Search

SUMMARYThe opportunistic pathogenPseudomonas viridiflavacolonizes more than fifty agricultural crop species and is the most commonPseudomonasin the phyllosphere of EuropeanArabidopsis thalianapopulations. Belonging to theP. syringaecomplex, it is genetically and phenotypically distinct from well-characterizedP. syringae sensu stricto. Despite its prevalence, we lack knowledge of howA. thalianaresponds to its native isolates at the molecular level. Here, we characterize the host response in anA. thaliana-P. viridiflavapathosystem.We measured host and pathogen growth in axenic infections, and used immune mutants, transcriptomics, and metabolomics to determine defense pathways influencing susceptibility toP. viridiflavainfection.Infection withP. viridiflavaincreased jasmonic acid (JA) levels and the expression of ethylene defense pathway marker genes. The immune response in a susceptible host accession was delayed compared to a tolerant one. Mechanical injury rescued susceptibility, consistent with an involvement of JA.The JA/ethylene pathway is important for suppression ofP. viridiflava, yet suppression capacity varies between accessions. Our results shed light on howA. thalianacan suppress the ever-presentP. viridiflava, but further studies are needed to understand howP. viridiflavaevades this suppression to spread broadly acrossA. thalianapopulations.

Sphingomonas is one of the most abundant bacterial genera in the phyllosphere of wild Arabidopsis thaliana, but relative to Pseudomonas, the ecology of Sphingomonas and its interaction with plants is poorly described. We analyzed the genomic features of over 400 Sphingomonas isolates collected from local A. thaliana populations, which revealed much higher intergenomic diversity than for the considerably more uniform Pseudomonas isolates found in the same host populations. Variation in Sphingomonas plasmid complements and additional genomic features suggest high adaptability of this genus, and the widespread presence of protein secretion systems hints at frequent biotic interactions. While some of the isolates showed plant-protective phenotypes in lab tests, this was a rare trait. To begin to understand the extent of strain sharing across alternate hosts, we employed amplicon sequencing and a bulk-culturing metagenomics approach on both A. thaliana and neighboring plants. Our data reveal that both Sphingomonas and Pseudomonas thrive on other diverse plant hosts, but that Sphingomonas is a poor in or dead leaves., Proceedings of the National Academy of Sciences of the United States of America, 119 (52), ISSN:0027-8424, ISSN:1091-6490

Although clonal propagation is frequently used in commercial plant breeding and plant biotechnology programs because it minimizes genetic variation, it is not uncommon to observe clonal plants with stable phenotypic changes, a phenomenon known as somaclonal variation. Several studies have shown that epigenetic modifications induced during regeneration are associated with this newly acquired phenotypic variation. However, the factors that determine the extent of somaclonal variation and the molecular changes associated with it remain poorly understood. To address this gap in our knowledge, we compared clonally propagated Arabidopsis thaliana plants derived from somatic embryogenesis using two different embryonic transcription factors-RWP-RK DOMAIN-CONTAINING 4 (RKD4) or LEAFY COTYLEDON2 (LEC2) and from two epigenetically distinct tissues. We found that both the epi(genetic) status of explant and the regeneration protocol employed play critical roles in shaping the molecular and phenotypic state of clonal plants. Phenotypic variation of regenerated plants can be largely explained by the inheritance of tissue-specific DNA methylation imprints, which are associated with specific transcriptional and metabolic changes in sexual progeny of clonal plants. Moreover, regenerants from roots were particularly affected by the inheritance of epigenetic imprints, which resulted in increased accumulation of salicylic acid in leaves and accelerated plant senescence. Collectively, our data reveal pathways for targeted manipulation of phenotypic variation in clonal plants.

Complex traits, such as growth and fitness, are typically controlled by a very large number of variants, which can interact in both additive and non-additive fashion. In an attempt to gauge the relative importance of both types of genetic interactions, we have turned to hybrids, which provide a facile means for creating many novel allele combinations. We focused on the interaction between alleles of the same locus and performed a transcriptomic study involving 141 random crosses between different accessions of the plant model speciesArabidopsis thaliana. Additivity is rare, consistently observed for only about 300 genes enriched for roles in stress response and cell death. Regulatory rare-allele burden affects the expression level of these genes but does not correlate with F1rosette size. Non-additive gene expression in F1hybrids is much more common, with the vast majority of genes (over 90%) being expressed below parental average. Unlike in the additive genes, regulatory rare-allele burden in the non-additive gene set is strongly correlated with F1rosette size, even though it only mildly covary with the expression level of these genes. Our study underscores under-dominance as the predominant gene action associated with emergence of rosette growth trajectories in theA. thalianahybrid model. Our work lays the foundation for understanding molecular mechanisms and evolutionary forces that lead to dominance complementation of rare regulatory alleles.

Trabajo presentado en el International XI Symposium on Grapevine Physiology and Biotechnology, celebrado en Stellenbosch (Sudáfrica), del 31 de octubre al 5 de noviembre de 2021, Tempranillo is the third most widely grown red wine cultivar worldwide and the most relevant in the Iberian Peninsula. Here we produced a diploid genome assembly of Tempranillo to study the origin of clonal improvement in this cultivar. After PacBio and Oxford Nanopore sequencing, a trio binning approach produced chromosome-arm level and complete assemblies for the two haplo-phases. Based on the de novo assembly, we searched for genetic variation in ten re-sequenced Tempranillo clones selected for variation in agronomically relevant traits including fruit colour, low fruit sugar accumulation adaptable to warmer climates or loose bunches that are less prone to fungal diseases. Thousands of somatic mutations were detected in the ten clones, which, combined with gene annotations based on PacBio Iso-seq, identified candidate dominant missense variants related with selected phenotypes. We could also phase inter-chromosomal structural variation (SV) in the origin of white gape and loose cluster phenotypes. Our results provide a quality genome assembly for inter- and intra-cultivar comparisons and release candidate missense and SV mutations that could be responsible for clonal diversification and improvement.

Linking plant phenotype to genotype is a common goal to both plant breeders and geneticists. However, collecting phenotypic data for large numbers of plants remain a bottleneck. Plant phenotyping is mostly image based and therefore requires rapid and robust extraction of phenotypic measurements from image data. However, because segmentation tools usually rely on color information, they are sensitive to background or plant color deviations. We have developed a versatile, fully open-source pipeline to extract phenotypic measurements from plant images in an unsupervised manner. ARADEEPOPSIS (https://github.com/Gregor-Mendel-Institute/aradeepopsis) uses semantic segmentation of top-view images to classify leaf tissue into three categories: healthy, anthocyanin rich, and senescent. This makes it particularly powerful at quantitative phenotyping of different developmental stages, mutants with aberrant leaf color and/or phenotype, and plants growing in stressful conditions. On a panel of 210 natural Arabidopsis (Arabidopsis thaliana) accessions, we were able to not only accurately segment images of phenotypically diverse genotypes but also to identify known loci related to anthocyanin production and early necrosis in genome-wide association analyses. Our pipeline accurately processed images of diverse origin, quality, and background composition, and of a distantly related Brassicaceae. ARADEEPOPSIS is deployable on most operating systems and high-performance computing environments and can be used independently of bioinformatics expertise and resources.

