Thirty Years of the DICOM Standard.

Digital Imaging and Communications in Medicine (DICOM) is an international standard that defines a format for storing medical images and a protocol to enable and facilitate data communication among medical imaging systems. The DICOM standard has been instrumental in transforming the medical imaging world over the last three decades. Its adoption has been a significant experience for manufacturers, healthcare users, and research scientists. In this review, thirty years after introducing the standard, we discuss the innovation, advantages, and limitations of adopting the DICOM and its possible future directions.

Comparison of three freeware software packages for 18F-FDG PET texture feature calculation.

PURPOSE: To compare texture feature estimates obtained from 18F-FDG-PET images using three different software packages. METHODS: PET images from 15 patients with head and neck cancer were processed with three different freeware software: CGITA, LIFEx, and Metavol. For each lesion, 38 texture features were extracted from each software package. To evaluate the statistical agreement among the features across packages a non-parametric Kruskal-Wallis test was used. Differences in the features between each couple of software were assessed using a subsequent Dunn test. Correlation between texture features was evaluated via the Spearman coefficient. RESULTS: Twenty-three of 38 features showed a significant agreement across the three software (P < 0.05). The agreement was better between LIFEx vs. Metavol (36 of 38) and worse between CGITA and Metavol (24 of 38), and CGITA vs. LIFEx (23 of 38). All features resulted correlated (ρ > = 0.70, P < 0.001) in comparing LIFEx vs. Metavol. Seven of 38 features were found not in agreement and slightly or not correlated (ρ < 0.70, P < 0.001) in comparing CGITA vs. LIFEx, and CGITA vs. Metavol. CONCLUSION: Some texture discrepancies across software packages exist. Our findings reinforce the need to continue the standardization process, and to succeed in building a reference dataset to be used for comparisons.

Integration of stimulated Raman gain and stimulated Raman losses detection modes in a single nonlinear microscope.

This work addresses some key challenges in the fields of bio and nanophotonics by stimulated Raman microscopy. We present the design and the implementation of a femtosecond stimulated Raman scattering microscope, equipped with three femtosecond laser sources, which allows the coexistence of stimulated Raman gain (SRG) and stimulate Raman losses (SRL) detection modes in a single microscopy setup and to generate images of the same region in succession, without adding or removing components. In order to demonstrate the switching between the two detection modes, SRL and SRG images of polystyrene beads are acquired and the images quality are evaluated and compared.

Self-Trained Supervised Segmentation of Subcortical Brain Structures Using Multispectral Magnetic Resonance Images.

The aim of this paper is investigate the feasibility of automatically training supervised methods, such as k-nearest neighbor (kNN) and principal component discriminant analysis (PCDA), and to segment the four subcortical brain structures: caudate, thalamus, pallidum, and putamen. The adoption of supervised classification methods so far has been limited by the need to define a representative training dataset, operation that usually requires the intervention of an operator. In this work the selection of the training data was performed on the subject to be segmented in a fully automated manner by registering probabilistic atlases. Evaluation of automatically trained kNN and PCDA classifiers that combine voxel intensities and spatial coordinates was performed on 20 real datasets selected from two publicly available sources of multispectral magnetic resonance studies. The results demonstrate that atlas-guided training is an effective way to automatically define a representative and reliable training dataset, thus giving supervised methods the chance to successfully segment magnetic resonance brain images without the need for user interaction.

Genetic deletion in uncoupling protein 3 augments 18F-fluorodeoxyglucose cardiac uptake in the ischemic heart.

