Metabolic tumor volume predicts outcome in patients with advanced stage follicular lymphoma from the RELEVANCE trial.

BACKGROUND: We investigated the prognostic value of baseline positron emission tomography (PET) parameters for patients with treatment-naïve follicular lymphoma (FL) in the phase III RELEVANCE trial, comparing the immunomodulatory combination of lenalidomide and rituximab (R2) versus R-chemotherapy (R-chemo), with both regimens followed by R maintenance therapy. PATIENTS AND METHODS: Baseline characteristics of the entire PET-evaluable population (n = 406/1032) were well balanced between treatment arms. The maximal standard uptake value (SUVmax) and the standardized maximal distance between tow lesions (SDmax) were extracted, the standardized distance between two lesions the furthest apart, were extracted. The total metabolic tumor volume (TMTV) was computed using the 41% SUVmax method. RESULTS: With a median follow-up of 6.5 years, the 6-year progression-free survival (PFS) was 57.8%, the median TMTV was 284 cm3, SUVmax was 11.3 and SDmax was 0.32 m-1, with no significant difference between arms. High TMTV (>510 cm3) and FLIPI were associated with an inferior PFS (P = 0.013 and P = 0.006, respectively), whereas SUVmax and SDmax were not (P = 0.08 and P = 0.12, respectively). In multivariable analysis, follicular lymphoma international prognostic index (FLIPI) and TMTV remained significantly associated with PFS (P = 0.0119 and P = 0.0379, respectively). These two adverse factors combined stratified the overall population into three risk groups: patients with no risk factors (40%), with one factor (44%), or with both (16%), with a 6-year PFS of 67.7%, 54.5%, and 41.0%, respectively. No significant interaction between treatment arms and TMTV or FLIPI (P = 0.31 or P = 0.59, respectively) was observed. The high-risk group (high TMTV and FLIPI 3-5) had a similar PFS in both arms (P = 0.45) with a median PFS of 68.4% in the R-chemo arm versus 71.4% in the R2 arm. CONCLUSIONS: Baseline TMTV is predictive of PFS, independently of FLIPI, in patients with advanced FL even in the context of antibody maintenance.

Risk stratification in diffuse large B-cell lymphoma using lesion dissemination and metabolic tumor burden calculated from baseline PET/CT†.

BACKGROUND: We analyzed the prognostic value of a new baseline positron emission tomography (PET) parameter reflecting the spread of the disease, the largest distance between two lesions (Dmax). We tested its complementarity to metabolic tumor volume (MTV) in a large cohort of diffuse large B-cell lymphoma (DLBCL) patients from the REMARC trial (NCT01122472). PATIENTS AND METHODS: MTVs were defined using the 41% maximum standardized uptake value threshold. From the three-dimensional coordinates, the centroid of each lesion was automatically obtained and considered as the lesion location. The distances between all pairs were calculated. Dmax was obtained for each patient and normalized with the body surface area [standardized Dmax (SDmax)]. RESULTS: From the REMARC trial, 290 patients aged 60-80 years were included: 91% had an advanced stage and 71% International Prognostic Index (IPI) ≥3. High versus low SDmax significantly impacted progression-free survival (PFS) (P < 0.0001) and overall survival (OS) (P = 0.0027). Patients with SDmax > 0.32 m-1 (n = 82) had a 4-year PFS and OS of 46% and 71%, respectively, against 77% and 87%, respectively, for patients with low SDmax. High SDmax and high MTV were independent prognostic factors of PFS (P = 0.0001 and P = 0.0010, respectively) and OS (P = 0.0028 and P = 0.0004, respectively). Combining MTV and SDmax yielded three risk groups with no (n = 109), one (n = 122) or two (n = 59) factors (P < 0.0001 for both PFS and OS). The 4-year PFS were 90%, 63%, 41%, respectively, and the 4-year OS were 95%, 79%, 66%, respectively. In addition, patients with at least two of the three factors including high SDmax, high MTV, Eastern Cooperative Oncology Group (ECOG) ≥2 had a higher number of central nervous system relapse (P = 0.017). CONCLUSIONS: SDmax is a simple feature that captures lymphoma dissemination, independent from MTV. These two PET metrics, SDmax and MTV, are complementary to characterize the disease, reflecting the tumor burden and its spread. This score appeared promising for DLBCL baseline risk stratification.