Sphingomonas is one of the most abundant bacterial genera in the phyllosphere of wild Arabidopsis thaliana, but relative to Pseudomonas, the ecology of Sphingomonas and its interaction with plants remains elusive. We analyzed the genomic features of over 400 Sphingomonas isolates collected from local A. thaliana populations, which revealed high intergenomic diversity, in contrast to genetically much more uniform Pseudomonas isolates found in the same host populations. Variation in Sphingomonas plasmid complements and additional genomic features suggest high adaptability of this genus, and the widespread presence of protein secretion systems hints at frequent biotic interactions. While some of the isolates showed plant-protective properties in lab tests, this was a rare trait. To begin to understand the extent of strain sharing across alternate hosts, we employed amplicon sequencing and a novel bulk-culturing metagenomics approach on both A. thaliana and neighboring plants. Our data reveal that Sphingomonas and Pseudomonas both thrive on other diverse plant hosts, but that Sphingomonas is a poor competitor in dying or dead leaves.

New combinations of genetic material brought together through hybridization can lead to unfit offspring as a result of outbreeding or inbreeding depression. In selfing plants such asArabidopsis thaliana, outbreeding depression is typically the result of pairwise deleterious epistatic interactions between two alleles that can geographically co-occur. What remains elusive is how often alleles resulting in genetic incompatibilities co-occur in natural populations of outcrossing plant species. To address this question, we screened over two thousand five hundred wildArabidopsis arenosahybrid plants in search for potential genetic mismatches. We show that although abnormal deleterious phenotypes are common, the transcriptional profiles of these abnormalA. arenosaplants differ substantially from those seen in incompatibleA. thalianahybrids. The abnormal hybrid phenotypes inA. arenosahad different underlying genetic architectures, yet a repeated theme was increased homozygosity, indicating that inbreeding rather than outbreeding depression gives rise to some of the deleterious phenotypes segregating in wildA. arenosapopulations.

Linking plant phenotype to genotype, i.e., identifying genetic determinants of phenotypic traits, is a common goal of both plant breeders and geneticists. While the ever-growing genomic resources and rapid decrease of sequencing costs have led to enormous amounts of genomic data, collecting phenotypic data for large numbers of plants remains a bottleneck. Many phenotyping strategies rely on imaging plants, which makes it necessary to extract phenotypic measurements from these images rapidly and robustly. Common image segmentation tools for plant phenotyping mostly rely on color information, which is error-prone when either background or plant color deviate from the underlying expectations. We have developed a versatile, fully open-source pipeline to extract phenotypic measurements from plant images in an unsupervised manner. aradeepopsis was built around the deep-learning model DeepLabV3+ that was re-trained for segmentation of Arabidopsis thaliana rosettes. It uses semantic segmentation to classify leaf tissue into up to three categories: healthy, anthocyanin-rich, and senescent. This makes aradeepopsis particularly powerful at quantitative phenotyping from early to late developmental stages, of mutants with aberrant leaf color and/or phenotype, and of plants growing in stressful conditions where leaf color may deviate from green. Using our tool on a panel of 210 natural Arabidopsis accessions, we were able to not only accurately segment images of phenotypically diverse genotypes but also to map known loci related to anthocyanin production and early necrosis using the aradeepopsis output in genome-wide association analyses. Our pipeline is able to handle images of diverse origins, image quality, and background composition, and could even accurately segment images of a distantly related Brassicaceae. Because it can be deployed on virtually any common operating system and is compatible with several high-performance computing environments, aradeepopsis can be used independently of bioinformatics expertise and computing resources. aradeepopsis is available at https://github.com/Gregor-Mendel-Institute/aradeepopsis.

Linking plant phenotype to genotype is a common goal to both plant breeders and geneticists. However, collecting phenotypic data for large numbers of plants remain a bottleneck. Plant phenotyping is mostly image based and therefore requires rapid and robust extraction of phenotypic measurements from image data. However, because segmentation tools usually rely on color information, they are sensitive to background or plant color deviations. We have developed a versatile, fully open-source pipeline to extract phenotypic measurements from plant images in an unsupervised manner. ARADEEPOPSIS (https://github.com/Gregor-Mendel-Institute/aradeepopsis) uses semantic segmentation of top-view images to classify leaf tissue into three categories: healthy, anthocyanin rich, and senescent. This makes it particularly powerful at quantitative phenotyping of different developmental stages, mutants with aberrant leaf color and/or phenotype, and plants growing in stressful conditions. On a panel of 210 natural Arabidopsis (Arabidopsis thaliana) accessions, we were able to not only accurately segment images of phenotypically diverse genotypes but also to identify known loci related to anthocyanin production and early necrosis in genome-wide association analyses. Our pipeline accurately processed images of diverse origin, quality, and background composition, and of a distantly related Brassicaceae. ARADEEPOPSIS is deployable on most operating systems and high-performance computing environments and can be used independently of bioinformatics expertise and resources.

A central goal in microbiome research is to learn what distinguishes a healthy from a dysbiotic microbial community. Shifts in diversity and taxonomic composition are important indicators of dysbiosis, but a full understanding also requires knowledge of absolute microbial population sizes. In addition to the number of microbial cells, information on taxonomic composition can provide important insight into microbiome function and disease state. Here we use shotgun metagenomics to simultaneously assess microbiome composition and microbial load in the phyllosphere of wild populations of the plant Arabidopsis thaliana. We find that wild plants vary substantially in the load of colonizing microbes, and that high loads are typically associated with the proliferation of single taxa, with only a few putatively pathogenic taxa achieving high abundances in the field. Our results suggest (i) that the inside of a plant leaf is on average sparsely colonized with an estimated two bacterial genomes per plant genome and an order of magnitude fewer eukaryotic microbial genomes, and (ii) that higher levels of microbial cells often indicate successful colonization by pathogens. Lastly, our results show that load is a significant explanatory variable for loss of estimated Shannon diversity in phyllosphere microbiomes, implying that reduced diversity may be a significant predictor of microbial dysbiosis in a plant leaf.

