Clinical evaluation of deep learning-enhanced lymphoma pet imaging with accelerated acquisition.

PURPOSE: This study aims to evaluate the clinical performance of a deep learning (DL)-enhanced two-fold accelerated PET imaging method in patients with lymphoma. METHODS: A total of 123 cases devoid of lymphoma underwent whole-body 18F-FDG-PET/CT scans to facilitate the development of an advanced SAU2Net model, which combines the advantages of U2Net and attention mechanism. This model integrated inputs from simulated 1/2-dose (0.07 mCi/kg) PET acquisition across multiple slices to generate an estimated standard dose (0.14 mCi/kg) PET scan. Additional 39 cases with confirmed lymphoma pathology were utilized to evaluate the model's clinical performance. Assessment criteria encompassed peak-signal-to-noise ratio (PSNR), structural similarity index (SSIM), a 5-point Likert scale rated by two experienced physicians, SUV features, image noise in the liver, and contrast-to-noise ratio (CNR). Diagnostic outcomes, including lesion numbers and Deauville score, were also compared. RESULTS: Images enhanced by the proposed DL method exhibited superior image quality (P < 0.001) in comparison to low-dose acquisition. Moreover, they illustrated equivalent image quality in terms of subjective image analysis and lesion maximum standardized uptake value (SUVmax) as compared to the standard acquisition method. A linear regression model with y = 1.017x + 0.110 ( R 2 = 1.00 ${R^2} = \;1.00$ ) can be established between the enhanced scans and the standard acquisition for lesion SUVmax. With enhancement, increased signal-to-noise ratio (SNR), CNR, and reduced image noise were observed, surpassing those of the standard acquisition. DL-enhanced PET images got diagnostic results essentially equavalent to standard PET images according to two experienced readers. CONCLUSION: The proposed DL method could facilitate a 50% reduction in PET imaging duration for lymphoma patients, while concurrently preserving image quality and diagnostic accuracy.

Fast SPECT/CT planar bone imaging enabled by deep learning enhancement.

BACKGROUND: The application of deep learning methods in rapid bone scintigraphy is increasingly promising for minimizing the duration of SPECT examinations. Recent works showed several deep learning models based on simulated data for the synthesis of high-count bone scintigraphy images from low-count counterparts. Few studies have been conducted and validated on real clinical pairs due to the misalignment inherent in multiple scan procedures. PURPOSE: To generate high quality whole-body bone images from 2× and 3× fast scans using deep learning based enhancement method. MATERIALS AND METHODS: Seventy-six cases who underwent whole-body bone scans were enrolled in this prospective study. All patients went through a standard scan at a speed of 20 cm/min, which followed by fast scans consisting of 2× and 3× accelerations at speeds of 40 and 60 cm/min. A content-attention image restoration approach based on Residual-in-Residual Dense Block (RRDB) is introduced to effectively recover high-quality images from fast scans with fine-details and less noise. Our approach is robust with misalignment introduced from patient's metabolism, and shows valid count-level consistency. Learned Perceptual Image Patch Similarity (LPIPS) and Fréchet Inception Distance (FID) are employed in evaluating the similarity to the standard bone images. To further prove our method practical in clinical settings, image quality of the anonymous images was evaluated by two experienced nuclear physicians on a 5-point Likert scale (5 =  excellent) . RESULTS: The proposed method reaches the state-of-the-art performance on FID and LPIPS with 0.583 and 0.176 for 2× fast scans and 0.583 and 0.185 for 3× fast scans. Clinic evaluation further demonstrated the restored images had a significant improvement compared to fast scan in image quality, technetium 99m-methyl diphosphonate (Tc-99 m MDP) distribution, artifacts, and diagnostic confidence. CONCLUSIONS: Our method was validated for accelerating whole-body bone scans by introducing real clinical data. Confirmed by nuclear medicine physicians, the proposed method can effectively enhance image diagnostic value, demonstrating potential for efficient high-quality fast bone imaging in practical settings.

Deep learning enhanced ultra-fast SPECT/CT bone scan in patients with suspected malignancy: quantitative assessment and clinical performance.