Fully 4D list-mode reconstruction applied to respiratory-gated PET scans.

(18)F-fluoro-deoxy-glucose ((18)F-FDG) positron emission tomography (PET) is one of the most sensitive and specific imaging modalities for the diagnosis of non-small cell lung cancer. A drawback of PET is that it requires several minutes of acquisition per bed position, which results in images being affected by respiratory blur. Respiratory gating techniques have been developed to deal with respiratory motion in the PET images. However, these techniques considerably increase the level of noise in the reconstructed images unless the acquisition time is increased. The aim of this paper is to evaluate a four-dimensional (4D) image reconstruction algorithm that combines the acquired events in all the gates whilst preserving the motion deblurring. This algorithm was compared to classic ordered subset expectation maximization (OSEM) reconstruction of gated and non-gated images, and to temporal filtering of gated images reconstructed with OSEM. Two datasets were used for comparing the different reconstruction approaches: one involving the NEMA IEC/2001 body phantom in motion, the other obtained using Monte-Carlo simulations of the NCAT breathing phantom. Results show that 4D reconstruction reaches a similar performance in terms of the signal-to-noise ratio (SNR) as non-gated reconstruction whilst preserving the motion deblurring. In particular, 4D reconstruction improves the SNR compared to respiratory-gated images reconstructed with the OSEM algorithm. Temporal filtering of the OSEM-reconstructed images helps improve the SNR, but does not achieve the same performance as 4D reconstruction. 4D reconstruction of respiratory-gated images thus appears as a promising tool to reach the same performance in terms of the SNR as non-gated acquisitions while reducing the motion blur, without increasing the acquisition time.

Experimental characterisation of a proton kernel model for pencil beam scanning techniques.

The aim of this work is to perform Monte Carlo simulations of a proton pencil beam scanning machine, characterise the low-dose envelope of scanned proton beams and assess the differences between various approximations for nozzle geometry. Measurements and Monte Carlo simulations were carried out in order to describe the dose distribution of a proton pencil beam in water for energies between 100 and 220 MeV. Dose distributions were simulated by using a Geant4 Monte Carlo platform (TOPAS), and were measured in water using a two-dimensional ion chamber array detector. The beam source in air was adjusted for each configuration. Double Gaussian parameterisation was proposed for definition of the beam source model in order to improve simulations starting at the nozzle exit. Absolute dose distributions and field size factors were measured and compared with simulations. The influence of the high-density components present in the treatment nozzle was also investigated by analysis of proton phase spaces at the nozzle exit. An excellent agreement was observed between experimental dose distributions and simulations for energies higher than 160 MeV. However, minor differences were observed between 100 and 160 MeV, suggesting poorer modelling of the beam when the full treatment head was not taken into account. We found that the first ionisation chamber was the main cause of the tail component observed for low proton beam energies. In this work, various parameterisations of proton sources were proposed, thereby allowing reproduction of the low-dose envelope of proton beams and excellent agreement with measured data.

Perfusion-weighted MR imaging studies in brain hypervascular diseases: comparison of arterial input function extractions for perfusion measurement.

BACKGROUND AND PURPOSE: Brain hypervascular diseases are complex and induce hemodynamic disturbances on brain parenchyma, which are difficult to accurately evaluate by using perfusion-weighted (PWI) MR imaging. Our purpose was to test and to assess the best AIF estimation method among 4 patients with brain hypervascular disease and healthy volunteers. METHODS: Thirty-three patients and 10 healthy volunteers underwent brain perfusion studies by using a 1.5T MR imaging scanner with gadolinium-chelate bolus injection. PWI was performed with the indicator dilution method. AIF estimation methods were performed with local, regional, regional scaled, and global estimated arterial input function (AIF), and PWI measurements (cerebral blood volume [CBV] and cerebral blood flow [CBF]) were performed with regions of interest drawn on the thalami and centrum semiovale in all subjects, remote from the brain hypervascular disease nidus. Abnormal PWI results were assessed by using Z Score, and evaluation of the best AIF estimation method was performed by using a no gold standard evaluation method. RESULTS: From 88% to 97% of patients had overall abnormal perfusion areas of hypo- (decreased CBV and CBF) and/or hyperperfusion (increased CBV and CBF) and/or venous congestion (increased CBV, normal or decreased CBF), depending on the AIF estimation method used for PWI computations. No gold standard evaluation of the 4 AIF estimates found the regional and the regional scaled methods to be the most accurate. CONCLUSION: Brain hypervascular disease induces remote brain perfusion abnormalities that can be better detected by using PWI with regional or regional scaled AIF estimation methods.