BACKGROUND: We investigated the effects of uncoupling protein 3 (UCP3) genetic deletion on 18F-fluorodeoxyglucose (FDG) cardiac uptake by positron emission tomography (PET)/computed tomography (CT) dedicated animal system after permanent coronary artery ligation. METHODS: Cardiac 18F-FDG PET/CT was performed in UCP3 knockout (UCP3-/-) and wild-type (WT) mice one week after induction of myocardial infarction or sham procedure. RESULTS: In sham-operated mice no difference in left ventricular (LV) volume was detectable between WT and UCP3-/-. After myocardial infarction, LV volume was higher in both WT and UCP3-/- compared to sham animals, with a significant interaction (p < 0.05) between genotype and myocardial infarction. In sham-operated animals no difference in FDG standardized uptake value (SUV) was detectable between WT (1.8 ± 0.6) and UCP3-/- (1.8 ± 0.6). After myocardial infarction SUV was significantly higher in remote areas than in infarcted territories in both UCP3-/- and WT mice (both p < 0.01). Moreover, in remote areas, SUV was significantly higher (p < 0.001) in UCP3-/- as compared to WT, while in the infarcted territory SUV was comparable (p = 0.29). A significant relationship (r = 0.68, p < 0.001) between LV volume and SUV was found. CONCLUSIONS: In a mice model of permanent coronary occlusion, UCP3 deficiency results in a metabolic shift that favored glycolytic metabolism and increased FDG uptake in remote areas.

Evaluation of supervised methods for the classification of major tissues and subcortical structures in multispectral brain magnetic resonance images.

This work investigates the capability of supervised classification methods in detecting both major tissues and subcortical structures using multispectral brain magnetic resonance images. First, by means of a realistic digital brain phantom, we investigated the classification performance of various Discriminant Analysis methods, K-Nearest Neighbor and Support Vector Machine. Then, using phantom and real data, we quantitatively assessed the benefits of integrating anatomical information in the classification, in the form of voxels coordinates as additional features to the intensities or tissue probabilistic atlases as priors. In addition we tested the effect of spatial correlations between neighboring voxels and image denoising. For each brain tissue we measured the classification performance in terms of global agreement percentage, false positive and false negative rates and kappa coefficient. The effectiveness of integrating spatial information or a tissue probabilistic atlas has been demonstrated for the aim of accurately classifying brain magnetic resonance images.

Medical image file formats.

Image file format is often a confusing aspect for someone wishing to process medical images. This article presents a demystifying overview of the major file formats currently used in medical imaging: Analyze, Neuroimaging Informatics Technology Initiative (Nifti), Minc, and Digital Imaging and Communications in Medicine (Dicom). Concepts common to all file formats, such as pixel depth, photometric interpretation, metadata, and pixel data, are first presented. Then, the characteristics and strengths of the various formats are discussed. The review concludes with some predictive considerations about the future trends in medical image file formats.

Metabolic tumor volume assessed by 18F-FDG PET/CT for the prediction of outcome in patients with multiple myeloma.

UNLABELLED: (18)F-FDG PET/CT allows the direct measurement of metabolic tumor burden in a variety of different malignancies. The aim of this study was to assess whether metabolic tumor volume (MTV) determined by (18)F-FDG PET/CT could be used in the prediction of progression-free and overall survival in multiple myeloma patients. METHODS: Forty-seven patients (18 women, 29 men; mean age ± SD, 63 ± 11 y) with stage IIIA disease who had undergone whole-body (18)F-FDG PET/CT were retrospectively evaluated. Images underwent a 3-dimensional region-of-interest analysis including all focal lesions with a maximum standardized uptake value > 2.5. The MTV of each lesion was calculated using an automated contouring program based on the standardized uptake value and developed with a threshold of 40% of the maximum standardized uptake value. The total MTV of each patient was defined as the sum of metabolic volume of all focal lesions. Patients were treated and then subjected to a mean follow-up period of 24 mo. RESULTS: In the 47 patients studied, MTV range was 1.3-316.3 mL, with a median of 23.7 mL. A direct, significant correlation was found between MTV and the percentage of diffuse infiltration of bone marrow by plasma cells (r = 0.46, P = 0.006), whereas hemoglobin levels were inversely correlated with MTV (r = -0.56, P = 0.0001). At follow-up, patients who developed progressive disease (n = 18) showed a significantly higher MTV (74.7 ± 19.3 vs. 29.8 ± 5.1 mL, P = 0.009) than patients without progressive disease (n = 29). Furthermore, patients who died of myeloma (n = 9) had a significantly higher MTV (123.2 ± 30.6 vs. 28.9 ± 4.2 mL, P = 0.0001) than survivors (n = 38). No differences in age, plasma cell infiltration, M protein, albumin, ß2-microglobulin, performance status, International Staging System score, and presence or absence of a bone marrow transplant were found between groups. The MTV cutoff level was determined by receiver-operating-characteristic curve analysis, and the best discriminative value found for predicting progression-free and overall survival was 42.2 and 77.6 mL, respectively. By Kaplan-Meier analysis and log-rank testing, progression-free and overall survival at follow-up were significantly better in patients showing an MTV lower than the cutoff than in those having an MTV higher than the cutoff (χ(2) = 3.9, P = 0.04, and χ(2) = 56.3, P < 0.0001, respectively). CONCLUSION: The direct measurement of tumor burden obtained by calculating MTV on (18)F-FDG PET/CT images may be used in the prediction of progression-free and overall survival in myeloma patients.