Objectives. To evaluate the clinical performance of deep learning-enhanced ultrafast single photon emission computed tomography/computed tomography (SPECT/CT) bone scans in patients with suspected malignancy.Approach. In this prospective study, 102 patients with potential malignancy were enrolled and underwent a 20 min SPECT/CT and a 3 min SPECT scan. A deep learning model was applied to generate algorithm-enhanced images (3 min DL SPECT). The reference modality was the 20 min SPECT/CT scan. Two reviewers independently evaluated general image quality, Tc-99m MDP distribution, artifacts, and diagnostic confidence of 20 min SPECT/CT, 3 min SPECT/CT, and 3 min DL SPECT/CT images. The sensitivity, specificity, accuracy, and interobserver agreement were calculated. The lesion maximum standard uptake value (SUVmax) of the 3 min DL and 20 min SPECT/CT images was analyzed. The peak signal-to-noise ratio (PSNR) and structure similarity index measure (SSIM) were evaluated.Main results. The 3 min DL SPECT/CT images showed significantly superior general image quality, Tc-99m MDP distribution, artifacts, and diagnostic confidence than the 20 min SPECT/CT images (P< 0.0001). The diagnostic performance of the 20 min and 3 min DL SPECT/CT images was similar for reviewer 1 (pairedX2= 0.333,P= 0.564) and reviewer 2 (pairedX2= 0.05,P= 0.823). The diagnosis results for the 20 min (kappa = 0.822) and 3 min DL (kappa = 0.732) SPECT/CT images showed high interobserver agreement. The 3 min DL SPECT/CT images had significantly higher PSNR and SSIM than the 3 min SPECT/CT images (51.44 versus 38.44,P< 0.0001; 0.863 versus 0.752,P< 0.0001). The SUVmaxof the 3 min DL and 20 min SPECT/CT images showed a strong linear relationship (r= 0.991;P< 0.0001).Significance.Ultrafast SPECT/CT with a 1/7 acquisition time can be enhanced by a deep learning method to achieve comparable image quality and diagnostic value to those of standard acquisition.

Clinical ultra-high resolution CT scans enabled by using a generative adversarial network.

BACKGROUND: Ultra-high resolution computed tomography (UHRCT) has shown great potential for the detection of pulmonary diseases. However, UHRCT scanning generally induces increases in scanning time and radiation exposure. Super resolution is a gradually prosperous application in CT imaging despite higher radiation dose. Recent works have proved that the convolution neural network especially the generative adversarial network (GAN) based model could generate high-resolution CT using phantom images or simulated low resolution data without extra dose. Research that used clinical CT particularly lung images are rare due to the difficulty in collecting paired dataset. PURPOSE: To generate clinical UHRCT in lung from low resolution computed tomography (LRCT) using a GAN model. METHODS: 43 clinical scans with LRCT and UHRCT were collected in this study. Paired patches were selected using the structural similarity index measure (SSIM) and the peak signal-to-noise ratio (PSNR) threshold. A relativistic GAN with gradient guidance was trained to learn the mapping from LRCT to UHRCT. The performance of the proposed method was evaluated using PSNR and SSIM. A reader study with five-point Likert score (five for the worst and one for the best) is also applied to assess the proposed method in terms of general quality, diagnostic confidence, sharpness and denoise level. RESULTS: Experimental results show that our method got PSNR 32.60 ± 2.92 and SSIM 0.881 ± 0.057 on our clinical CT dataset, outperforming other state-of-the-art methods based on the simulated scenarios. Moreover, reader study shows that our method reached the good clinical performance in terms of general quality (1.14 ± 0.36), diagnostic confidence (1.36 ± 0.49), sharpness (1.07 ± 0.27) and high denoise level (1.29 ± 0.61) compared to other SR methods. CONCLUSION: This study demonstrated the feasibility of generating UHRCT images from LRCT without longer scanning time or increased radiation dose.

Ultra high speed SPECT bone imaging enabled by a deep learning enhancement method: a proof of concept.