SummaryHybrid offspring can look very different from their parents, including having greatly increased or decreased fitness. In many plant species, conflicts between divergent elements of the immune system can cause hybrids to express autoimmunity, a generally deleterious syndrome known as hybrid necrosis. We are investigating multiple hybrid necrosis cases in Arabidopsis thaliana that are caused by allele-specific interactions between different variants at two unlinked resistance (R) gene clusters. One is the RESISTANCE TO PERONOSPORA PARASITICA 7 (RPP7) cluster, which encodes an intracellular nucleotide binding site-leucine rich repeat (NLR) immune receptors that confer strain-specific resistance to oomycetes. The other is the RESISTANCE TO POWDERY MILDEW 8 (RPW8)/HOMOLOG OF RPW8 (HR) locus, which encodes atypical resistance proteins that can confer broad-spectrum resistance to filamentous pathogens. There is extensive structural variation in the RPW8/HR cluster, both at the level of gene copy number and at the level of C-terminal protein repeats of unknown function. We demonstrate that the number of RPW8/HR repeats correlate, albeit in a complex manner, with the severity of hybrid necrosis when these alleles are combined with specific RPP7 variants. This observation suggests that gross structural differences, rather than individual amino acid polymorphisms, guide the genetic interaction between RPW8/HR and RPP7 alleles. We discuss these findings in light of the similarity of RPW8/HR proteins with pore-forming toxins, MLKL and HET-S, from mammals and fungi.

OBJECTIVE The aim of this study was to evaluate the clinical applicability and technical feasibility of fluorodeoxyglucose (FDG) positron emission tomography (PET)/magnetic resonance imaging (MRI) compared with FDG PET/computed tomography (CT) in young children focusing on lesion detection, PET quantification, and potential savings in radiation exposure. METHODS Twenty examinations (10 PET/CT and 10 PET/MRI examinations) were performed prospectively in 9 patients with solid tumors (3 female, 6 male; mean age, 4.8 [1-6] years). Fluorodeoxyglucose PET/CT and FDG PET/MRI were performed sequentially after a single tracer injection. Lesion detection and analysis were performed independently in PET/CT and PET/MRI. Potential changes in diagnostic or therapeutic patient management were recorded. Positron emission tomography quantification in PET/MRI was evaluated by comparing standardized uptake values resulting from MRI-based and CT-based attenuation correction. Effective radiation doses of PET and CT were estimated. RESULTS Twenty-one PET-positive lesions were found congruently in PET/CT and PET/MRI. Magnetic resonance imaging enabled significantly better detection of morphologic PET correlates compared with CT. Eight suspicious PET-negative lesions were identified by MRI, of which one was missed in CT. Sensitivity, specificity, and accuracy for correct lesion classification were not significantly different (90%, 47%, and 62% in PET/CT; 100%, 68%, and 79% in PET/MRI, respectively). In 4 patients, the use of PET/MRI resulted in a potential change in diagnostic management compared with PET/CT, as local and whole-body staging could be performed within 1 single examination. In 1 patient, PET/MRI initiated a change in therapeutic management. Positron emission tomography quantification using MRI-based attenuation correction was accurate compared with CT-based attenuation correction. Higher standardized uptake value deviations of about 18% were observed in the lungs due to misclassification in MRI-based attenuation maps. Potential reduction in radiation dose was 48% in PET/MRI compared with PET/CT (P < 0.05). CONCLUSIONS FDG PET/MRI is at least equivalent to FDG PET/CT for oncologic imaging in young children. Specifically, superior soft tissue contrast of MRI results in higher confidence in lesion interpretation. Substantial savings in radiation exposure can be achieved, and the number of necessary imaging examinations can be reduced using PET/MRI compared with PET/CT.

Significance Imaging provides an insight into biological patho-mechanisms of diseases. However, the link between the imaging phenotype and the underlying molecular processes is often not well understood. Methods such as metabolomics and proteomics reveal detailed information about these processes. Unfortunately, they provide no spatial information and thus cannot be easily correlated with functional imaging. We have developed an image-guided milling machine and unique workflows to precisely isolate tissue samples based on imaging data. The tissue samples remain cooled during the entire procedure, preventing sample degradation. This enables us to correlate, at an unprecedented spatial precision, comprehensive imaging information with metabolomics and proteomics data, leading to a better understanding of diseases., Phenotypic heterogeneity is commonly observed in diseased tissue, specifically in tumors. Multimodal imaging technologies can reveal tissue heterogeneity noninvasively in vivo, enabling imaging-based profiling of receptors, metabolism, morphology, or function on a macroscopic scale. In contrast, in vitro multiomics, immunohistochemistry, or histology techniques accurately characterize these heterogeneities in the cellular and subcellular scales in a more comprehensive but ex vivo manner. The complementary in vivo and ex vivo information would provide an enormous potential to better characterize a disease. However, this requires spatially accurate coregistration of these data by image-driven sampling as well as fast sample-preparation methods. Here, a unique image-guided milling machine and workflow for precise extraction of tissue samples from small laboratory animals or excised organs has been developed and evaluated. The samples can be delineated on tomographic images as volumes of interest and can be extracted with a spatial accuracy better than 0.25 mm. The samples remain cooled throughout the procedure to ensure metabolic stability, a precondition for accurate in vitro analysis.

Recent findings suggest that alternative splicing has a critical role in controlling the responses of plants to temperature variations. However, alternative splicing factors in plants are largely uncharacterized. Here we establish the putative splice regulator, PORCUPINE (PCP), as temperature-specific regulator of development in Arabidopsis thaliana. Our findings point to the misregulation of WUSCHEL and CLAVATA3 as the possible cause for the meristem defects affecting the pcp-1 loss-of-function mutants at low temperatures.

Multiparametric medical imaging data can be large and are often complex. Machine learning algorithms can assist in image interpretation when reliable training data exist. In most cases, however, knowledge about ground truth (e.g. histology) and thus training data is limited, which makes application of machine learning algorithms difficult. The purpose of this study was to design and implement a learning algorithm for classification of multidimensional medical imaging data that is robust and accurate even with limited prior knowledge and that allows for generalization and application to unseen data. Local prostate cancer was chosen as a model for application and validation. 16 patients underwent combined simultaneous [11C]-choline positron emission tomography (PET)/MRI. The following imaging parameters were acquired: T2 signal intensities, apparent diffusion coefficients, parameters Ktrans and Kep from dynamic contrast-enhanced MRI, and PET standardized uptake values (SUVs). A spatially constrained fuzzy c-means algorithm (sFCM) was applied to the single datasets and the resulting labeled data were used for training of a support vector machine (SVM) classifier. Accuracy and false positive and false negative rates of the proposed algorithm were determined in comparison with manual tumor delineation. For five of the 16 patients rates were also determined in comparison with the histopathological standard of reference. The combined sFCM/SVM algorithm proposed in this study revealed reliable classification results consistent with the histopathological reference standard and comparable to those of manual tumor delineation. sFCM/SVM generally performed better than unsupervised sFCM alone. We observed an improvement in accuracy with increasing number of imaging parameters used for clustering and SVM training. In particular, including PET SUVs as an additional parameter markedly improved classification results. A variety of applications are conceivable, especially for imaging of tissues without easily available histopathological correlation. Copyright © 2015 John Wiley & Sons, Ltd.

