analyses is not significantly different from visual analysis by an experienced reader.

I-123 mIBG and Tc-99m myocardial SPECT imaging to predict inducibility of ventricular arrhythmia on electrophysiology testing: a retrospective analysis.

OBJECTIVES: The purpose of this study is to assess mIBG uptake in scar border zone and its relation with ventricular arrhythmia (VA) inducibility on electrophysiology (EP) testing using I-123 mIBG SPECT and resting Tc-99m SPECT myocardial perfusion imaging (MPI). METHODS: Forty-seven patients from a previous clinical trial were retrospectively analyzed. These patients underwent I-123 mIBG and resting Tc-99m tetrofosmin SPECT, and EP testing. Twenty-eight patients were positive (EP+) and 19 patients were negative (EP-) for inducibility of sustained (>30 seconds) VA on EP testing. MPI scar extent, border zone extent, and mIBG uptake in border zone were used to predict VA inducibility on EP testing, respectively. RESULTS: There was no significant difference in scar extent between the EP+ and EP- groups. The EP+ group had significantly larger border zone and lower mIBG uptake ratio in the border zone than the EP- group. Receiver operating characteristic (ROC) curve analysis showed that the prediction accuracy for border zone extent (area under ROC = 0.75) was better than scar extent (area under ROC = 0.66). The prediction accuracy was further improved (area under ROC = 0.78), when assessing mIBG uptake in the border zone. CONCLUSION: A new tool has been developed to measure scar and border zone and to assess mIBG uptake in scar and border zone from combined I-123 MIBG SPECT and resting Tc-99m SPECT MPI. The mIBG uptake in the border zone predicted VA inducibility on EP testing with a promising accuracy.

Automatic detection of left and right ventricles from CTA enables efficient alignment of anatomy with myocardial perfusion data.

BACKGROUND: Accurate alignment between cardiac CT angiographic studies (CTA) and nuclear perfusion images is crucial for improved diagnosis of coronary artery disease. This study evaluated in an animal model the accuracy of a CTA fully automated biventricular segmentation algorithm, a necessary step for automatic and thus efficient PET/CT alignment. METHODS AND RESULTS: Twelve pigs with acute infarcts were imaged using Rb-82 PET and 64-slice CTA. Post-mortem myocardium mass measurements were obtained. Endocardial and epicardial myocardial boundaries were manually and automatically detected on the CTA and both segmentations used to perform PET/CT alignment. To assess the segmentation performance, image-based myocardial masses were compared to experimental data; the hand-traced profiles were used as a reference standard to assess the global and slice-by-slice robustness of the automated algorithm in extracting myocardium, LV, and RV. Mean distances between the automated and the manual 3D segmented surfaces were computed. Finally, differences in rotations and translations between the manual and automatic surfaces were estimated post-PET/CT alignment. The largest, smallest, and median distances between interactive and automatic surfaces averaged 1.2 ± 2.1, 0.2 ± 1.6, and 0.7 ± 1.9 mm. The average angular and translational differences in CT/PET alignments were 0.4°, -0.6°, and -2.3° about x, y, and z axes, and 1.8, -2.1, and 2.0 mm in x, y, and z directions. CONCLUSIONS: Our automatic myocardial boundary detection algorithm creates surfaces from CTA that are similar in accuracy and provide similar alignments with PET as those obtained from interactive tracing. Specific difficulties in a reliable segmentation of the apex and base regions will require further improvements in the automated technique.

Myocardial perfusion imaging-derived left ventricular strain: Regional abnormalities associated with transthyretin cardiac amyloidosis.

Background: Transthyretin (ATTR) cardiac amyloidosis is associated with an apical-sparing strain pattern on TTE. We hypothesize that strain indices derived from myocardial perfusion imaging (MPI) can identify this abnormality. Methods: A group with ATTR amyloidosis was compared to age-matched controls with LVH but without amyloidosis who underwent PET or SPECT MPI. Strain values were used to calculate the apical strain index (ASI), apex-to-base ratio (ABR), and ejection fraction to global strain ratio in multiple planes. Results: A direct comparison using Welch's t-tests reveals 6 statistically significant metrics. After regression analysis, the circumferential ASI and ABR at rest remain significantly greater in the ATTR group compared to controls. Conclusion: MPI-derived strain from the circumferential plane at rest may distinguish cardiac amyloidosis from other forms of LVH. If these findings are confirmed with validation studies, routine MPI-derived strain analysis could identify patients with subclinical amyloidosis who may benefit from further testing.