BACKGROUND: To generate high-quality bone scan SPECT images from only 1/7 scan time SPECT images using deep learning-based enhancement method. MATERIALS AND METHODS: Normal-dose (925-1110 MBq) clinical technetium 99 m-methyl diphosphonate (99mTc-MDP) SPECT/CT images and corresponding SPECT/CT images with 1/7 scan time from 20 adult patients with bone disease and a phantom were collected to develop a lesion-attention weighted U2-Net (Qin et al. in Pattern Recognit 106:107404, 2020), which produces high-quality SPECT images from fast SPECT/CT images. The quality of synthesized SPECT images from different deep learning models was compared using PSNR and SSIM. Clinic evaluation on 5-point Likert scale (5 = excellent) was performed by two experienced nuclear physicians. Average score and Wilcoxon test were constructed to assess the image quality of 1/7 SPECT, DL-enhanced SPECT and the standard SPECT. SUVmax, SUVmean, SSIM and PSNR from each detectable sphere filled with imaging agent were measured and compared for different images. RESULTS: U2-Net-based model reached the best PSNR (40.8) and SSIM (0.788) performance compared with other advanced deep learning methods. The clinic evaluation showed the quality of the synthesized SPECT images is much higher than that of fast SPECT images (P < 0.05). Compared to the standard SPECT images, enhanced images exhibited the same general image quality (P > 0.999), similar detail of 99mTc-MDP (P = 0.125) and the same diagnostic confidence (P = 0.1875). 4, 5 and 6 spheres could be distinguished on 1/7 SPECT, DL-enhanced SPECT and the standard SPECT, respectively. The DL-enhanced phantom image outperformed 1/7 SPECT in SUVmax, SUVmean, SSIM and PSNR in quantitative assessment. CONCLUSIONS: Our proposed method can yield significant image quality improvement in the noise level, details of anatomical structure and SUV accuracy, which enabled applications of ultra fast SPECT bone imaging in real clinic settings.

Developmental trajectory of magnetic susceptibility in the healthy rhesus macaque brain.

Quantitative susceptibility mapping (QSM) is used to quantify iron deposition in non-human primates in our study. Although QSM has many applications in detecting iron deposits in the human brain, including the distribution of iron deposits in specific brain regions, the change of iron deposition with aging, and the comparison of iron deposits between diseased groups and healthy controls, few studies have applied QSM to non-human primates, while most animal brain experiments focus on biochemical and anatomical results instead of non-invasive experiments. Additionally, brain imaging in children's research is difficult, but can be substituted using young rhesus monkeys, which are very similar to humans, as research animals. Therefore, understanding the relationship between iron deposition and age in rhesus macaques' brains can offer insights into both the developmental trajectory of magnetic susceptibility in the animal model and the correlated evidence in children's research. Twenty-three healthy rhesus macaque monkeys (23 ± 7.85 years, range 2-29 years) were included in this research. Seven regions of interest (ROIs-globus pallidus, substantia nigra, dentate nucleus, caudate nucleus, putamen, thalamus, red nucleus) have been analyzed in terms of QSM and R2 * (apparent relaxation rate). Susceptibility in most ROIs correlated significantly with the growth of age, similarly to the results for R2 *, but showed different trends in the thalamus and red nucleus, which may be caused by the different sensitivities of myelination and iron deposition in R2 * and QSM analysis. By assessing the correlation between iron content and age in healthy rhesus macaques' brains using QSM, we provide a piece of pilot information on normality for advanced animal disease models. Meanwhile, this study also could serve as the normative basis for further clinical studies using QSM for iron content quantification. Due to the comparison of the susceptibility on the same experimental objects, this research can also provide practical support for future research on characteristics for QSM and R2 *.

Decompose quantitative susceptibility mapping (QSM) to sub-voxel diamagnetic and paramagnetic components based on gradient-echo MRI data.

PURPOSE: A method named DECOMPOSE-QSM is developed to decompose bulk susceptibility measured with QSM into sub-voxel paramagnetic and diamagnetic components based on a three-pool complex signal model. METHODS: Multi-echo gradient echo signal is modeled as a summation of three weighted exponentials corresponding to three types of susceptibility sources: reference susceptibility, diamagnetic and paramagnetic susceptibility relative to the reference. Paramagnetic component susceptibility (PCS) and diamagnetic component susceptibility (DCS) maps are constructed to represent the sub-voxel compartments by solving for linear and nonlinear parameters in the model. RESULTS: Numerical forward simulation and phantom validation confirmed the ability of DECOMPOSE-QSM to separate the mixture of paramagnetic and diamagnetic components. The PCS obtained from temperature-variant brainstem imaging follows the Curie's Law, which further validated the model and the solver. Initial in vivo investigation of human brain images showed the ability to extract sub-voxel PCS and DCS sources that produce visually enhanced contrast between brain structures comparing to threshold QSM.

Deep learning-based methods may minimize GBCA dosage in brain MRI.