Pediatric imaging is regarded as a key application for combined PET/MR imaging systems. Because existing MR-based attenuation-correction methods were not designed specifically for pediatric patients, we assessed the impact of 2 potentially influential factors: inter- and intrapatient variability of attenuation coefficients and anatomic variability. Furthermore, we evaluated the quantification accuracy of 3 methods for MR-based attenuation correction without (SEGbase) and with bone prediction using an adult and a pediatric atlas (SEGwBONEad and SEGwBONEpe, respectively) on PET data of pediatric patients. Methods: The variability of attenuation coefficients between and within pediatric (5–17 y, n = 17) and adult (27–66 y, n = 16) patient collectives was assessed on volumes of interest (VOIs) in CT datasets for different tissue types. Anatomic variability was assessed on SEGwBONEad/pe attenuation maps by computing mean differences to CT-based attenuation maps for regions of bone tissue, lungs, and soft tissue. PET quantification was evaluated on VOIs with physiologic uptake and on 80% isocontour VOIs with elevated uptake in the thorax and abdomen/pelvis. Inter- and intrapatient variability of the bias was assessed for each VOI group and method. Results: Statistically significant differences in mean VOI Hounsfield unit values and linear attenuation coefficients between adult and pediatric collectives were found in the lungs and femur. The prediction of attenuation maps using the pediatric atlas showed a reduced error in bone tissue and better delineation of bone structure. Evaluation of PET quantification accuracy showed statistically significant mean errors in mean standardized uptake values of −14% ± 5% and −23% ± 6% in bone marrow and femur-adjacent VOIs with physiologic uptake for SEGbase, which could be reduced to 0% ± 4% and −1% ± 5% using SEGwBONEpe attenuation maps. Bias in soft-tissue VOIs was less than 5% for all methods. Lung VOIs showed high SDs in the range of 15% for all methods. For VOIs with elevated uptake, mean and SD were less than 5% except in the thorax. Conclusion: The use of a dedicated atlas for the pediatric patient collective resulted in improved attenuation map prediction in osseous regions and reduced interpatient bias variation in femur-adjacent VOIs. For the lungs, in which intrapatient variation was higher for the pediatric collective, a patient- or group-specific attenuation coefficient might improve attenuation map accuracy. Mean errors of −14% and −23% in bone marrow and femur-adjacent VOIs can affect PET quantification in these regions when bone tissue is ignored.

To compare positron emission tomography (PET)/magnetic resonance (MR) imaging and PET/computed tomography (CT) for lesion detection and interpretation, quantification of fluorine 18 ((18)F) fluorodeoxyglucose (FDG) uptake, and accuracy of MR-based PET attenuation correction in pediatric patients with solid tumors. Materials and Methods This prospective study had local ethics committee and German Federal Institute for Drugs and Medical Devices approval. Written informed consent was obtained from all patients and legal guardians. Twenty whole-body (18)F-FDG PET/CT and (18)F-FDG PET/MR examinations were performed in 18 pediatric patients (median age, 14 years; range, 11-17 years). (18)F-FDG PET/CT and (18)F-FDG PET/MR data were acquired sequentially on the same day for all patients. PET standardized uptake values (SUVs) were quantified with volume of interest measurements in lesions and healthy tissues. MR-based PET attenuation correction was compared with CT-derived attenuation maps (µ-maps). Lesion detection was assessed with separate reading of PET/CT and PET/MR data. Estimates of radiation dose were derived from the applied doses of (18)F-FDG and CT protocol parameters. Descriptive statistical analyses were performed to report correlation coefficients and relative deviations for comparison of SUVs, rates of lesion detection, and percentage reductions in radiation dose.PET SUVs showed strong correlations between PET of PET/CT (PETCT) and PET of PET/MR (PETMR) (r0.85 for most tissues). Apart from drawbacks of MR-based PET attenuation correction in osseous structures and lungs, similar SUVs were found on PET images corrected with CT-based µ-maps (13.1% deviation of SUVs for bone marrow and5% deviation for other tissues). Lesion detection rate with PET/MR imaging was equivalent to that with PET/CT (61 areas of focal uptake on PETMR images vs 62 areas on PETCT images). Advantages of PET/MR were observed especially in soft-tissue regions. Furthermore, PET/MR offered significant dose reduction (73%) compared with PET/CT.Pediatric oncologic PET/MR is technically feasible, showing satisfactory performance for PET quantification with SUVs similar to those of PET/CT. Compared with PET/CT, PET/MR demonstrates equivalent lesion detection rates while offering markedly reduced radiation exposure. Thus, PET/MR is a promising modality for the clinical work-up of pediatric malignancies. Online supplemental material is available for this article.

Objective A potential major application of simultaneous avalanche photodiode-based positron emission tomography (PET)/magnetic resonance imaging (MRI) systems are quantitative brain studies for cerebral blood flow measurements in combination with blood-oxygen-level-dependent or perfusion MRI, requiring a high performance for both modalities. Thus, we evaluated PET quantification accuracy and homogeneity for 2 different simultaneous PET/MRI systems (whole-body and brain scanner) compared with those of a state-of-the-art PET detector (PET/computed tomography) using phantom studies. In addition, we investigated the long-term stability of PET and quality of functional MRI measurements of a clinical whole-body PET/MRI scanner. Materials and methods Phantom measurements were conducted using spheres filled with [F]-fluoride distributed in a homogeneous cylinder phantom at different positions inside the PET field of view. Recovery values and standard deviations were extracted from resulting PET images. The influence of magnetic resonance-based attenuation correction and that of activity outside the PET field of view on the recovery values of these spheres was evaluated. Furthermore, long-term PET stability of the whole-body PET/MRI system was assessed by evaluating position profiles, energy spectra, count rates, and recovery values from [Ge]-phantom scans. Functional MRI applicability was tested in accordance with the functional Biomedical Information Research Network procedure. Results The BrainPET system showed high recovery values (up to 99%) but also increased variability (up to 7.4%). Significant underestimations in PET quantification near activity outside the PET field of view were found (up to 80%). Using magnetic resonance-based attenuation correction led to an underestimation in PET activity of approximately 7%. In distinction, the whole-body PET/MRI system revealed performance similar to the PET/computed tomographic scanner (recovery values up to approximately 60% with a variability of approximately 4%). Long-term stability and fMRI performance of the whole-body PET/MRI scanner showed no degradation compared with stand-alone systems. Conclusions Homogeneity and accuracy of avalanche photodiode-based PET detectors is comparable with those of the state-of-the-art detectors based on photomultiplier tubes. However, attenuation correction on PET/MRI systems has to be adapted carefully for quantitative PET measurements. The BrainPET system needs improved scatter correction to perform quantitative brain studies. The whole-body PET/MRI scanner, however, is applicable for quantitative brain studies.