Validation of the Monte Carlo simulator GATE for indium-111 imaging.

Monte Carlo simulations are useful for optimizing and assessing single photon emission computed tomography (SPECT) protocols, especially when aiming at measuring quantitative parameters from SPECT images. Before Monte Carlo simulated data can be trusted, the simulation model must be validated. The purpose of this work was to validate the use of GATE, a new Monte Carlo simulation platform based on GEANT4, for modelling indium-111 SPECT data, the quantification of which is of foremost importance for dosimetric studies. To that end, acquisitions of (111)In line sources in air and in water and of a cylindrical phantom were performed, together with the corresponding simulations. The simulation model included Monte Carlo modelling of the camera collimator and of a back-compartment accounting for photomultiplier tubes and associated electronics. Energy spectra, spatial resolution, sensitivity values, images and count profiles obtained for experimental and simulated data were compared. An excellent agreement was found between experimental and simulated energy spectra. For source-to-collimator distances varying from 0 to 20 cm, simulated and experimental spatial resolution differed by less than 2% in air, while the simulated sensitivity values were within 4% of the experimental values. The simulation of the cylindrical phantom closely reproduced the experimental data. These results suggest that GATE enables accurate simulation of (111)In SPECT acquisitions.

Fully 3D Monte Carlo reconstruction in SPECT: a feasibility study.

In single photon emission computed tomography (SPECT) with parallel hole collimation, image reconstruction is usually performed as a set of bidimensional (2D) analytical or iterative reconstructions. This approach ignores the tridimensional (3D) nature of scatter and detector response function that affects the detected signal. To deal with the 3D nature of the image formation process, iterative reconstruction can be used by considering a 3D projector modelling the 3D spread of photons. In this paper, we investigate the value of using accurate Monte Carlo simulations to determine the 3D projector used in a fully 3D Monte Carlo (F3DMC) reconstruction approach. Given the 3D projector modelling all physical effects affecting the imaging process, the reconstruction problem is solved using the maximum likelihood expectation maximization (MLEM) algorithm. To validate the concept, three data sets were simulated and F3DMC was compared with two other 3D reconstruction strategies using analytical corrections for attenuation, scatter and camera point spread function. Results suggest that F3DMC improves spatial resolution, relative and absolute quantitation and signal-to-noise ratio. The practical feasibility of the approach on real data sets is discussed.

Assessment of a fully 3D Monte Carlo reconstruction method for preclinical PET with iodine-124.

Iodine-124 is a radionuclide well suited to the labeling of intact monoclonal antibodies. Yet, accurate quantification in preclinical imaging with I-124 is challenging due to the large positron range and a complex decay scheme including high-energy gammas. The aim of this work was to assess the quantitative performance of a fully 3D Monte Carlo (MC) reconstruction for preclinical I-124 PET. The high-resolution small animal PET Inveon (Siemens) was simulated using GATE 6.1. Three system matrices (SM) of different complexity were calculated in addition to a Siddon-based ray tracing approach for comparison purpose. Each system matrix accounted for a more or less complete description of the physics processes both in the scanned object and in the PET scanner. One homogeneous water phantom and three heterogeneous phantoms including water, lungs and bones were simulated, where hot and cold regions were used to assess activity recovery as well as the trade-off between contrast recovery and noise in different regions. The benefit of accounting for scatter, attenuation, positron range and spurious coincidences occurring in the object when calculating the system matrix used to reconstruct I-124 PET images was highlighted. We found that the use of an MC SM including a thorough modelling of the detector response and physical effects in a uniform water-equivalent phantom was efficient to get reasonable quantitative accuracy in homogeneous and heterogeneous phantoms. Modelling the phantom heterogeneities in the SM did not necessarily yield the most accurate estimate of the activity distribution, due to the high variance affecting many SM elements in the most sophisticated SM.

Monte-Carlo simulations of clinically realistic respiratory gated (18)F-FDG PET: application to lesion detectability and volume measurements.