PET/CT imaging in mouse models of myocardial ischemia.

Different species have been used to reproduce myocardial infarction models but in the last years mice became the animals of choice for the analysis of several diseases, due to their short life cycle and the possibility of genetic manipulation. Many techniques are currently used for cardiovascular imaging in mice, including X-ray computed tomography (CT), high-resolution ultrasound, magnetic resonance imaging, and nuclear medicine procedures. Cardiac positron emission tomography (PET) allows to examine noninvasively, on a molecular level and with high sensitivity, regional changes in myocardial perfusion, metabolism, apoptosis, inflammation, and gene expression or to measure changes in anatomical and functional parameters in heart diseases. Currently hybrid PET/CT scanners for small laboratory animals are available, where CT adds high-resolution anatomical information. This paper reviews mouse models of myocardial infarction and discusses the applications of dedicated PET/CT systems technology, including animal preparation, anesthesia, radiotracers, and images postprocessing.

Reproducibility and accuracy of non-invasive measurement of infarct size in mice with high-resolution PET/CT.

BACKGROUND: This study assessed the reproducibility and accuracy of 2-deoxy-2[(18)F]fluoro-D-glucose ((18)F-FDG) for non-invasive quantification of myocardial infarct size in mice by a high-resolution positron emission tomography (PET)/computed tomography (CT) system. METHODS AND RESULTS: Mice were studied by (18)F-FDG PET/CT 1 week after induction of myocardial infarction by permanent coronary occlusion or sham procedure. In a subset of mice, PET/CT was repeated 2 days apart to assess the reproducibility of infarct size measurements. Histological analysis was used as reference method to validate imaging data. The average difference in infarct size measurements between the first and the second study was -0.42% ± 2.07% (95% confidence interval -2.6 to 1.75) with a repeatability coefficient of 4.05%. At Bland-Altman analysis, the lower and upper limits of agreement between the two repeated studies were -4.46% and 3.63%, respectively, and no correlation between difference and mean was found (P = .89). The concordance correlation coefficient was 0.99 (P < .001) and the intraclass coefficient of correlation was 0.99. A high correlation between PET/CT and histology was found for measurement of infarct size (P < .001). Using Bland-Altman analysis, the mean difference in infarct size measurement (PET/CT minus histology) was 1.9% (95% confidence interval 0.94% to 2.86%). CONCLUSIONS: In a mice model of permanent coronary occlusion non-invasive measurement of infarct size with high-resolution (18)F-FDG, PET/CT has excellent reproducibility and accuracy. These findings support the use of this methodology in serial studies.

Descent algorithms on oblique manifold for source-adaptive ICA contrast.