Hybrid PET/MR systems have rapidly progressed from the prototype stage to systems that are increasingly being used in the clinics. This review provides an overview of developments in hybrid PET/MR systems and summarizes the current state of the art in PET/MR instrumentation, correction techniques, and data analysis. The strong magnetic field requires considerable changes in the manner by which PET images are acquired and has led, among others, to the development of new PET detectors, such as silicon photomultipliers. During more than a decade of active PET/MR development, several system designs have been described. The technical background of combined PET/MR systems is explained and related challenges are discussed. The necessity for PET attenuation correction required new methods based on MR data. Therefore, an overview of recent developments in this field is provided. Furthermore, MR-based motion correction techniques for PET are discussed, as integrated PET/MR systems provide a platform for measuring motion with high temporal resolution without additional instrumentation. The MR component in PET/MR systems can provide functional information about disease processes or brain function alongside anatomic images. Against this background, we point out new opportunities for data analysis in this new field of multimodal molecular imaging.

Combined positron emission tomography (PET) and magnetic resonance imaging (MRI) is a new tool to study functional processes in the brain. Here we study brain function in response to a barrel-field stimulus simultaneously using PET, which traces changes in glucose metabolism on a slow time scale, and functional MRI (fMRI), which assesses fast vascular and oxygenation changes during activation. We found spatial and quantitative discrepancies between the PET and the fMRI activation data. The functional connectivity of the rat brain was assessed by both modalities: the fMRI approach determined a total of nine known neural networks, whereas the PET method identified seven glucose metabolism-related networks. These results demonstrate the feasibility of combined PET-MRI for the simultaneous study of the brain at activation and rest, revealing comprehensive and complementary information to further decode brain function and brain networks.

Hybrid whole-body magnetic resonance/positron emission tomography (MR/PET) systems are a new diagnostic tool enabling the simultaneous acquisition of morphologic and multiple functional data and thus allowing for a diversified characterization of oncological diseases.The aim of this study was to investigate the image and alignment quality of MR/PET in patients with pulmonary lesions and to compare the congruency of the 2 functional measurements of diffusion-weighted imaging (DWI) in MR imaging and 2-deoxy-[18F] fluoro-2-D-glucose (FDG) uptake in PET.A total of 15 patients were examined with a routine positron emission tomography/computer tomography (PET/CT) protocol and, subsequently, in a whole-body MR/PET scanner allowing for simultaneous PET and MR data acquisition. The PET and MR image quality was assessed visually using a 4-point score (1, insufficient; 4, excellent). The alignment quality of the rigidly registered PET/CT and MR/PET data sets was investigated on the basis of multiple anatomic landmarks of the lung using a scoring system from 1 (no alignment) to 4 (very good alignment). In addition, the alignment quality of the tumor lesions in PET/CT and MR/PET as well as for retrospective fusion of PET from PET/CT and MR images was assessed quantitatively and was compared between lesions strongly or less influenced by respiratory motion. The correlation of the simultaneously acquired DWI and FDG uptake in the pulmonary masses was analyzed using the minimum and mean apparent diffusion coefficient (ADC min and ADC mean) as well as the maximum and mean standardized uptake value (SUV max and SUV mean), respectively. In addition, the correlation of SUV max from PET/CT data was investigated as well. On lesions 3 cm or greater, a voxelwise analysis of ADC and SUV was performed.The visual evaluation revealed excellent image quality of the PET images (mean [SD] score, 3.6 [0.5]) and overall good image quality of DWI (mean [SD] score of 2.5 [0.5] for ADC maps and 2.7 [0.5] for diffusion-weighted images, respectively). The alignment quality of the data sets was very good in both MR/PET and PET/CT without significant differences (overall mean [SD] score of MR/PET, 3.8 [0.4]; PET/CT 3.6 [0.5]). Also, the alignment quality of the tumor lesions showed no significant differences between PET/CT and MR/PET (mean cumulative misalignment of MR/PET, 7.7 mm; PET/CT, 7.0 mm; P = 0.705) but between both modalities and a retrospective fusion (mean cumulative misalignment, 17.1 mm; P = 0.002 and P = 0.008 for PET/CT and MR/PET, respectively). Also, the comparison of the lesions strongly or less influenced by respiratory motion showed significant differences only for the retrospective fusion (21.3 mm vs 11.5 mm, respectively; P = 0.043). The ADC min and SUV max as measures of the cell density and glucose metabolism showed a significant reverse correlation (r = -0.80; P = 0.0006). No significant correlation was found between ADC mean and SUV mean (r = -0.42; P = 0.1392). Also, SUV max from the PET/CT data showed significant reverse correlation to ADC min (r = -0.62; P = 0.019). The voxelwise analysis of 5 pulmonary lesions each showed weak but significant negative correlation between ADC and SUV.Examinations of pulmonary lesions in a simultaneous whole-body MR/PET system provide diagnostic image quality in both modalities. Although DWI and FDG-PET reflect different tissue properties, there may very well be an association between the measures of both methods most probably because of increased cellularity and glucose metabolism of FDG-avid pulmonary lesions. A voxelwise DWI and FDG-PET correlation might provide a more sophisticated spatial characterization of pulmonary lesions.

Recent progress has allowed hybrid positron emission tomography/magnetic resonance (PET/MR) systems to make the transition from research prototypes to systems with full potential for clinical imaging. Options for directly measuring the attenuation maps, as is possible with PET/computed tomography or PET transmission scans, are not included in PET/MR scanners. New methods to compute attenuation maps from MR data have therefore been developed.

Purpose The aim of the study was to investigate the influence of lung density changes as well as bone proximity on the attenuation correction of lung standardized uptake values (SUVs). Methods and materials 15 patients with mostly oncologic diseases were examined in 18F-FDG-PET/CT and subsequently in a fully integrated PET/MR scanner. From each PET dataset acquired in PET/MR, four different PET reconstructions were computed using different attenuation maps (μ-maps): i) CT-based μ-map (gold standard); ii) CT-based μ-map in which the linear attenuation coefficients (LAC) of the lung tissue was replaced by the lung LAC from the MR-based segmentation method; iii) based on reconstruction ii), the LAC of bone structures was additionally replaced with the LAC from the MR-based segmentation method; iv) the vendor-provided MR-based μ-map (segmentation-based method). Those steps were performed using MATLAB. CT Hounsfield units (HU) and SUVmean was acquired in different levels and regions of the lung. Relative differences between the differently corrected PETs were computed. Results Compared to the gold standard, reconstruction ii), iii) and iv) led to a relative underestimation of SUV in the posterior regions of -9.0%, -13.4% and -14.0%, respectively. Anterior and middle regions were less affected with an overestimation of about 6–8% in reconstructions ii)–iv). Conclusion It could be shown that both, differences in lung density and the vicinity of bone tissue in the μ-map may have an influence on SUV, mostly affecting the posterior lung regions.

