Improving evaluation for TTR amyloidosis by interactive filtering of Tc-99 m PYP SPECT images. The role for "clean blood pool" imaging.

BACKGROUND: Myocardial imaging with bone agents such as Tc-99 m PYP and HMDP has assumed a central role in the evaluation of patients with suspected transthyretin (TTR) amyloidosis. Visual scoring (VS) (0-3 +) and the heart to contralateral lung ratio (HCL) classify many patients as equivocal when mediastinal uptake is apparent but cannot be further differentiated into myocardial uptake versus blood pool. SPECT imaging has been recommended but current reconstruction protocols frequently produce amorphous mediastinal activity that also fails to discriminate between myocardial activity and blood pool. We hypothesized that interactive filtering interactively using a deconvolving filter would improve this. METHODS: We identified 176 sequential patients referred for TTR amyloid imaging. All patients had planar imaging, 101 had planar imaging with a large field of view camera that allowed HCL measurements. SPECT imaging was performed on a 3-headed digital camera with lead fluorescence attenuation correction. One study was excluded for technical reasons. We created software to allow interactive filtering while reconstructing the images then overlay them on attenuation mu maps to assist localization of myocardial/mediastinal uptake. Conventional Butterworth and an interactive inverse Gaussian filters were employed to differentiate myocardial uptake from residual blood pool. We defined "clean blood pool" (CBP) as recognizable blood pool with no activity in the surrounding myocardium. A scan was determined diagnostic if it showed CBP, positive uptake or no identifiable mediastinal uptake. RESULTS: 76/175 (43%) were equivocal (1 +) by visual uptake. Of these 22 (29%) were diagnostic by Butterworth but 71 (93%) were by inverse gaussian (p < .0001). 71/101 (70%) were equivocal by HCL (1-1.5). Of these, 25 (35%) were diagnostic by Butterworth but 68 (96%) were diagnostic by inverse gaussian (p < .0001). This was driven by a greater than threefold increase in the identification of CBP by inverse gaussian filtering. CONCLUSION: CBP can be identified in the vast majority of patients with equivocal PYP scans using optimized reconstruction and can greatly reduce the number of equivocal scans.

Feasibility of SPECT myocardial perfusion imaging in the super-obese using a multi-head semiconductor camera with attenuation correction.

BACKGROUND: Obesity is not only associated with an increased risk of coronary artery disease, but also decreases the accuracy of many diagnostic modalities pertinent to this disease. Advances in myocardial perfusion imaging (MPI) have mitigated somewhat the effects of obesity, although the feasibility of MPI in the super-obese (defined as a BMI > 50) is currently untested. We undertook this study to assess the practicality of MPI in the super-obese using a multi-headed solid-state gamma camera with attenuation correction. METHODS: We retrospectively identified consecutive super-obese patients referred for MPI at our institution. The images were interpreted by 3 blinded, experienced readers and graded for quality and diagnosis, and subjectively evaluated the contribution of attenuation correction. Clinical follow-up was obtained from review of medical records. RESULTS: 72 consecutive super-obese patients were included. Their BMI ranged from 50 to 67 (55.7 ± 5.1). Stress image quality was considered good or excellent in 45 (63%), satisfactory in 24 (33%), poor in 3 (4%), and uninterpretable in 0 patients. Rest images were considered good or excellent in 34 (49%), satisfactory in 23 (33%), poor in 13 (19%), and uninterpretable in 0 patients. Attenuation correction changed the interpretation in 34 (47%) of studies. CONCLUSIONS: MPI is feasible and provides acceptable image quality for super-obese patients, although it may be camera and protocol dependent.

Limitations of parametric modeling of the left ventricle using first harmonic analysis: possible role for gaussian modeling.

BACKGROUND: Fourier (cosine) analysis of time activity curves (TAC) of radionuclide ventriculography (RVG) may oversimplify the TAC and has limitations. METHODS: We identified 21 patients who had undergone 24 frame planar RVG with ejection fractions ranging from 8% to 76% (43% ± 19%). The TAC for each pixel was fitted to both a cosine and gaussian function then analyzed on a pixel-by-pixel basis then over the entire LV. Second, mathematical simulations were performed to analyze the stability of each fit in the presence of low amplitude and noise. RESULTS: The fit was slightly but significantly better for the gaussian compared to the cosine function (RMS gaussian 13.0% ± 2.5% vs 13.5% ± 2.1% cosine; P = .016). There was near exact correlation with amplitude and between gaussian mu and cosine phase. The SD of phase from the cosine fit correlated strongly with the SD of the mu from the gaussian fit. The proposed new measure of dyssynchrony, the sigma parameter of the gaussian fit, correlated with the SD of the cosine phase (r = 0.520, P = .016). Simulations showed gradual but modest deviation of the sigma parameter from the gaussian fit with lower amplitudes whereas the deviation of the calculated SD of phase increased exponentially with decreasing amplitude. CONCLUSIONS: First harmonic (cosine) fitting has significant limitations. Gaussian fitting is an alternative way to model the LV TAC. The sigma from the gaussian may provide additional information LV dyssynchrony and is less influenced by image noise. Gaussian fitting merits further evaluation for modeling LV function.

The 'ramp filter' artifact in filtered back projection is due to attenuation not ramp filtering.