A Riemannian manifold optimization strategy is proposed to facilitate the relaxation of the orthonormality constraint in a more natural way in the course of performing independent component analysis (ICA) that employs a mutual information-based source-adaptive contrast function. Despite the extensive development of manifold techniques catering to the orthonormality constraint, only a limited number of works have been dedicated to oblique manifold (OB) algorithms to intrinsically handle the normality constraint, which has been empirically shown to be superior to other Riemannian and Euclidean approaches. Imposing the normality constraint implicitly, in line with the ICA definition, essentially guarantees a substantial improvement in the solution accuracy, by way of increased degrees of freedom while searching for an optimal unmixing ICA matrix, in contrast with the orthonormality constraint. Designs of the steepest descent, conjugate gradient with Hager-Zhang or a hybrid update parameter, quasi-Newton, and cost-effective quasi-Newton methods intended for OB are presented in this paper. Their performance is validated using natural images and systematically compared with the popular state-of-the-art approaches in order to assess the performance effects of the choice of algorithm and the use of a Riemannian rather than Euclidean framework. We surmount the computational challenge associated with the direct estimation of the source densities using the improved fast Gauss transform in the evaluation of the contrast function and its gradient. The proposed OB schemes may find applications in the offline image/signal analysis, wherein, on one hand, the computational overhead can be tolerated, and, on the other, the solution quality holds paramount interest.

An MRI digital brain phantom for validation of segmentation methods.

Knowledge of the exact spatial distribution of brain tissues in images acquired by magnetic resonance imaging (MRI) is necessary to measure and compare the performance of segmentation algorithms. Currently available physical phantoms do not satisfy this requirement. State-of-the-art digital brain phantoms also fall short because they do not handle separately anatomical structures (e.g. basal ganglia) and provide relatively rough simulations of tissue fine structure and inhomogeneity. We present a software procedure for the construction of a realistic MRI digital brain phantom. The phantom consists of hydrogen nuclear magnetic resonance spin-lattice relaxation rate (R1), spin-spin relaxation rate (R2), and proton density (PD) values for a 24 × 19 × 15.5 cm volume of a "normal" head. The phantom includes 17 normal tissues, each characterized by both mean value and variations in R1, R2, and PD. In addition, an optional tissue class for multiple sclerosis (MS) lesions is simulated. The phantom was used to create realistic magnetic resonance (MR) images of the brain using simulated conventional spin-echo (CSE) and fast field-echo (FFE) sequences. Results of mono-parametric segmentation of simulations of sequences with different noise and slice thickness are presented as an example of possible applications of the phantom. The phantom data and simulated images are available online at http://lab.ibb.cnr.it/.

In vivo targeting and growth inhibition of the A20 murine B-cell lymphoma by an idiotype-specific peptide binder.

B-cell lymphoma is a clonal expansion of neoplastic cells that may result in fatal outcomes. Here, we report the in vivo targeting and growth inhibition of aggressive A20 murine B-cell lymphoma by idiotype-specific peptide pA20-36. pA20-36 was selected from random peptide libraries and bound specifically to the B-cell receptor (BCR) of A20 cells in mice engrafted with A20 lymphoma, as shown by histology and positron emission tomographic analysis. BCR cross-linking of A20 cells with pA20-36 resulted in massive apoptosis of targeted tumor cells and in an increased survival of the diseased animals without any detectable evidence of toxicity. The pA20-36 treatment reverted the immune suppression of the tumor microenvironment as shown by reduced expression of vascular endothelial growth factor, interleukin-10, and transforming growth factor-beta cytokines together with a lower number of CD11b+Gr-1+ inhibitor myeloid-derived suppressor cells and Foxp3+CD4+ Treg cells. Furthermore, pA20-36 treatment was associated with an increased number of tumor-infiltrating, activated CD8+ T cells that exerted a tumor-specific cytolytic activity. These findings show that a short peptide that binds specifically to the complementarity-determining regions of the A20 BCR allows in vivo detection of neoplastic cells together with significant inhibition of tumor growth in vivo.

Segmentation of magnetic resonance brain images through discriminant analysis.