Clinical PET/MR acquisition protocols entail the use of MR contrast agents (MRCA) that could potentially affect PET quantification following MR-based attenuation correction (AC). We assessed the effect of oral and intravenous (IV) MRCA on PET quantification in PET/MR imaging.We employed two MRCA: Lumirem (oral) and Gadovist (IV). First, we determined their reference PET attenuation values using a PET transmission scan (ECAT-EXACT HR+, Siemens) and a CT scan (PET/CT Biograph 16 HI-REZ, Siemens). Second, we evaluated the attenuation of PET signals in the presence of MRCA. Phantoms were filled with clinically relevant concentrations of MRCA in a background of water and (18)F-fluoride, and imaged using a PET/CT scanner (Biograph 16 HI-REZ, Siemens) and a PET/MR scanner (Biograph mMR, Siemens). Third, we investigated the effect of clinically relevant volumes of MRCA on MR-based AC using human pilot data: a patient study employing Gadovist (IV) and a volunteer study employing two different oral MRCA (Lumirem and pineapple juice). MR-based attenuation maps were calculated following Dixon-based fat-water segmentation and an external atlas-based and pattern recognition (ATPR) algorithm.IV and oral MRCA in clinically relevant concentrations were found to have PET attenuation values similar to those of water. The phantom experiments showed that under clinical conditions IV and oral MRCA did not yield additional attenuation of PET emission signals. Patient scans showed that PET attenuation maps are not biased after the administration of IV MRCA but may be biased, however, after ingestion of iron oxide-based oral MRCA when segmentation-based AC algorithms are used. Alternative AC algorithms, such as ATPR, or alternative oral contrast agents, such as pineapple juice, can yield unbiased attenuation maps.In clinical PET/MR scenarios MRCA are not expected to lead to markedly increased attenuation of the PET emission signals. MR-based attenuation maps may be biased by oral iron oxide-based MRCA unless advanced AC algorithms are used.

PET/MRI is an emerging dual-modality imaging technology that requires new approaches to PET attenuation correction (AC). We assessed 2 algorithms for whole-body MRI-based AC (MRAC): a basic MR image segmentation algorithm and a method based on atlas registration and pattern recognition (AT&PR). Methods: Eleven patients each underwent a whole-body PET/CT study and a separate multibed whole-body MRI study. The MR image segmentation algorithm uses a combination of image thresholds, Dixon fat–water segmentation, and component analysis to detect the lungs. MR images are segmented into 5 tissue classes (not including bone), and each class is assigned a default linear attenuation value. The AT&PR algorithm uses a database of previously aligned pairs of MRI/CT image volumes. For each patient, these pairs are registered to the patient MRI volume, and machine-learning techniques are used to predict attenuation values on a continuous scale. MRAC methods are compared via the quantitative analysis of AC PET images using volumes of interest in normal organs and on lesions. We assume the PET/CT values after CT-based AC to be the reference standard. Results: In regions of normal physiologic uptake, the average error of the mean standardized uptake value was 14.1% ± 10.2% and 7.7% ± 8.4% for the segmentation and the AT&PR methods, respectively. Lesion-based errors were 7.5% ± 7.9% for the segmentation method and 5.7% ± 4.7% for the AT&PR method. Conclusion: The MRAC method using AT&PR provided better overall PET quantification accuracy than the basic MR image segmentation approach. This better quantification was due to the significantly reduced volume of errors made regarding volumes of interest within or near bones and the slightly reduced volume of errors made regarding areas outside the lungs.

Attenuation correction (AC) in fully integrated positron emission tomography (PET)/magnetic resonance (MR) systems plays a key role for the quantification of tracer uptake. The aim of this prospective study was to assess the accuracy of standardized uptake value (SUV) quantification using MR-based AC in direct comparison with computed tomography (CT)-based AC of the same PET data set on a large patient population.Sixty-six patients (22 female; mean [SD], 61 [11] years) were examined by means of combined PET/CT and PET/MR (11C-choline, 18F-FDG, or 68Ga-DOTATATE) subsequently. Positron emission tomography images from PET/MR examinations were corrected with MR-derived AC based on tissue segmentation (PET(MR)). The same PET data were corrected using CT-based attenuation maps (μ-maps) derived from PET/CT after nonrigid registration of the CT to the MR-based μ-map (PET(MRCT)). Positron emission tomography SUVs were quantified placing regions of interest or volumes of interest in 6 different body regions as well as PET-avid lesions, respectively.The relative differences of quantitative PET values when using MR-based AC versus CT-based AC were varying depending on the organs and body regions assessed. In detail, the mean (SD) relative differences of PET SUVs were as follows: -7.8% (11.5%), blood pool; -3.6% (5.8%), spleen; -4.4% (5.6%)/-4.1% (6.2%), liver; -0.6% (5.0%), muscle; -1.3% (6.3%), fat; -40.0% (18.7%), bone; 1.6% (4.4%), liver lesions; -6.2% (6.8%), bone lesions; and -1.9% (6.2%), soft tissue lesions. In 10 liver lesions, distinct overestimations greater than 5% were found (up to 10%). In addition, overestimations were found in 2 bone lesions and 1 soft tissue lesion adjacent to the lung (up to 28.0%).Results obtained using different PET tracers show that MR-based AC is accurate in most tissue types, with SUV deviations generally of less than 10%. In bone, however, underestimations can be pronounced, potentially leading to inaccurate SUV quantifications. In addition, SUV overestimations were found for some lesions close to lung borders. This has to be taken into account when comparing PET/CT- and PET/MR-derived SUVs.

