MRI-based automatic segmentation of rectal cancer using 2D U-Net on two independent cohorts

BACKGROUND

Tumor delineation is time- and labor-intensive and prone to inter- and intraobserver variations. Magnetic resonance imaging (MRI) provides good soft tissue contrast, and functional MRI captures tissue properties that may be valuable for tumor delineation. We explored MRI-based automatic segmentation of rectal cancer using a deep learning (DL) approach. We first investigated potential improvements when including both anatomical T2-weighted (T2w) MRI and diffusion-weighted MR images (DWI). Secondly, we investigated generalizability by including a second, independent cohort.

MATERIAL AND METHODS

Two cohorts of rectal cancer patients (C1 and C2) from different hospitals with 109 and 83 patients, respectively, were subject to 1.5 T MRI at baseline. T2w images were acquired for both cohorts and DWI (b-value of 500 s/mm²) for patients in C1. Tumors were manually delineated by three radiologists (two in C1, one in C2). A 2D U-Net was trained on T2w and T2w + DWI. Optimal parameters for image pre-processing and training were identified on C1 using five-fold cross-validation and patient Dice similarity coefficient (DSC_p) as performance measure. The optimized models were evaluated on a C1 hold-out test set and the generalizability was investigated using C2.

RESULTS

For cohort C1, the T2w model resulted in a median DSC_p of 0.77 on the test set. Inclusion of DWI did not further improve the performance (DSC_p 0.76). The T2w-based model trained on C1 and applied to C2 achieved a DSC_p of 0.59.

CONCLUSION

T2w MR-based DL models demonstrated high performance for automatic tumor segmentation, at the same level as published data on interobserver variation. DWI did not improve results further. Using DL models on unseen cohorts requires caution, and one cannot expect the same performance.

Diffusion-Weighted MRI Findings of Caudate Nucleus and Putamen in Patients With Obsessive-Compulsive Disorder

Objective The purpose of this study was to establish the diffusion-weighted magnetic resonance imaging (DW-MRI) findings of the caudate nucleus and putamen in patients with obsessive-compulsive disorder (OCD) and to obtain new information on the etiopathogenesis of OCD, which is still unclear. Methods The study comprised 20 patients with OCDs and 20 healthy volunteers. In these cases, DW-MRI and diffusion-weighted echo-planar images (DW-EPI) at b600 and b1000 gradient values were taken and the measurements were made using the apparent diffusion coefficient (ADC) maps of each group at b600 and b1000 values from the caudate nucleus and putamen. Results When the DW-MRI examination in patients with OCD was compared with the control group, the mean ADC values in the caudate nucleus and putamen were not found to have statistically significantly changed. In addition, there were no significant differences regarding the right and left caudate nuclei and putamen ADC values at the b600 and b1000 in the patients with OCD or the control group. Conclusion There are still many unknowns about the neurobiology of OCD. When the DW-MRI examination of the patients with OCD was compared with the control group in our study, no significant difference was found between the ADC values of the caudate nucleus and putamen. Further studies are required for this present study on DW-MRI in patients with OCD to be meaningful.

The prognosis of brain magnetic resonance imaging injury pattern for outcomes of hypothermia-treated infants

Magnetic resonance imaging (MRI) can be a tool that allows the observation of structural injury patterns after cooling. The aim of this study was to determine the early pattern of brain injury in the MRIs of infants with hypoxic ischemic encephalopathy (HIE) after cooling and to search for any clinical factors related to abnormal MRI findings.The study retrospectively recruited 118 infants who were treated with therapeutic hypothermia (TH) between 2013 and 2016.Forty-three patients had normal brain MRI, and 75 had abnormal brain MRI findings. The TH-treated infants with abnormal brain MRI readings showed significantly more clinical seizures and the use of additional antiepileptic drugs (AEDs) than the normal MRI group. As a long-term outcome, more lesions in the basal ganglia and thalamus, posterior limb of internal capsule, or severe white matter lesions were associated with abnormal neurodevelopmental outcomes at 18 to 24 months of age.A higher frequency of clinical seizures and AED use were related to abnormal brain injury on MRI. A significant risk for poor long-term outcomes was found in the abnormal brain MRI group.