Segmentation (tissue classification) of medical images obtained from a magnetic resonance (MR) system is a primary step in most applications of medical image post-processing. This paper describes nonparametric discriminant analysis methods to segment multispectral MR images of the brain. Starting from routinely available spin-lattice relaxation time, spin-spin relaxation time, and proton density weighted images (T1w, T2w, PDw), the proposed family of statistical methods is based on: (i) a transform of the images into components that are statistically independent from each other; (ii) a nonparametric estimate of probability density functions of each tissue starting from a training set; (iii) a classic Bayes 0-1 classification rule. Experiments based on a computer built brain phantom (brainweb) and on eight real patient data sets are shown. A comparison with parametric discriminant analysis is also reported. The capability of nonparametric discriminant analysis in improving brain tissue classification of parametric methods is demonstrated. Finally, an assessment of the role of multispectrality in classifying brain tissues is discussed.

Brain tissue volume changes in relapsing-remitting multiple sclerosis: correlation with lesion load.

The aim of this study was to simultaneously measure in vivo volumes of gray matter (GM), normal white matter (WM), abnormal white matter (aWM), and cerebro-spinal fluid (CSF), and to assess their relationship in 50 patients with relapsing-remitting multiple sclerosis (RR-MS) (age range, 21-59; mean EDSS, 2.5; mean disease duration, 9.9 years), using an unsupervised multiparametric segmentation procedure applied to brain MR studies. Tissue volumes were normalized to total intracranial volume providing corresponding fractional volumes (fGM, faWM, fWM, and fCSF), subsequently corrected for aWM-related segmentation inaccuracies and adjusted to mean patients' age according to age-related changes measured in 54 normal volunteers (NV) (age range 16-70). In MS patients aWM was 23.8 +/- 29.8 ml (range 0.4-138.8). A significant decrease in fGM was present in MS patients as compared to NV (49.5 +/- 3.2% vs 53.3 +/- 2.1%; P < 0.0001), with a corresponding increase in fCSF (13.0 +/- 3.8% vs 9.1 +/- 2.4%; P < 0.0001). No difference could be detected between the two groups for fWM (37.5 +/- 2.6% vs 37.6 +/- 2.2%). faWM correlated inversely with fGM (R = -0.434, P < 0.001 at regression analysis), and directly with fCSF (R = 0.473, P < 0.001), but not with fWM. There was a significant correlation between disease duration and EDSS, while no relationship was found between EDSS or disease duration and fractional volumes. Brain atrophy in RR-MS is mainly related to GM loss, which correlates with faWM. Both measures do not appear to significantly affect EDSS, which correlates to disease duration.

Stereotaxy-based regional brain volumetry applied to segmented MRI: validation and results in deficit and nondeficit schizophrenia.

A method for postprocessing of segmented routine brain MRI studies providing automated definition of major structures (frontal, parietal, occipital, and temporal lobes; cerebellar hemispheres; and lateral ventricles) according to the Talairach atlas is presented. The method was applied to MRI studies from 25 normal subjects (NV), 14 patients with deficit schizophrenia (DS), and 14 with nondeficit schizophrenia (NDS), to evaluate their gray matter and CSF regional volumes. The two patient groups did not differ in mean age at illness onset, duration of illness, severity of psychotic symptoms, or disorganization; DS had more severe avolition and worse social functioning than NDS. For validation purposes, brain structures were manually outlined on original MR images in 10 studies, thus obtaining reference measures. Manual and automated measures were repeated 1 month apart to measure reproducibilities of both methods. The automated method required less than 1 min/operator per study vs more than 30 min for manual assessment. Mean absolute difference per structure between the two techniques was 4.8 ml. Overall reproducibility did not significantly differ between the two methods. In subjects with schizophrenia, a significant decrease in GM and increase in CSF were found. GM loss was confined to frontal and temporal lobes. Lateral ventricles were significantly larger bilaterally in NDS compared to NV and only on the right in NDS compared to DS. The finding of greater structural brain abnormalities in NDS adds to the evidence that deficit schizophrenia does not represent just the more severe end of the schizophrenia continuum.