OBJECTIVES: To evaluate the clinical performance of a deep learning (DL)-based method for brain MRI exams with reduced gadolinium-based contrast agent (GBCA) dose to provide better understanding of the readiness and limitations of this method. METHODS: Eighty-three consecutive patients (from March 2019 to August 2019) who underwent brain contrast-enhanced (CE) MRI were included. Three 3D T1-weighted images with zero-dose, low-dose (10%), and full-dose (100%) GBCA were collected. The first 30 cases were used to train a DL model to synthesize the full-dose GBCA images from the zero-dose and low-dose image pairs. The remaining 53 cases were used for testing. The enhancement pattern, number, and location of enhancing lesions were recorded. Overall image quality, image signal noise ratio (SNR), lesion conspicuity, and lesion enhancement were assessed. RESULTS: Lesion detection from the DL-synthesized CE-MRI image accurately matched those from the true full-dose CE-MRI images in 48 of 53 cases (90.6%). The DL method identified the lesions in 34 of 36 cases (94.4%) with a single enhanced lesion and all lesions in 3 of 6 cases (50.0%) in cases with multiple enhancing lesions. The agreement between synthesized and true full-dose CE-MRI images were 0.73, 0.63, 0.89, and 0.87 for image quality, image SNR, lesion conspicuity, and lesion enhancement, respectively. CONCLUSIONS: The proposed DL method is a feasible way to minimize the dosage of GBCAs in brain MRI without sacrificing the diagnostic information. Missing enhancement of small lesions in patients with multiple lesions was observed, requiring improvements in algorithms or dosage design. KEY POINTS: • This study evaluated the clinical performance of a DL-based reconstruction method for significant dose reduction in GBCA contrast-enhanced MRI exams. • The proposed DL method has the potential to satisfy the routine radiological diagnosis needs in certain clinical applications.

Imaging microstructure with diffusion and susceptibility MR: neuronal density correlation in Disrupted-in-Schizophrenia-1 mutant mice.

PURPOSE: To probe cerebral microstructural abnormalities and assess changes of neuronal density in Disrupted-in-Schizophrenia-1 (DISC1) mice using non-Gaussian diffusion and quantitative susceptibility mapping (QSM). MATERIALS AND METHODS: Brain specimens of transgenic DISC1 mice (n = 8) and control mice (n = 7) were scanned. Metrics of neurite orientation dispersion and density imaging (NODDI) and diffusion kurtosis imaging (DKI), as well as QSM, were acquired. Cell counting was performed on Nissl-stained sections. Group differences of imaging metrics and cell density were assessed. Pearson correlations between imaging metrics and cell densities were also examined. RESULTS: Significant increases of neuronal density were observed in the hippocampus of DISC1 mice. DKI metrics such as mean kurtosis exhibited significant group differences in the caudate putamen (P = 0.015), cerebral cortex (P = 0.021), and hippocampus (P = 0.011). However, DKI metrics did not correlate with cell density. In contrast, significant positive correlation between density of neurons and the neurite density index of NODDI in the hippocampus was observed (r = 0.783, P = 0.007). Significant correlation between density of neurons and susceptibility (r = 0.657, P = 0.039), as well as between density of neuroglia and susceptibility (r = 0.750, P = 0.013), was also observed in the hippocampus. CONCLUSION: The imaging metrics derived from DKI were not sensitive specifically to cell density, while NODDI could provide diffusion metrics sensitive to density of neurons. The magnetic susceptibility values derived from the QSM method can serve as a sensitive biomarker for quantifying neuronal density.

Evaluating Heterogeneity of Primary Lung Tumor Using Clinical Routine Magnetic Resonance Imaging and a Tumor Heterogeneity Index.

OBJECTIVE: To improve the assessment of primary tumor heterogeneity in magnetic resonance imaging (MRI) of non-small cell lung cancer (NSCLC), we proposed a method using basic measurements from T1- and T2-weighted MRI. METHODS: One hundred and four NSCLC patients with different T stages were studied. Fifty-two patients were analyzed as training group and another 52 as testing group. The ratios of standard deviation (SD)/mean signal value of primary tumor from T1-weighted (T1WI), T1-enhanced (T1C), T2-weighted (T2WI), and T2 fat suppression (T2fs) images were calculated. In the training group, correlation analyses were performed between the ratios and T stages. Then an ordinal regression model was built to generate the tumor heterogeneous index (THI) for evaluating the heterogeneity of tumor. The model was validated in the testing group. RESULTS: There were 11, 32, 40, and 21 patients with T1, T2, T3, and T4 disease, respectively. In the training group, the median SD/mean on T1WI, T1C, T2WI, and T2fs sequences was 0.11, 0.19, 0.16, and 0.15 respectively. The SD/mean on T1C (p=0.003), T2WI (p=0.000), and T2fs sequences (p=0.002) correlated significantly with T stages. Patients with more advanced T stage showed higher SD/mean on T2-weighted, T2fs, and T1C sequences. The median THI in the training group was 2.15. THI correlated with T stage significantly (p=0.000). In the testing group, THI was also significantly related to T stages (p=0.001). Higher THI had relevance to more advanced T stage. CONCLUSIONS: The proposed ratio measurements and THI based on MRI can serve as functional radiomic markers that correlated with T stages for evaluating heterogeneity of lung tumors.