Quantitative analyses of the correlation between dynamic contrast-enhanced MRI and intravoxel incoherent motion DWI in thyroid nodules

This study investigated the correlation between dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and intravoxel incoherent motion diffusion-weighted imaging (IVIM DWI) to differentiate thyroid nodules. Quantitative DCE-MRI parameters, including the transfer constant (K ^trans), rate constant (K _ep) and volume fraction of the extracellular extravascular space (V _e), were calculated. The diffusion coefficient (D), pseudo-diffusion coefficient (D^* ), and perfusion fraction (f) were derived from biexponential fitting of IVIM DWI. A total of 38 nodules, including 22 malignant and 16 benign nodules, were analyzed. The K ^trans, K _ep and V _e for benign lesions were 1.32 ± 0.76 min^-1, 6.44 ± 1.44 min^-1, and 2.02 ± 0.89 min^-1, respectively, and for malignant lesions, the values were 0.84 ± 0.30 min^-1, 5.43 ± 1.38 min^-1, and 1.71 ± 0.83 min^-1, respectively (P = 0.027, 0.036, and 0.257, respectively). The D, f, and D^* for benign lesions were 1.51 ± 0.52 mm²/s, 26.63 ± 8.75%, and 15.84 ± 8.71 mm²/s, respectively, and for malignant lesions, the values were 0.68 ± 0.17 mm²/s, 31.63 ± 10.72%, and 11.10 ± 4.21 mm²/s, respectively (P [< 0.0001, 0.135, 0.058], respectively). No significant correlations were found between IVIM DWI and DCE-MRI quantitative parameters (all P > 0.05). In benign nodules, a moderate inverse correlation was found between D and K _ep (r = -0.54, P = 0.031). IVIM DWI shows no significant correlation with perfusion parameters derived from DCE-MRI; however, IVIM DWI combined with quantitative DCE-MRI may be a useful imaging tool for the assessment of thyroid nodules in clinical studies.

Evaluation of Malignant Breast Lesions Using High-resolution Readout-segmented Diffusion-weighted Echo-planar Imaging: Comparison with Pathology

PURPOSE

We aimed to investigate the performance of high resolution-diffusion-weighted imaging (HR-DWI) using readout-segmented echo-planar imaging in visualizing malignant breast lesions and evaluating their extent, using pathology as a reference.

METHODS

This retrospective study included patients who underwent HR-DWI with surgically confirmed malignant breast lesions. Two radiologists blinded to the final diagnosis evaluated HR-DWI independently and identified the lesions, measuring their maximum diameters. Another radiologist confirmed if those lesions were identical to the pathology. The maximum diameters of the lesions between HR-DWI and pathology were compared, and their correlations were calculated using Spearman's correlation coefficient. Apparent diffusion coefficient (ADC) values of the lesions were measured.

RESULTS

Ninety-five mass/64 non-mass lesions were pathologically confirmed in 104 females. Both radiologists detected the same 93 mass lesions (97.9%). Spearman's correlation coefficient for mass lesions were 0.89 and 0.90 (P < 0.0001 and 0001) for the two radiologists, respectively. The size differences within 10 mm were 90.3% (84/93) and 94.6% (88/93) respectively. One radiologist detected 35 non-mass lesions (54.7%) and another radiologist detected 32 non-mass lesions (50.0%), of which 28 lesions were confirmed as identical. Spearman's correlation coefficient for non-mass lesions were 0.59 and 0.22 (P = 0.0002 and 0.22), respectively. The mean ADC value of mass lesions and non-mass lesions were 0.80 and 0.89 × 10^-3 mm²/s, respectively.

CONCLUSION

Using HR-DWI, malignant mass lesions were depicted with excellent agreement with the pathological evaluation. Approximately half of the non-mass lesions could not be identified, suggesting a current limitation of HR-DWI.