Impact of age on myocardial uptake of ¹²³I-mIBG in older adult subjects without coronary heart disease.

BACKGROUND: The purpose of this study was to examine the relationship between myocardial uptake of (123)I-mIBG and age in older normal adult subjects. METHODS: 94 subjects (age 29-82, mean 58.5) without coronary heart disease were studied. All subjects underwent early and delayed planar and 4-hour SPECT (123)I-mIBG imaging. (123)I-mIBG uptake was quantified as heart/mediastinum ratio on planar images (H/M p) and on SPECT images (H/M s) reconstructed by filtered backprojection, ordered subsets-expectation maximization (OSEM), and OSEM with compensation for collimator septal penetration (DSP). Relationships between age and (123)I-mIBG uptake were examined by correlation analysis, t-tests, and analysis of variance. RESULTS: There was no significant correlation between age and H/M p, reflecting comparable increases in activity in the two regions of interest with age. Results on SPECT analyses were comparable, with no significant correlation between age and H/M s. Using DSP, (123)I-mIBG H/M s was significantly higher in subjects ≥70 of age compared with younger subjects. CONCLUSIONS: Both cardiac and background uptake of (123)I-mIBG increase with age in older subjects without coronary heart disease, resulting in stability of H/M results (planar and SPECT). This study suggests that prognostic analyses of quantitative (123)I-mIBG uptake in patients with heart disease do not require adjustment for patient age.

Left Ventricular Strain from Myocardial Perfusion PET Imaging: Method Development and Comparison to 2-Dimensional Echocardiography.

This study aimed to develop a measure of longitudinal, radial, and circumferential myocardial strain at rest and regadenoson during pharmacologic stress using 82Rb PET electrocardiography-gated myocardial perfusion imaging (MPI). Methods: We retrospectively identified 80 patients who underwent rest and regadenoson-stress CT attenuation-corrected 82Rb PET and had a standard resting transthoracic echocardiogram (TTE) with global longitudinal strain (GLS) analysis within 3 mo. A method was developed to compute longitudinal, radial, and circumferential strain from PET MPI at stress and rest. PET MPI-derived strain and left ventricular function were compared with resting TTE measures as the clinical reference standard. Interobserver agreement of PET MPI strain and left ventricular ejection fraction processing was reported. Results: Longitudinal strain assessed with resting TTE GLS showed good correlation with PET MPI at stress (r = 0.68, P < 0.001) and rest (r = 0.58, P < 0.001). Resting TTE GLS also correlated with PET MPI radial strain at stress (r = -0.70, P < 0.001) and rest (r = -0.59, P < 0.001) and circumferential strain at stress (r = 0.67, P < 0.001) and rest (r = 0.69, P < 0.001). The left ventricular ejection fraction showed good correlation between resting TTE and PET MPI at stress (r = 0.83, P < 0.001) and rest (r = 0.80, P < 0.001). Bland-Altman analysis indicated positive bias of TTE GLS compared with PET MPI longitudinal strain at stress (mean difference = 5.1%, 95% CI = [-2.5, 12.7]) and rest (mean difference = 4.2%, 95% CI = [-4.3, 12.8]). Reproducibility of PET MPI longitudinal strain showed good agreement at stress (concordance correlation coefficient = 0.73, P < 0.001) and rest (concordance correlation coefficient = 0.74, P < 0.001), with Bland-Altman analysis showing a small bias in the longitudinal direction at stress (mean difference = -0.2%) and rest (mean difference = -1.0%). Conclusion: Strain measured with PET MPI using an automated technique correlated well with resting GLS strain obtained by TTE, and the measure is reproducible. Strain from PET MPI should be investigated further to establish reference ranges and assess its value in routine clinical practice.

Computer assisted interpretation of Tc-99m mercaptoacetyltriglycine diuretic scintigraphy enhances resident performance.