The aim of this study was to evaluate the clinical applicability and technical feasibility of fluorodeoxyglucose (FDG) positron emission tomography (PET)/magnetic resonance imaging (MRI) compared with FDG PET/computed tomography (CT) in young children focusing on lesion detection, PET quantification, and potential savings in radiation exposure.Twenty examinations (10 PET/CT and 10 PET/MRI examinations) were performed prospectively in 9 patients with solid tumors (3 female, 6 male; mean age, 4.8 [1-6] years). Fluorodeoxyglucose PET/CT and FDG PET/MRI were performed sequentially after a single tracer injection. Lesion detection and analysis were performed independently in PET/CT and PET/MRI. Potential changes in diagnostic or therapeutic patient management were recorded. Positron emission tomography quantification in PET/MRI was evaluated by comparing standardized uptake values resulting from MRI-based and CT-based attenuation correction. Effective radiation doses of PET and CT were estimated.Twenty-one PET-positive lesions were found congruently in PET/CT and PET/MRI. Magnetic resonance imaging enabled significantly better detection of morphologic PET correlates compared with CT. Eight suspicious PET-negative lesions were identified by MRI, of which one was missed in CT. Sensitivity, specificity, and accuracy for correct lesion classification were not significantly different (90%, 47%, and 62% in PET/CT; 100%, 68%, and 79% in PET/MRI, respectively). In 4 patients, the use of PET/MRI resulted in a potential change in diagnostic management compared with PET/CT, as local and whole-body staging could be performed within 1 single examination. In 1 patient, PET/MRI initiated a change in therapeutic management. Positron emission tomography quantification using MRI-based attenuation correction was accurate compared with CT-based attenuation correction. Higher standardized uptake value deviations of about 18% were observed in the lungs due to misclassification in MRI-based attenuation maps. Potential reduction in radiation dose was 48% in PET/MRI compared with PET/CT (P0.05).FDG PET/MRI is at least equivalent to FDG PET/CT for oncologic imaging in young children. Specifically, superior soft tissue contrast of MRI results in higher confidence in lesion interpretation. Substantial savings in radiation exposure can be achieved, and the number of necessary imaging examinations can be reduced using PET/MRI compared with PET/CT.

Zielsetzung: Ziel dieser prospektiven Studie war die Untersuchung der technischen Machbarkeit und diagnostischen Gute der kombinierten FDG-PET/MRT zur onkologischen Diagnostik bei Kleinkindern. Dies erfolgte im direkten Vergleich zur FDG-PET/CT. Material und Methodik: 10 Doppeluntersuchungen wurden an 9 Patienten (4 Madchen; 3,8 ± 1,8J.) durchgefuhrt. Zunachst erfolgte die PET/CT 60 min nach Injektion von 115 ± 13 MBq 18F-FDG. Im Anschluss wurde 120 min nach FDG-Injektion die PET/MRT durchgefuhrt. Neben einer T1-Gradientenecho-Sequenz zur PET-Schwachungskorrektur wurden folgende MR-Sequenzen akquiriert: GK-STIR cor, GK-DWI, axiale STIR des Halses/Thorax, axiale T2 TSE des Abdomens/Beckens, axiale GK-T1-Flash-3D. PET/CT und PET/MRT wurden unabhangig voneinander durch zwei Befunderpaare klinisch ausgewertet. Zudem erfolgte eine quantitative Auswertung der PET-Datensatze hinsichtlich Genauigkeit der PET-Quantifizierung. Eff. Dosen wurden anhand des Dosislangenproduktes und der applizierten FDG-Dosis errechnet. Ergebnisse: Insgesamt wurden mittels PET-CT bzw PET-MRT jeweils 21 PET-positive und 7 bzw 8 PET-negative Lasionen detektiert. Dabei konnte mittels CT bzw MRT in 13 bzw 20 von 21 Fallen ein sicheres morphologisches PET- Korrelat identifiziert werden. In 5 von 10 Untersuchungen fuhrte die PET/MRT im Vergleich zur PET/CT zu einer potentiellen Anderung des weiteren klinischen Vorgehens. Die quantitative Auswertung der PET-Komponente ergab einen hohen Grad der Ubereinstimmung der PET-SUVs zwischen PET/CT und PET/MRT. Lediglich intrapulmonal und intraossar ergaben sich relevante Abweichungen der SUV-Werte um 10 bis 15%, was auf Ungenauigkeiten der MR-basierten PET-Schwachungskarte zuruckzufuhren war. Die theoretische Reduktion der eff. Dosis betrug ca 48%, Schlussfolgerungen: Die FDG-PET/MRT ist in der onkologischen Diagnostik bei Kleinkindern zuverlassig durchfuhrbar und zeigt zur PET/CT diagnostisch gleichwertige Ergebnisse mit Vorteilen in der Detektion von Weichteillasionen. Dabei ist mindestens eine Halbierung der eff. Dosis moglich.

OBJECTIVES The objective of this study was to evaluate the frequency and characteristics of artifacts in segmentation-based attenuation correction maps (μ-maps) of positron emission tomography/magnetic resonance (PET/MR) and their impact on PET interpretation and the standardized uptake value (SUV) quantification in normal tissue and lesions. MATERIALS AND METHODS The study was approved by the local institutional review board. Attenuation maps of 100 patients with PET/MR and preceding PET/computed tomography examination were retrospectively inspected for artifacts (tracers: 2-deoxy-2-[¹⁸F]fluoro-D-glucose (¹⁸F-FDG), ¹¹C-Choline, ⁶⁸Ga-DOTATOC, ⁶⁸Ga-DOTATATE, ¹¹C-Methionine). The artifacts were subdivided into 9 different groups on the basis of their localization and appearance. The impact of μ-map artifacts in normal tissue and lesions on PET interpretation was evaluated qualitatively via visual analysis in synopsis with the non-attenuation-corrected (NAC) PET as well as quantitatively by comparing the SUV in artifact regions to reference regions. RESULTS Attenuation map artifacts were found in 72% of the head/neck data sets, 61% of the thoracic data sets, 25% of the upper abdominal data sets, and 26% of the pelvic data sets. The most frequent localizations of the overall 276 artifacts were around metal implants (16%), in the lungs (19%), and outer body contours (31%). Twenty-one percent of all PET-avid lesions (38 of 184 lesions) were affected by artifacts in the majority without further consequences for visual PET interpretation. However, 9 PET-avid lung lesions were masked owing to μ-map artifacts and, thus, were only detectable on the NAC PET or additional MR imaging sequences. Quantitatively, μ-map artifacts led to significant SUV changes in areas with erroneous assignment of air instead of soft tissue (ie, metal artifacts) and of soft tissue instead of lung. Nevertheless, no change in diagnosis would have been caused by μ-map artifacts. CONCLUSIONS Attenuation map artifacts that occur in a considerable percentage of PET/MR data sets have the potential to falsify PET quantification and visual PET interpretation. Nevertheless, on the basis of the present data, in the clinical interpretation setup, no changes in diagnosis due to μ-map artifacts may occur, especially when the μ-maps are checked for artifacts and PET/MR is read in synopsis with the NAC PET, if artifacts are present.

Especially for neuroscience and the development of new biomarkers, a direct correlation between in vivo imaging and histology is essential. However, this comparison is hampered by deformation and shrinkage of tissue samples caused by fixation, dehydration and paraffin embedding. We used magnetic resonance (MR) imaging and computed tomography (CT) imaging to analyze the degree of shrinkage on murine brains for various fixatives. After in vivo imaging using 7 T MRI, animals were sacrificed and the brains were dissected and immediately placed in different fixatives, respectively: zinc-based fixative, neutral buffered formalin (NBF), paraformaldehyde (PFA), Bouin-Holland fixative and paraformaldehyde-lysine-periodate (PLP). The degree of shrinkage based on mouse brain volumes, radiodensity in Hounsfield units (HU), as well as non-linear deformations were obtained. The highest degree of shrinkage was observed for PLP (68.1%, P0.001), followed by PFA (60.2%, P0.001) and NBF (58.6%, P0.001). The zinc-based fixative revealed a low shrinkage with only 33.5% (P0.001). Compared to NBF, the zinc-based fixative shows a slightly higher degree of deformations, but is still more homogenous than PFA. Tissue shrinkage can be monitored non-invasively with CT and MR. Zinc-based fixative causes the smallest degree of brain shrinkage and only small deformations and is therefore recommended for in vivo ex vivo comparison studies.