In PET/CT thoracic imaging, respiratory motion reduces image quality. A solution consists in performing respiratory gated PET acquisitions. The aim of this study was to generate clinically realistic Monte-Carlo respiratory PET data, obtained using the 4D-NCAT numerical phantom and the GATE simulation tool, to assess the impact of respiratory motion and respiratory-motion compensation in PET on lesion detection and volume measurement. To obtain reconstructed images as close as possible to those obtained in clinical conditions, a particular attention was paid to apply to the simulated data the same correction and reconstruction processes as those applied to real clinical data. The simulations required 140,000h (CPU) generating 1.5 To of data (98 respiratory gated and 49 ungated scans). Calibration phantom and patient reconstructed images from the simulated data were visually and quantitatively very similar to those obtained in clinical studies. The lesion detectability was higher when the better trade-off between lesion movement limitation (compared to ungated acquisitions) and image statistic preservation is considered (respiratory cycle sampling in 3 frames). We then compared the lesion volumes measured on conventional PET acquisitions versus respiratory gated acquisitions, using an automatic segmentation method and a 40%-threshold approach. A time consuming initial manual exclusion of noisy structures needed with the 40%-threshold was not necessary when the automatic method was used. The lesion detectability along with the accuracy of tumor volume estimates was largely improved with the gated compared to ungated PET images.

Comparative assessment of nine scatter correction methods based on spectral analysis using Monte Carlo simulations.

UNLABELLED: We compared nine scatter correction methods based on spectral analysis which process SPECT projections. METHODS: Monte Carlo simulation was used to generate histories of photons emitted from a realistic 99mTc phantom. A particular projection was considered. Information regarding the history, location and energy of the photons detected in this projection was analyzed to test the assumptions underlying each scatter correction method. Relative and absolute quantification and signal-to-noise ratio were assessed for each scatter corrected image. RESULTS: For the simulated data, two methods do not enable activity quantification. Among the methods requiring some parameters to be calibrated, the dual-energy window method shows the best compromise between accuracy and ease of implementation but introduces a bias in relative quantification. In this respect, a triple-energy window technique is more accurate than the dual-window method. A factor analysis approach results in more stable quantitative accuracy (error approximately 10%) for a wide range of activity but requires a more sophisticated acquisition mode (30 energy windows). CONCLUSION: These results show that a scatter correction method using spectral analysis can be used to substantially improve accurate quantification.

Impact of attenuation correction by simultaneous emission/transmission tomography on visual assessment of 201Tl myocardial perfusion images.

UNLABELLED: It has been shown in clinical studies that for subjects with a low likelihood of coronary artery disease (CAD), attenuation correction (AC) improves the specificity of defect detection in the inferior wall (right coronary artery [RCA] region). The aim of this study was to investigate the effect of AC on the visual interpretation of the RCA and anteroseptal (corresponding to the left anterior descending artery [LAD]) regions in CAD patients. METHODS: Fifty-six patients with suspected CAD underwent 20Tl stress/4 h-delayed imaging SPECT using a simultaneous 201Tl emission/99mTc transmission imaging protocol. Images were reconstructed using the maximum likelihood-expectation maximum algorithm without and with AC. The stress/4 h-delayed images were interpreted blindly for reversible or fixed defects in the RCA and LAD regions by three experienced physicians. Coronary angiography, electrocardiography and enzyme findings were used to establish diagnoses of ischemia or infarction, and receiver operating characteristic (ROC) analyses were performed. RESULTS: Statistical testing of ROC curve areas showed that defect detection performance improved with AC when compared with performance without AC in the RCA region. This was mainly the result of a systematic increase in specificity of 12% or more (for any observer and any type of defect) for a similar sensitivity (no definite change in sensitivity values). However, defect detection performance significantly decreased in the LAD territory with AC images (P < 0.05) because of a systematic decrease in sensitivity of 20% or more, with no consistent change in specificity. Similar trends were observed when reversible and fixed defects were considered separately. CONCLUSION: AC significantly affects the visual interpretation of 201Tl stress/4 h-delayed SPECT images. This study confirmed the increase in specificity obtained with AC in the RCA territory. However, in the population considered, the studied AC was deleterious for the LAD territory assessment.

A "hybrid" method for measuring myocardial wall thickening from gated PET/SPECT images.