OBJECTIVE: iRENEX is a software module that incorporates scintigraphic and clinical data to interpret 99m Tc- mercaptoacetyltriglycine (MAG3) diuretic studies and provide reasons for their conclusions. Our objectives were to compare iRENEX interpretations with those of expert physicians, use iRENEX to evaluate resident performance and determine if iRENEX could improve the diagnostic accuracy of experienced residents. METHODS: Baseline and furosemide 99m Tc-MAG3 acquisitions of 50 patients with suspected obstruction (mean age ± SD, 58.7 ±â€…15.8 years, 60% female) were randomly selected from an archived database and independently interpreted by iRENEX, three expert readers and four nuclear medicine residents with one full year of residency. All raters had access to scintigraphic data and a text file containing clinical information and scored each kidney on a scale from +1.0 to -1.0. Scores ≥0.20 represented obstruction with higher scores indicating greater confidence. Scores +0.19 to -0.19 were indeterminate; scores ≤-0.20 indicated no obstruction. Several months later, residents reinterpreted the studies with access to iRENEX. Receiver operating characteristic (ROC) analysis and concordance correlation coefficient (CCC) quantified agreement. RESULTS: The CCC among experts was higher than that among residents, 0.84, versus 0.39, respectively, P  < 0.001. When residents reinterpreted the studies with iRENEX, their CCC improved from 0.39 to 0.73, P  < 0.001. ROC analysis showed significant improvement in the ability of residents to distinguish between obstructed and non-obstructed kidneys using iRENEX ( P  = 0.036). CONCLUSION: iRENEX interpretations were comparable to those of experts. iRENEX reduced interobserver variability among experienced residents and led to better agreement between resident and expert interpretations.

iRENEX: a clinically informed decision support system for the interpretation of 99mTc-MAG3 scans to detect renal obstruction.

PURPOSE: Decision support systems for imaging analysis and interpretation are rapidly being developed and will have an increasing impact on the practice of medicine. RENEX is a renal expert system to assist physicians evaluate suspected obstruction in patients undergoing mercaptoacetyltriglycine (MAG3) renography. RENEX uses quantitative parameters extracted from the dynamic renal scan data using QuantEM™II and heuristic rules in the form of a knowledge base gleaned from experts to determine if a kidney is obstructed; however, RENEX does not have access to and could not consider the clinical information available to diagnosticians interpreting these studies. We designed and implemented a methodology to incorporate clinical information into RENEX, implemented motion detection and evaluated this new comprehensive system (iRENEX) in a pilot group of 51 renal patients. METHODS: To reach a conclusion as to whether a kidney is obstructed, 56 new clinical rules were added to the previously reported 60 rules used to interpret quantitative MAG3 parameters. All the clinical rules were implemented after iRENEX reached a conclusion on obstruction based on the quantitative MAG3 parameters, and the evidence of obstruction was then modified by the new clinical rules. iRENEX consisted of a library to translate parameter values to certainty factors, a knowledge base with 116 heuristic interpretation rules, a forward chaining inference engine to determine obstruction and a justification engine. A clinical database was developed containing patient histories and imaging report data obtained from the hospital information system associated with the pertinent MAG3 studies. The system was fine-tuned and tested using a pilot group of 51 patients (21 men, mean age 58.2 ± 17.1 years, 100 kidneys) deemed by an expert panel to have 61 unobstructed and 39 obstructed kidneys. RESULTS: iRENEX, using only quantitative MAG3 data agreed with the expert panel in 87 % (34/39) of obstructed and 90 % (55/61) of unobstructed kidneys. iRENEX, using both quantitative and clinical data agreed with the expert panel in 95 % (37/39) of obstructed and 92 % (56/61) of unobstructed kidneys. The clinical information significantly (p < 0.001) increased iRENEX certainty in detecting obstruction over using the quantitative data alone. CONCLUSION: Our renal expert system for detecting renal obstruction has been substantially expanded to incorporate the clinical information available to physicians as well as advanced quality control features and was shown to interpret renal studies in a pilot group at a standardized expert level. These encouraging results warrant a prospective study in a large population of patients with and without renal obstruction to establish the diagnostic performance of iRENEX.

Quantitative I-123 mIBG SPECT in differentiating abnormal and normal mIBG myocardial uptake.

BACKGROUND: The purpose of this study was to evaluate global quantitation of cardiac uptake on I-123 mIBG SPECT. METHODS: The study included a pilot group of 67 subjects and a validation group of 1,051 subjects. SPECT images were reconstructed by filtered backprojection, ordered subsets expectation maximization, and deconvolution of septal penetration, respectively. SPECT heart-to-mediastinum ratio (H/M) was calculated by comparing the mean counts between heart and mediastinum volumes of interest drawn on transaxial images. Receiver operating characteristic (ROC) analysis was used to assess the capability of each SPECT method to differentiate the heart disease subjects from controls in comparison with that of the planar H/M. RESULTS: In the validation group, the areas under the ROC curves were not significantly different between the SPECT and planar H/M. Order subsets expectation maximization had significantly larger area under the ROC curve than the other two SPECT methods. CONCLUSION: H/M obtained from I-123 mIBG SPECT was equivalent to the planar H/M for differentiating between subjects with normal and abnormal mIBG uptake. Global quantification of cardiac I-123 mIBG SPECT may represent a viable alternative to the planar H/M.