Imaging beta amyloid aggregation and iron accumulation in Alzheimer's disease using quantitative susceptibility mapping MRI.

Beta amyloid is a protein fragment snipped from the amyloid precursor protein (APP). Aggregation of these peptides into amyloid plaques is one of the hallmarks of Alzheimer's disease. MR imaging of beta amyloid plaques has been attempted using various techniques, notably with T2* contrast. The non-invasive detectability of beta amyloid plaques in MR images has so far been largely attributed to focal iron deposition accompanying the plaques. It is believed that the T2* shortening effects of paramagnetic iron are the primary source of contrast between plaques and surrounding tissue. Amyloid plaque itself has been reported to induce no magnetic susceptibility effect. We hypothesized that aggregations of beta amyloid would increase electron density and induce notable changes in local susceptibility value, large enough to generate contrast relative to surrounding normal tissues that can be visualized by quantitative susceptibility mapping (QSM) MR imaging. To test this hypothesis, we first demonstrated in a phantom that beta amyloid is diamagnetic and can generate strong contrast on susceptibility maps. We then conducted experiments on a transgenic mouse model of Alzheimer's disease that is known to mimic the formation of human beta amyloid but without neurofibrillary tangles or neuronal death. Over a period of 18 months, we showed that QSM can be used to longitudinally monitor beta amyloid accumulation and accompanied iron deposition in vivo. Individual beta amyloid plaque can also be visualized ex vivo in high resolution susceptibility maps. Moreover, the measured negative susceptibility map and positive susceptibility map could provide histology-like image contrast for identifying deposition of beta amyloid plaques and iron. Finally, we demonstrated that the diamagnetic susceptibility of beta amyloid can also be observed in brain specimens of AD patients. The ability to assess beta amyloid aggregation non-invasively with QSM MR imaging may aid the diagnosis of Alzheimer's disease.

Microstructural alterations of cortical and deep gray matter over a season of high school football revealed by diffusion kurtosis imaging.

OBJECTIVES: To probe microstructural changes that are associated with subconcussive head impact exposure in deep and cortical gray matter of high school football players over a single season. METHODS: Players underwent diffusion kurtosis imaging (DKI) and quantitative susceptibility mapping (QSM) scans. Head impact data was recorded. Association between parametric changes and frequency of frontal head impact was assessed. RESULTS: In deep gray matter, significant decreases in mean kurtosis (MK) and increases in mean diffusivity (MD) over the season were observed in the thalamus and putamen. Correlations between changes in DKI metrics and frequency of frontal impacts were observed in the putamen and caudate. In cortical gray matter, decreases in MK were observed in regions including the pars triangularis and inferior parietal. In addition, increases in MD were observed in the rostral middle frontal cortices. Negative correlations between MK and frequency of frontal impacts were observed in the posterior part of the brain including the pericalcarine, lingual and middle temporal cortices. Magnetic susceptibility values exhibited no significant difference or correlation, suggesting these diffusion changes common within the group may not be associated with iron-related mechanisms. CONCLUSION: Microstructural alterations over the season and correlations with head impacts were captured by DKI metrics, which suggested that DKI imaging of gray matter may yield valuable biomarkers for evaluating brain injuries associated with subconcussive head impact. Findings of associations between frontal impacts and changes in posterior cortical gray matter also indicated that contrecoup injury rather than coup injury might be the dominant mechanism underlying the observed microstructural alterations. ADVANCES IN KNOWLEDGE: Significant microstructural changes, as reflected by DKI metrics, in cortical gray matter such as the rostral middle frontal cortices, and in deep gray matter such as the thalamus were observed in high school football players over the course of a single season without clinically diagnosed concussion. QSM showed no evidence of iron-related changes in the observed subconcussive brain injuries. The detected microstructural changes in cortical and deep gray matter correlated with frequency of subconcussive head impacts. IMPLICATIONS FOR PATIENT CARE: DKI may yield valuable biomarkers for evaluating the severity of brain injuries associated with subconcussive head impacts in contact sport athletes.