UNLABELLED: We introduce a hybrid index, HYB, which combines counts with geometric information to measure wall thickening from PET/SPECT gated images. Its accuracy is compared with that of a count-based index (MAX) and a geometric index (FWHM). METHODS: For each index, the index values versus thickness and the estimated thickening values versus true thickening were investigated using theoretical analyses, realistic simulated data obtained from clinical gated MR scans, phantom measurements and preliminary gated MRI and PET patient studies. Each index was studied for different spatial resolutions and noise and background conditions. The performance of each index was quantified using a parameter "Q" reflecting bias and variability of thickening estimates. RESULTS: HYB varied more linearly with thickness than MAX and FWHM, resulting in a better Q value than with MAX and FWHM for all noise, background and spatial resolutions. ROC analysis confirmed that HYB significantly increases the sensitivity and specificity for detection of wall thickening abnormalities (sensitivity = 100%; specificity = 85% for HYB, 95% and 50% for MAX and 100% and 0% for FWHM, respectively). CONCLUSION: Use of the hybrid index instead of conventional count-based or geometric indices should improve the classification of normal/abnormal wall thickening values in gated SPECT and PET.

Impact of scatter correction in planar scintimammography: a phantom study.

UNLABELLED: This study examines how scatter correction might affect lesion detection and quantitation of tumor-to-normal breast tissue activity ratio in planar scintimammography. METHODS: Forty-one phantom acquisitions were performed to mimic a wide variety of scintimammographic imaging conditions in which lesions would be close to the chest wall. For each acquisition, the images corresponding to a 10% energy window (110) and two scatter correction methods [the Jaszczak (JA) method and a factor analysis (FA)-based method] were obtained in addition to the conventional 20% image (120). A total of 368 images in which detection of the "tumor" was judged borderline were selected, and 10 independent observers were asked to detect lesions in these images. Receiver operating curve analyses were performed to assess detection performance. Tumor-to-normal tissue activity ratios were calculated for quantitative analysis. RESULTS: Detection performance significantly improved for the I10, JA and FA images compared to the 120 images, with an increase in sensitivity up to 8% for FA images. Sensitivity was especially increased for small lesions (13- and 16-mm3 spheres) and true heart-to-normal tissue activity ratios of > 12. Scatter correction also increased the certainty with which the readers gave their judgment. The tumor-to-normal tissue activity ratio was approximately 8% larger on JA or FA images and 1% larger on the I10 images compared to the 120 images. For a given image, the variability with which this ratio was estimated was reduced by approximately 4% on JA and FA images. CONCLUSION: Based on these phantom results, scatter correction might be used with benefit in scintimammography.

Relative impact of scatter, collimator response, attenuation, and finite spatial resolution corrections in cardiac SPECT.

UNLABELLED: We determined the relative effect of corrections for scatter, depth-dependent collimator response, attenuation, and finite spatial resolution on various image characteristics in cardiac SPECT. METHODS: Monte Carlo simulations and real acquisition of a 99mTc cardiac phantom were performed under comparable conditions. Simulated and acquired data were reconstructed using several correction schemes that combined different methods for scatter correction (3 methods), depth-dependent collimator response correction (frequency-distance principle), attenuation correction (nonuniform Chang correction or within an iterative reconstruction algorithm), and finite spatial resolution correction (use of recovery coefficients). Five criteia were considered to assess the effect of the processing schemes: bull's-eye map (BEM) uniformity, contrast between the left ventricle (LV) wall and the LV cavity, spatial resolution, signal-to-noise ratio (SNR), and percent errors with respect to the known LV wall and liver activities. RESULTS: Similar results were obtained for the simulated and acquired data. Scatter correction significantly improved contrast and absolute quantitation but did not have noticeable effects on BEM uniformity or on spatial resolution and reduced the SNR. Correction for the depth-dependent collimator response improved spatial resolution from 13.3 to 9.5 mm in the LV region, improved absolute quantitation and contrast, but reduced the SNR. Correcting for attenuation was essential for restoring BEM uniformity (78% and 89% without and with attenuation correction, respectively [ideal value being 100%]) and accurate absolute activity quantitation (errors in estimated LV wall and liver activity decreased from 90% without attenuation correction to approximately20% with attenuation correction only). Although accurate absolute activity quantitation was achieved in the liver using scatter and attenuation corrections only, correction for finite spatial resolution was needed to estimate LV wall activity within 10%. CONCLUSION: The respective effects of corrections for scatter, depth-dependent collimator response, attenuation, and finite spatial resolution on different image features in cardiac SPECT were quantified for a specific acquisition configuration. These results give indications regarding the improvements to be expected when using a specific processing scheme involving some or all corrections.