Associations between home literacy environment, brain white matter integrity and cognitive abilities in preschool-age children

AIM

Caregiver-child reading is advocated by health organisations, citing cognitive and neurobiological benefits. The influence of home literacy environment (HLE) on brain structure prior to kindergarten has not previously been studied.

METHODS

Preschool-age children completed assessments of language (EVT-2, CTOPP-2 Rapid Object Naming) and emergent literacy skills (Get Ready to Read!, The Reading House) followed by diffusion tensor imaging (DTI). Parents completed a survey of HLE (StimQ-P₂ READ), which has four subscales. DTI measures included axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD) and fractional anisotropy (FA).

RESULTS

Forty-seven children completed DTI (54 ± 7 months, range 36-63; 27 girls). StimQ-P₂ READ scores correlated with higher EVT-2, GRTR and TRH scores, controlling for age and gender (P < .01), and also with lower AD, RD and MD in tracts supporting language and literacy skills, controlling for age, gender and income (P < .05, family-wise error corrected). Correlations were strongest for the Bookreading Quantity subscale, including with higher scores on all cognitive measures including CTOPP-2, and also with higher FA in left-lateralised literacy-supporting tracts, controlling for age, gender and income.

CONCLUSION

More nurturing home reading environment prior to kindergarten may stimulate brain development supporting language and literacy skills, reinforcing the need for further study.

A Clinical Approach to Diffusion-Weighted Magnetic Resonance Imaging in Evaluating Chest Wall Invasion of Breast Tumors

Objective:

The purpose of this study is to evaluate diffusion weighted magnetic rsonance imaging (MRI) acquisitions in delineating posterior extent of breast tumors and in predicting chest wall invasion prior to treatment. To our knowledge, there has not been any literature specifically evaluating the utility of diffusion-weighted acquisitions in chest wall invasion of breast tumors.

Materials and Methods:

A retrospective review of our breast imaging database for keywords “chest wall invasion” and “breast MRI” was performed over the last 14 years. Diffusion sequences, T1 sequences (pre and post contrast), and T2 sequences were evaluated. Apparent diffusion coefficient (ADC) values in tumor and chest wall were assessed. Imaging findings were correlated with surgical pathology.

Results:

23 patients met inclusion criteria. All 23 had loss of fat plane on T2 sequences. 22/23 had loss of fat plane on postcontrast T1 sequences. Pectoralis muscle enhancement was present in 19/23 (83%) tumors and chest wall enhancement was present 9/23 (39%) tumors. Qualitative restricted diffusion within the pectoralis muscle was present in 18/23 (71%) tumors and in the chest wall was present in 8/23 (35%) tumors. Mean ADC values were 1.15 s/mm² in the tumor and 1.29 s/mm² in the chest wall. Sensitivity, specificity, positive predictive value and negative predictive value were 100%, 36%, 63%, and 100% for chest wall enhancement respectively and 69%, 36%, 61%, and 80% for chest wall diffusion-weighted imaging restriction respectively.

Conclusion:

Diffusion weighted sequences can be helpful in characterizing chest wall invasion of breast tumors.

Longitudinal investigation of ischemic stroke using magnetic resonance imaging: Animal model

BACKGROUND

Arterial embolism is a major cause of ischemic stroke. Currently, digital subtraction angiography (DSA) is the gold standard in clinical arterial embolization examinations. However, it is invasive and risky.

OBJECTIVE

This study aims to longitudinally assess the progression of carotid artery embolism in middle cerebral artery occlusion animal model (MCAO) using magnetic resonance imaging (MRI) techniques.

METHODS

Turbo spin echo (TSE), time of flight magnetic resonance angiography (TOF-MRA) and diffusion weighted magnetic resonance imaging (DWI) were used to evaluate the image characteristics of cerebral tissues at 1, 2, 3, 7, 14, 21 and 28 days after MCAO microsurgery on Sprague-Dawley (SD) rats. Quantitative analysis was performed and compared in MCAO hemisphere and contralateral normal hemisphere. Furthermore, pathologic section using triphenyl tetrazolium chloride (TTC) stain was performed as well.