Differential microstructural and morphological abnormalities in mild cognitive impairment and Alzheimer's disease: Evidence from cortical and deep gray matter.

One aim of this study is to use non-Gaussian diffusion kurtosis imaging (DKI) for capturing microstructural abnormalities in gray matter of Alzheimer's disease (AD). The other aim is to compare DKI metrics against thickness of cortical gray matter and volume of deep gray matter, respectively. A cohort of 18 patients with AD, 18 patients with amnestic mild cognitive impairment (MCI), and 18 normal controls underwent morphological and DKI MR imaging. Images were investigated using regions-of-interest-based analyses for deep gray matter and vertex-wise analyses for cortical gray matter. In deep gray matter, more regions showed DKI parametric abnormalities than atrophies at the early MCI stage. Mean kurtosis (MK) exhibited the largest number of significant abnormalities among all DKI metrics. At the later AD stage, diffusional abnormalities were observed in fewer regions than atrophies. In cortical gray matter, abnormalities in thickness were mainly in the medial and lateral temporal lobes, which fit the locations of known early pathological changes. Microstructural abnormalities were predominantly in the parietal and even frontal lobes, which fit the locations of known late pathological changes. In conclusion, MK can complement conventional diffusion metrics for detecting microstructural changes, especially in deep gray matter. This study also provides evidence supporting the notion that microstructural changes predate morphological changes. Hum Brain Mapp 38:2495-2508, 2017. © 2017 Wiley Periodicals, Inc.

Hemisphere, gender and age-related effects on iron deposition in deep gray matter revealed by quantitative susceptibility mapping.

The purpose of this work was to investigate the effects of hemispheric location, gender and age on susceptibility value, as well as the association between susceptibility value and diffusional metrics, in deep gray matter. Iron content was estimated in vivo using quantitative susceptibility mapping. Microstructure was probed using diffusional kurtosis imaging. Regional susceptibility and diffusional metrics were measured for the putamen, caudate nucleus, globus pallidus, thalamus, substantia nigra and red nucleus in 42 healthy adults (age range 25-78 years). Susceptibility value was significantly higher in the left than the right side of the caudate nucleus (P = 0.043) and substantia nigra (P < 0.001). Women exhibited lower susceptibility values than men in the thalamus (P < 0.001) and red nucleus (P = 0.032). Significant age-related increases of susceptibility were observed in the putamen (P < 0.001), red nucleus (P < 0.001), substantia nigra (P = 0.004), caudate nucleus (P < 0.001) and globus pallidus (P = 0.017). The putamen exhibited the highest rate of iron accumulation with aging (slope of linear regression = 0.73 × 10(-3) ppm/year), which was nearly twice those in substantia nigra (slope = 0.40 × 10(-3) ppm/year) and caudate nucleus (slope = 0.39 × 10(-3) ppm/year). Significant positive correlations between the susceptibility value and diffusion measurements were observed for fractional anisotropy (P = 0.045) and mean kurtosis (P = 0.048) in the putamen without controlling for age. Neither correlation was significant after controlling for age. Hemisphere, gender and age-related differences in iron measurements were observed in deep gray matter. Notably, the putamen exhibited the highest rate of increase in susceptibility with aging. Correlations between susceptibility value and microstructural measurements were inconclusive. These findings could provide new clues for unveiling mechanisms underlying iron-related neurodegenerative diseases.

Quantitative Susceptibility Mapping: Contrast Mechanisms and Clinical Applications.

Quantitative susceptibility mapping (QSM) is a recently developed MRI technique for quantifying the spatial distribution of magnetic susceptibility within biological tissues. It first uses the frequency shift in the MRI signal to map the magnetic field profile within the tissue. The resulting field map is then used to determine the spatial distribution of the underlying magnetic susceptibility by solving an inverse problem. The solution is achieved by deconvolving the field map with a dipole field, under the assumption that the magnetic field is a result of the superposition of the dipole fields generated by all voxels and that each voxel has its unique magnetic susceptibility. QSM provides improved contrast to noise ratio for certain tissues and structures compared to its magnitude counterpart. More importantly, magnetic susceptibility is a direct reflection of the molecular composition and cellular architecture of the tissue. Consequently, by quantifying magnetic susceptibility, QSM is becoming a quantitative imaging approach for characterizing normal and pathological tissue properties. This article reviews the mechanism generating susceptibility contrast within tissues and some associated applications.