BACKGROUND: Artifactually decreased counts in regions of the left ventricle adjacent to intense splanchnic activity have been described in single photon emission computed tomography imaging and attributed to the side lobes of ramp filtering, a crucial step in filtered back projection. However, filtered back projection is based on a mathematical proof that makes no assumptions about the distribution of activity implying there is another cause for this artifact. METHODS: Computer simulations of a left ventricle with varying intensity, proximity and size of adjacent splanchnic activity was created with and without attenuation. Both filtered back projection and an iterative expectation maximization algorithm were performed. Volume weighted bullseyes were generated then the ratio of average activity in regions in the anterior and inferior walls of the bullseye were calculated. RESULTS: In the absence of attenuation, there was no effect on uptake despite intense adjacent splanchnic activity. The effect of attenuation alone was mild but was greatly amplified by adjacent splanchnic uptake. This effect varied with the intensity, proximity and size of adjacent activity. Moving the attenuation to the apex of the left ventricle produced a paradoxical increase in counts to the inferior wall. The effect on iterative reconstruction was less pronounced than with filtered back projection and the paradoxical increase with attenuation at the apex did not occur. CONCLUSION: The ramp filter artifact in single photon emission computed tomography reconstruction is due to attenuation, not ramp filtering per se. Adjacent splanchnic activity can result in either an increase or decrease in counts. These effects are less pronounced and more consistent with iterative reconstruction.

The ratio of the apex/anterior wall: a marker of breast attenuation artifact in women.

BACKGROUND: Breast attenuation artifact is well known for reducing the accuracy of myocardial perfusion imaging in women. We have noticed the particular pattern of relative preservation of apical activity in women with breast attenuation and decreased anterior wall counts. This study was undertaken to see if this finding could be used to improve the accuracy of perfusion imaging in women. METHODS: We identified 295 women referred for exercise stress testing using (201)Tl single photon emission computed tomography (SPECT) of whom 193 had less than 5% probability of coronary disease and 102 had coronary artery disease documented by catheterization within 60 days of stress testing (mean of 1.8+/-0.8 vessels with=50% stenosis). Patients with documented myocardial infarction, pathologic Q waves, left bundle branch block, non-ischaemic cardiomyopathy or prior bypass grafting were excluded. Volume-weighted bullseye plots were generated and normalized to 100; next, regions of interest were drawn over the anterior wall and apex in all patients and the ratio of the mean counts of each region was calculated. The normals were further divided into those with breast attenuation (defined as mean anterior counts <70% maximum) and those without. Defect scores of all patients were calculated; a formula to adjust the score for patients with breast attenuation was developed. Accuracy was assessed by calculating the area under the receiver operating curve. RESULTS: The normals, overall, had a mean ratio of 1.0+/-0.08 vs. 0.9+/-0.16 for those with coronary disease (P<0.0001). In normals with breast attenuation the ratio was 1.1+/-0.08 compared to 0.99+/-0.07 (P<0.0001) without. By adjusting the anterior wall defect score in patients with apex/anterior ratio >1 we were able to improve the accuracy from 0.808+/-0.028 to 0.826+/-0.027 (P<0.01). CONCLUSIONS: A ratio of the apex to the anterior wall >1 is not physiological and suggests the presence of significant breast attenuation artifact. This finding can be used to produce a small but statistically significant improvement in the accuracy of quantitative thallium SPECT in women who have not undergone coronary bypass grafting.

Factor Analysis of the MSVT, NV-MSVT, and RBANS Memory Subtests.

OBJECTIVE: The current investigation sought to define the relationship between established performance validity tests and measures of memory via a factor analytic strategy first published by Heyanka, Thaler, Linck, Pastorek, Miller, Romesser, & Sim (2015). A Factor analytic approach to the validation of the Word Memory Test and Test of Memory Malingering as measures of effort and not memory. Archives of Clinical Neuropsychology, 30, 369-376. METHOD: The full range of Medical Symptom Validity Test (MSVT) and Non-Verbal Medical Symptom Validity Test (NV-MSVT) subtests were factor analyzed with the memory scales of the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) in a sample of 346 service members with a history of concussion. RESULTS: A two-factor solution was extracted with the MSVT and NV-MSVT effort and paired associate subtests loading on one factor and the RBANS subtests loading on a second factor. CONCLUSIONS: Results support the conclusion that the effort subtests of the MSVT and NV-MSVT tap a different construct from established memory measures.

An affective-motivational model of marijuana and alcohol problems among college students.

The study examined models of marijuana (n = 309) and alcohol (n = 731) problems. Impulsivity was directly associated with both marijuana- and alcohol-related problems. Negative mood regulation expectancies were indirectly associated with marijuana problems through coping motives. Sensation seeking was indirectly associated with alcohol problems through enhancement motives. Affect lability and negative affect were indirectly associated with alcohol problems though coping motives. In both models, coping motives were directly associated with use-related problems. A multigroup analysis indicated that the association between negative affect and coping motives as well as use and problems was stronger among participants using both alcohol and marijuana relative to alcohol only. Enhancement motives were a stronger predictor of alcohol use among participants using alcohol only.

Orthographic neighborhood size, number of word meanings, and number of higher frequency neighbors.

The present lexical decision task experiment examined orthographic (large, medium, small), number-of-meanings (ambiguous, unambiguous), and number of higher frequency neighbors (few, many), factors that have to date not been studied simultaneously. For both mean reaction time and percentage error, the critical three-way interaction between these factors was significant. Breakdown of this interaction revealed that the ambiguity effect (unambiguous-ambiguous) decreased as neighborhood size increased, but only when there were many higher frequency neighbors in the neighborhood. These results appear inconsistent with serial search models but are understandable within the context of interactive-activation models of word recognition.