RESULTS

TOF-MRA showed carotid signal void in the embolism side, which is evidence of artery occlusion. The used MRI techniques showed that edema gradually dissipated within one week, but there was no significant change afterwards. The time-varying signal intensity of MRI techniques in MCAO hemisphere changed significantly, but there were no significant changes in contralateral normal hemisphere. Cerebral injury was also confirmed by analysis of pathology images.

CONCLUSIONS

The MCAO animal model was successfully established on SD rats using the microsurgery to assess arterial embolization of intracranial tissue injury.

Whole-volume apparent diffusion coefficient-based entropy parameters for assessment of gastric cancer aggressiveness

PURPOSE

To explore the role of whole-volume apparent diffusion coefficient (ADC)-based entropy parameters in the preoperative assessment of gastric cancer's aggressiveness.

MATERIALS AND METHODS

In all, 64 patients with gastric cancers who underwent 3.0T magnetic resonance imaging (MRI) were retrospectively included. Regions of interest were drawn manually using in-house software, around gastric cancer lesions on each slice of the diffusion-weighted images and ADC maps. Entropy-related parameters based on ADC maps were calculated automatically: (1) first-order entropy; (2-5) second-order entropies, including entropy(H)₀ , entropy(H)₄₅ , entropy(H)₉₀ , and entropy(H)₁₃₅ ; (6) entropy(H)_mean ; and (7) entropy(H)_range . Correlations between entropy-related parameters and pathological characteristics were analyzed with the Spearman correlation test. The parameters were compared among different pathological characteristics with independent-samples Kruskal-Wallis or Mann-Whitney U-test. Additionally, diagnostic performances of parameters in differentiating different pathological characteristics were analyzed by receiver operating characteristic (ROC) curve analysis.

RESULTS

All the entropy-related parameters significantly correlated with T, N, and overall stages, especially the first-order entropy (r = 0.588, 0.585, and 0.677, respectively, all P < 0.05). All the entropy-related parameters showed significant differences in gastric cancers at different T, N, and overall stages, as well as at different status of vascular invasion (P < 0.001-0.027). And four parameters, including entropy, entropy(H)₀ , entropy(H)₄₅ , and entropy(H)₉₀ , showed significant differences between gastric cancers with and without perineural invasion (P 0.006-0.040).

CONCLUSION

Entropy-related parameters derived from whole-volume ADC texture analysis could help assess the aggressiveness of gastric cancers via analyzing intratumoral heterogeneity quantitatively, especially the first-order entropy.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:168-175.

[The possibilities for diagnostics of prescription of death coming based on the changes in the lumbar intervertebral disks (the comparison of the morphological, immunohistochemical and topographical findings)]

The objective of the present study was the comprehensive analysis of the postmortem changes in the lumbar intervertebral disks within different periods after death. A total of seven vertebromotor segments were distinguished in the lumbosacral region of the vertebral column based on the examination of 7 corpses. All these segments were divided into three groups in accordance with the prescription of death coming as follows: up to 12 hours (group 1), between 12 and 24 hours (group 2), and between 24 and 36 hours (group 3) after death. The models of the segments thus obtained were subjected to the study by means of diffusion weighted MRI. The removed intervertebral disks were used for morphological and immunohistochemical investigations. The comparison of the diffusion coefficients (DI) revealed the significant difference between the intervertebral disks assigned to groups 1 and 2 (p<0.01). The number of the cells in the pulpal core, the vertebral end plate, and the fibrous ring in all the above groups of the intervertebral disks was significantly reduced (p<0.01). The analysis of the correlation dependence between cell density and diffusion coefficients has demonstrated the well apparent relationship between these characteristics of the intervertebral disks comprising groups 1 and 2. It is concluded that diffusion weighted MRI in the combination with the calculation of diffusion coefficients for the intervertebral disks provides a tool for diagnostics of prescription of death coming as confirmed by the results of the morphometric studies and immunohistochemical analysis.