Clinically viable myocardial CCTA segmentation for measuring vessel-specific myocardial blood flow from dynamic PET/CCTA hybrid fusion.

BACKGROUND: Positron emission tomography (PET)-derived LV MBF quantification is usually measured in standard anatomical vascular territories potentially averaging flow from normally perfused tissue with those from areas with abnormal flow supply. Previously we reported on an image-based tool to noninvasively measure absolute myocardial blood flow at locations just below individual epicardial vessel to help guide revascularization. The aim of this work is to determine the robustness of vessel-specific flow measurements (MBFvs) extracted from the fusion of dynamic PET (dPET) with coronary computed tomography angiography (CCTA) myocardial segmentations, using flow measured from the fusion with CCTA manual segmentation as the reference standard. METHODS: Forty-three patients' 13NH3 dPET, CCTA image datasets were used to measure the agreement of the MBFvs profiles after the fusion of dPET data with three CCTA anatomical models: (1) a manual model, (2) a fully automated segmented model and (3) a corrected model, where major inaccuracies in the automated segmentation were briefly edited. Pairwise accuracy of the normality/abnormality agreement of flow values along differently extracted vessels was determined by comparing, on a point-by-point basis, each vessel's flow to corresponding vessels' normal limits using Dice coefficients (DC) as the metric. RESULTS: Of the 43 patients CCTA fully automated mask models, 27 patients' borders required manual correction before dPET/CCTA image fusion, but this editing process was brief (2-3 min) allowing a 100% success rate of extracting MBFvs in clinically acceptable times. In total, 124 vessels were analyzed after dPET fusion with the manual and corrected CCTA mask models yielding 2225 stress and 2122 rest flow values. Forty-seven vessels were analyzed after fusion with the fully automatic masks producing 840 stress and 825 rest flow samples. All DC coefficients computed globally or by territory were ≥ 0.93. No statistical differences were found in the normal/abnormal flow classifications between manual and corrected or manual and fully automated CCTA masks. CONCLUSION: Fully automated and manually corrected myocardial CCTA segmentation provides anatomical masks in clinically acceptable times for vessel-specific myocardial blood flow measurements using dynamic PET/CCTA image fusion which are not significantly different in flow accuracy and within clinically acceptable processing times compared to fully manually segmented CCTA myocardial masks.

Dynamic cardiac PET motion correction using 3D normalized gradient fields in patients and phantom simulations.

This work expands on the implementation of three-dimensional (3D) normalized gradient fields to correct for whole-body motion and cardiac creep in [N-13]-ammonia patient studies and evaluates its accuracy using a dynamic phantom simulation model. METHODS: A full rigid-body algorithm was developed using 3D normalized gradient fields including a multi-resolution step and sampling off the voxel grid to reduce interpolation artifacts. Optimization was performed using a weighted similarity metric that accounts for opposing gradients between images of blood pool and perfused tissue without the need for segmentation. Forty-three retrospective dynamic [N-13]-ammonia PET/CT rest/adenosine-stress patient studies were motion corrected and the mean motion parameters plotted at each frame time point. Motion correction accuracy was assessed using a comprehensive dynamic XCAT simulation incorporating published physiologic parameters of the heart's trajectory following adenosine infusion as well as corrupted attenuation correction commonly observed in clinical studies. Accuracy of the algorithm was assessed objectively by comparing the errors between isosurfaces and centers of mass of the motion corrected XCAT simulations. RESULTS: In the patient studies, the overall mean cranial-to-caudal translation was 7 mm at stress over the duration of the adenosine infusion. Noninvasive clinical measures of relative flow reserve and myocardial flow reserve were highly correlated with their invasive analogues. Motion correction accuracy assessed with the XCAT simulations showed an error of <1 mm in late perfusion frames that broadened gradually to <3 mm in earlier frames containing blood pool. CONCLUSION: This work demonstrates that patients undergoing [N-13]-ammonia dynamic PET/CT exhibit a large cranial-to-caudal translation related to cardiac creep primarily at stress and to a lesser extent at rest, which can be accurately corrected by optimizing their 3D normalized gradient fields. Our approach provides a solution to the challenging condition where the image intensity and its gradients are opposed without the need for segmentation and remains robust in the presence of PET-CT mismatch.