Aging in deep gray matter and white matter revealed by diffusional kurtosis imaging.

Diffusion tensor imaging has already been extensively used to probe microstructural alterations in white matter tracts, and scarcely, in deep gray matter. However, results in literature regarding age-related degenerative mechanisms in white matter tracts and parametric changes in the putamen are inconsistent. Diffusional kurtosis imaging is a mathematical extension of diffusion tensor imaging, which could more comprehensively mirror microstructure, particularly in isotropic tissues such as gray matter. In this study, we used the diffusional kurtosis imaging method and a white-matter model that provided metrics of explicit neurobiological interpretations in healthy participants (58 in total, aged from 25 to 84 years). Tract-based whole-brain analyses and regions-of-interest (anterior and posterior limbs of the internal capsule, cerebral peduncle, fornix, genu and splenium of corpus callosum, globus pallidus, substantia nigra, red nucleus, putamen, caudate nucleus, and thalamus) analyses were performed to examine parametric differences across regions and correlations with age. In white matter tracts, evidence was found supportive for anterior-posterior gradient and not completely supportive for retrogenesis theory. Age-related degenerations appeared to be broadly driven by axonal loss. Demyelination may also be a major driving mechanism, although confined to the anterior brain. In terms of deep gray matter, higher mean kurtosis and fractional anisotropy in the globus pallidus, substantia nigra, and red nucleus reflected higher microstructural complexity and directionality compared with the putamen, caudate nucleus, and thalamus. In particular, the unique age-related positive correlations for fractional anisotropy, mean kurtosis, and radial kurtosis in the putamen opposite to those in other regions call for further investigation of exact underlying mechanisms. In summary, the results suggested that diffusional kurtosis can provide measurements in a new dimension that were complementary to diffusivity metrics. Kurtosis together with diffusivity can more comprehensively characterize microstructural compositions and age-related changes than diffusivity alone. Combined with proper model, it may also assist in providing neurobiological interpretations of the identified alterations.

Tumor volumes measured from static and dynamic 18F-fluoro-2-deoxy-D-glucose positron emission tomography-computed tomography scan: comparison of different methods using magnetic resonance imaging as the criterion standard.

OBJECTIVE: The objective of this study was to compare the accuracy of calculating the primary tumor volumes using a gradient-based method and fixed threshold methods on the standardized uptake value (SUV) maps and the net influx of FDG (Ki) maps from positron emission tomography-computed tomography (PET-CT) images. MATERIALS AND METHODS: Newly diagnosed patients with head and neck cancer were recruited, and dynamic PET-CT scan and T2-weighted magnetic resonance imaging were performed. The maps of Ki and SUV were calculated from PET-CT images. The tumor volumes were calculated using a gradient-based method and a fixed threshold method at 40% of maximal SUV or maximal Ki. Four kinds of volumes, VOLKi-Gra (from the Ki maps using the gradient-based method), VOLKi-40% (from the Ki maps using the threshold of 40% maximal Ki), VOLSUV-Gra (from the SUV maps using the gradient-based method), and VOLSUV-40% (from the SUV maps using the threshold of 40% maximal SUV), were acquired and compared with VOLMRI (the volumes acquired on T2-weighted images) using the Pearson correlation, paired t test, and similarity analysis. RESULTS: Eighteen patients were studied, of which 4 had poorly defined tumors (PDT). The positron emission tomography-derived volumes were as follows: VOLSUV-40%, 2.1 to 41.2 cm (mean [SD], 12.3 [10.6]); VOLSUV-Gra, 2.2 to 28.1 cm (mean [SD], 13.2 [8.4]); VOLKi-Gra, 2.4 to 17.0 cm (mean [SD], 9.5 [4.6]); and VOLKi-40%, 2.7 to 20.3 cm (mean [SD], 12.0 [6.0]). The VOLMRI ranged from 2.9 to 18.1 cm (mean [SD], 9.1 [3.9]). The VOLKi-Gra significantly correlated with VOLMRI with the highest correlation coefficient (PDT included, R = 0.673, P = 0.002; PDT excluded, R = 0.841, P < 0.001) and presented no difference from VOLMRI (P = 0.672 or 0.561, respectively, PDT included and excluded). The difference between VOLKi-Gra and VOLMRI was also the smallest. CONCLUSIONS: The tumor volumes delineated on the Ki maps using the gradient-based method are more accurate than those on the SUV maps and using the fixed threshold methods.