Multiparametric medical imaging data can be large and are often complex. Machine learning algorithms can assist in image interpretation when reliable training data exist. In most cases, however, knowledge about ground truth (e.g. histology) and thus training data is limited, which makes application of machine learning algorithms difficult. The purpose of this study was to design and implement a learning algorithm for classification of multidimensional medical imaging data that is robust and accurate even with limited prior knowledge and that allows for generalization and application to unseen data. Local prostate cancer was chosen as a model for application and validation. 16 patients underwent combined simultaneous [(11) C]-choline positron emission tomography (PET)/MRI. The following imaging parameters were acquired: T2 signal intensities, apparent diffusion coefficients, parameters Ktrans and Kep from dynamic contrast-enhanced MRI, and PET standardized uptake values (SUVs). A spatially constrained fuzzy c-means algorithm (sFCM) was applied to the single datasets and the resulting labeled data were used for training of a support vector machine (SVM) classifier. Accuracy and false positive and false negative rates of the proposed algorithm were determined in comparison with manual tumor delineation. For five of the 16 patients rates were also determined in comparison with the histopathological standard of reference. The combined sFCM/SVM algorithm proposed in this study revealed reliable classification results consistent with the histopathological reference standard and comparable to those of manual tumor delineation. sFCM/SVM generally performed better than unsupervised sFCM alone. We observed an improvement in accuracy with increasing number of imaging parameters used for clustering and SVM training. In particular, including PET SUVs as an additional parameter markedly improved classification results. A variety of applications are conceivable, especially for imaging of tissues without easily available histopathological correlation.

PET Attenuation Correction (AC) in small animal imaging is generally performed via transmission (TX) scans. In simultaneous PET/MR scanners, no transmission source is available, and MR-based AC (MRAC) or AC from emission data is necessary. However, MR-invisible objects such as the bed and the MR coils cannot be corrected from MR data. We quantified the effects of a Medres warming bed and a local brain coil on AC for small animal PET/MR imaging using a dedicated Inveon PET tomograph. Additionally, the effect of segmentation-based MRAC of rats in brain regions was evaluated using MR images separately acquired on a preclinical 7T MRI system.

Hybrid PET/MR systems have recently entered clinical practice. Thus, the accuracy of MR-based attenuation correction in simultaneously acquired data can now be investigated. We assessed the accuracy of 4 methods of MR-based attenuation correction in lesions within soft tissue, bone, and MR susceptibility artifacts: 2 segmentation-based methods (SEG1, provided by the manufacturer, and SEG2, a method with atlas-based susceptibility artifact correction); an atlas- and pattern recognition-based method (ATPR), which also used artifact correction; and a new method combining ATPR and SEG2 (SEG2wBONE).Attenuation maps were calculated for the PET/MR datasets of 10 patients acquired on a whole-body PET/MR system, allowing for simultaneous acquisition of PET and MR data. Eighty percent iso-contour volumes of interest were placed on lesions in soft tissue (n = 21), in bone (n = 20), near bone (n = 19), and within or near MR susceptibility artifacts (n = 9). Relative mean volume-of-interest differences were calculated with CT-based attenuation correction as a reference.For soft-tissue lesions, none of the methods revealed a significant difference in PET standardized uptake value relative to CT-based attenuation correction (SEG1, -2.6% ± 5.8%; SEG2, -1.6% ± 4.9%; ATPR, -4.7% ± 6.5%; SEG2wBONE, 0.2% ± 5.3%). For bone lesions, underestimation of PET standardized uptake values was found for all methods, with minimized error for the atlas-based approaches (SEG1, -16.1% ± 9.7%; SEG2, -11.0% ± 6.7%; ATPR, -6.6% ± 5.0%; SEG2wBONE, -4.7% ± 4.4%). For lesions near bone, underestimations of lower magnitude were observed (SEG1, -12.0% ± 7.4%; SEG2, -9.2% ± 6.5%; ATPR, -4.6% ± 7.8%; SEG2wBONE, -4.2% ± 6.2%). For lesions affected by MR susceptibility artifacts, quantification errors could be reduced using the atlas-based artifact correction (SEG1, -54.0% ± 38.4%; SEG2, -15.0% ± 12.2%; ATPR, -4.1% ± 11.2%; SEG2wBONE, 0.6% ± 11.1%).For soft-tissue lesions, none of the evaluated methods showed statistically significant errors. For bone lesions, significant underestimations of -16% and -11% occurred for methods in which bone tissue was ignored (SEG1 and SEG2). In the present attenuation correction schemes, uncorrected MR susceptibility artifacts typically result in reduced attenuation values, potentially leading to highly reduced PET standardized uptake values, rendering lesions indistinguishable from background. While ATPR and SEG2wBONE show accurate results in both soft tissue and bone, SEG2wBONE uses a two-step approach for tissue classification, which increases the robustness of prediction and can be applied retrospectively if more precision in bone areas is needed.

Introduction There has recently been a growing interest in combining PET and MR. While promising hardware prototypes have been presented [1], the software for combined imaging is in its infancy. PET attenuation correction (AC), which accounts for radiation attenuation properties of the tissue, is mandatory for quantitative PET. Usually the attenuation map in stand-alone PET or PET/CT is obtained from a transmission scan or from a CT image. In the case of PET/MR, there is insufficient space for the rotating source so that the tissue attenuation information needs to be determined from the MR image. This is inherently difficult: The PET attenuation value is not related to the proton density and tissue relaxation properties expressed in the MR intensities. For example bone compacta and air both typically do not contribute a MR signal, whereas their attenuation values are maximally distinct. Here, we report our progress on evaluating suitable MR sequences and methodology for calculating the attenuation images (“Pseudo-CT”) based on the MR image. We present results for head and whole body imaging. To our knowledge, our contribution is the first to show the use of ultra short echo time (UTE) sequences for AC in clinical examples. It is also among the first to show results for human whole body AC.

Data preprocessing and a careful selection of the training and classification method are key steps for building a predictive or classification model with high performance. Here, we present the winner approaches submitted to the 2005 GfKl Data Mining Competition. The task to be solved for the competition was the prediction of a possible liquidity crisis of a company. The binary classification was to be based on a set of 26 variables describing attributes of the companies with unknown semantics.