Statistical distribution of factors and factor images in factor analysis of medical image sequences.

From a time or energy image sequence, factor analysis of medical image sequences (FAMIS) estimates factors, representing kinetics or spectra in a given physiological compartment, and associated factor images, showing the compartments corresponding to each curve. In this paper, we show that the statistical properties of factor images and associated factors can be determined using a well known result from elementary probability theory. Numerical experiments are conducted to demonstrate that the variance observed in factor images can be predicted when the statistical properties of the original data are known. It is shown how these theoretical results can be used to relax the non-negativity constraints during FAMIS oblique analysis and to improve the quantitative interpretation of the factor images by associating a confidence interval with each pixel value.

A new correction method for gamma camera non-uniformity due to energy response variability.

We present a new uniformity correction (Fourier energy correction) which is designed to correct for gamma camera non-uniformity caused by variations of the energy response function within a wide spectral range. A convolution model is used to describe the spatial distortions of the energy response function. The model is solved in Fourier space. A preliminary flood acquisition is required to obtain energy-dependent Fourier weights which are used to correct subsequent acquisitions. The influence of the parameters involved in the correction procedure is studied and the Fourier energy correction is compared to a conventional multiplicative energy correction for different acquisition geometries. The Fourier energy correction appears especially useful when the energy information associated with each detected photon is analysed using a fine sampling, or when windows different from the photopeak window are used.

Target apex-seeking in factor analysis of medical image sequences.

The aim of factor analysis of medical image sequences (FAMIS) is to estimate a limited number of physical or physiological fundamental functions. Its oblique rotation stage strongly affects the quality and the interpretation of the resulting estimates (factors and factor images). A new target apex-seeking method which integrates physical or physiological knowledge in this stage is described. This knowledge concerns some of the fundamental functions and reacts on the determination of all the factors. A simulated spectral study illustrates the method. We discuss its properties in comparison with the other approaches using a priori physical or physiological information.

A statistical model for the determination of the optimal metric in factor analysis of medical image sequences (FAMIS).

A statistical model is added to the conventional physical model underlying factor analysis of medical image sequences (FAMIS). It allows a derivation of the optimal metric to be used for the orthogonal decomposition involved in FAMIS. The oblique analysis of FAMIS is extended to take this optimal metric into account. The case of scintigraphic image sequences is used. We derive in this case that the optimal decomposition is obtained by correspondence analysis. A scintigraphic dynamic study illustrates the practical consequences of the use of the optimal metric in FAMIS.

Detection of inter-hemispheric asymmetries of brain perfusion in SPECT.

Technetium-99m HMPAO and technetium-99m ECD single photon emission computed tomography (SPECT) imaging is commonly used to highlight brain regions with altered perfusion. It is particularly useful in the investigation of intractable partial epilepsy. However, SPECT suffers from poor spatial resolution that makes interpretation difficult. In this context, we propose an unsupervised voxel neighbourhood based method to assist the detection of significant functional inter-hemispheric asymmetries in brain SPECT, using anatomical information from MRI. For each MRI voxel, the anatomically homologous voxel in the contralateral hemisphere is identified. Both homologous voxel coordinates are then mapped into the SPECT volume using SPECT-MRI registration. Neighbourhoods are then defined around each SPECT voxel and compared to obtain a volume of inter-hemispheric differences. A volume including only the statistically significant inter-hemispheric differences is deduced from this volume using a non-parametric approach. The method was validated using realistic analytical simulated SPECT data including known asymmetries (in size and amplitude) as ground truth (gold standard). Detection performance was assessed using an ROC (receiver operating characteristic) approach based on the measures of the overlap between known and detected asymmetries. Validation with computer-simulated data demonstrates the ability to detect asymmetric zones with relatively small extension and amplitude. The registration of these detected functional asymmetries on the MRI enables good anatomical localization to be achieved.