Increasing the accuracy of volume and ADC delineation for heterogeneous tumor on diffusion-weighted MRI: correlation with PET/CT.

PURPOSE: To improve the accuracy of volume and apparent diffusion coefficient (ADC) measurements in diffusion-weighted magnetic resonance imaging (MRI), we proposed a method based on thresholding both the b0 images and the ADC maps. METHODS AND MATERIALS: In 21 heterogeneous lesions from patients with metastatic gastrointestinal stromal tumors (GIST), gross lesion were manually contoured, and corresponding volumes and ADCs were denoted as gross tumor volume (GTV) and gross ADC (ADC(g)), respectively. Using a k-means clustering algorithm, the probable high-cellularity tumor tissues were selected based on b0 images and ADC maps. ADC and volume of the tissues selected using the proposed method were denoted as thresholded ADC (ADC(thr)) and high-cellularity tumor volume (HCTV), respectively. The metabolic tumor volume (MTV) in positron emission tomography (PET)/computed tomography (CT) was measured using 40% maximum standard uptake value (SUV(max)) as the lower threshold, and corresponding mean SUV (SUV(mean)) was also measured. RESULTS: HCTV had excellent concordance with MTV according to Pearson's correlation (r=0.984, P<.001) and linear regression (slope = 1.085, intercept = -4.731). In contrast, GTV overestimated the volume and differed significantly from MTV (P=.005). ADC(thr) correlated significantly and strongly with SUV(mean) (r=-0.807, P<.001) and SUV(max) (r=-0.843, P<.001); both were stronger than those of ADC(g). CONCLUSIONS: The proposed lesion-adaptive semiautomatic method can help segment high-cellularity tissues that match hypermetabolic tissues in PET/CT and enables more accurate volume and ADC delineation on diffusion-weighted MR images of GIST.

Imaging biomarker with T1ρ and T2 mappings in osteoarthritis - in vivo human articular cartilage study.

INTRODUCTION: Osteoarthritis (OA) of the knee is a common and disabling disease worldwide. Its prevalence is increasing in view of the aging population. Changes in collagen content, its orientation and GAG content in the articular cartilage with age are the major features in knee osteoarthritis. These changes in collagen and GAG contents show no manifestation in plain radiography and conventional magnetic resonance imaging (MRI). Nevertheless, early diagnosis of the knee osteoarthritis is of paramount importance clinically in view of the evolution of putative interventions in its early stage. The aim of this project is to identify the relationships between the two imaging biomarkers (i.e. T1ρ and T2 mappings) and the GAG concentration in living human symptomatic cartilage. METHODOLOGY: 28 patients with clinical diagnosis of knee osteoarthritis were enrolled. 7 males and 16 females were recruited and their mean age was 68.1 (ranges from 53 to 84). Conventional PD sequence, T1ρ and T2 mappings were performed for each subject within 1 week before total knee arthroplasty. Articular cartilage from the lateral tibial plateau was harvested and the GAG content in cartilage was determined by using dimethylmethylene blue method. T1ρ mean and T2 values were calculated and correlate with GAG concentration statistically. RESULTS: The mean value T1ρ was 40.3±13.5ms, ranging from 15.3 to 69.3ms and the mean value T2 was 31.0±10.5ms, ranging from 16.1 to 46.9ms. The mean value of GAG content was 80.1±33.3mg, ranging from 24.9 to 166.0mg while the mean value of GAG concentration was 267.4±165.9mg/cm(3), ranging from 91.3 to 760.5mg/cm(3). T2 values were inversely correlated with GAG concentration with R(2)=0.375, p=0.001 while T1ρ values were also inversely correlated with GAG concentration with R(2)=0.200, p=0.025. CONCLUSION: This in vivo study confirmed that T1ρ and T2 values correlate with the GAG concentration in living human knee cartilages which corroborate with the previous works. The later (T2 values) is found more reliable in our study and less controversial in literatures. We postulate that T2 values can serve as a non-invasive imaging biomarker in the progress of knee osteoarthritis in terms of both disease diagnosis and treatment response monitoring.