Inter-Observer Agreement on Diffusion-Weighted Magnetic Resonance Imaging Interpretation for Diagnosis of Acute Ischemic Stroke Among Emergency Physicians

Objectives

Diffusion-weighted magnetic resonance imaging (DW-MRI) is a highly sensitive tool for the detection of early ischemic stroke and is excellent at detecting small and early infarcts. Nevertheless, conflict may arise and judgments may differ among different interpreters. Inter-observer variability shows the systematic difference among different observers and is expressed as the kappa (Κ) coefficient. In this study, we aimed to determinate the inter-observer variability among emergency physicians in the use of DW-MRI for the diagnosis of acute ischemic stroke.

Methods

Cranial DW-MRI images of 50 patients were interpreted in this retrospective observational cross-sectional study. Patients who were submitted to DW-MRI imaging for a suspected acute ischemic stroke were included in the study, unless the scans were ordered by any of the reviewers or they were absent in the system. The scans were blindly and randomly interpreted by four emergency physicians. Inter-observer agreement between reviewers was evaluated using Fleiss’ Κ statistics.

Results

The mean kappa value for high signal on diffusion-weighted images (DWI) and for reduction on apparent diffusion coefficient (ADC) were substantial (k=0.67) and moderate (k=0.60) respectively. The correlation for detection of the presence of ischemia and location was substantial (k: 0.67). There were 18 false-positive and 4 false-negative evaluations of DWI, 15 false positive and 8 false-negative evaluations of ADC.

Conclusions

Our data suggest that DW-MRI is reliable in screening for ischemic stroke when interpreted by emergency physicians in the emergency department. The levels of stroke identification and variability show that emergency physicians may have an acceptable level of agreement.

Diagnostic value of diffusion weighted MRI and ADC in differential diagnosis of cavernous hemangioma of the liver

AIMS

To investigate the use of diffusion weighted magnetic resonance imaging (DWI) and the apparent diffusion coefficient (ADC) values in the diagnosis of hemangioma.

MATERIALS AND METHODS

The study population consisted of 72 patients with liver masses larger than 1 cm (72 focal lesions). DWI examination with a b value of 600 s/mm2 was carried out for all patients. After DWI examination, an ADC map was created and ADC values were measured for 72 liver masses and normal liver tissue (control group). The average ADC values of normal liver tissue and focal liver lesions, the "cut-off" ADC values, and the diagnostic sensitivity and specificity of the ADC map in diagnosing hemangioma, benign and malignant lesions were researched.

RESULTS

Of the 72 liver masses, 51 were benign and 21 were malignant. Benign lesions comprised 38 hemangiomas and 13 simple cysts. Malignant lesions comprised 9 hepatocellular carcinomas, and 12 metastases. The highest ADC values were measured for cysts (3.782±0.53×10(-3) mm(2)/s) and hemangiomas (2.705±0.63×10(-3) mm(2)/s). The average ADC value of hemangiomas was significantly higher than malignant lesions and the normal control group (p<0.001). The average ADC value of cysts were significantly higher when compared to hemangiomas and normal control group (p<0.001). To distinguish hemangiomas from malignant liver lesions, the "cut-off" ADC value of 1.800×10(-3) mm(2)/s had a sensitivity of 97.4% and a specificity of 90.9%. To distinguish hemangioma from normal liver parenchyma the "cut-off" value of 1.858×10(-3) mm(2)/s had a sensitivity of 97.4% and a specificity of 95.7%. To distinguish benign liver lesions from malignant liver lesions the "cut-off" value of 1.800×10(-3) mm(2)/s had a sensitivity of 96.1% and a specificity of 90.0%.

CONCLUSION

DWI and quantitative measurement of ADC values can be used in differential diagnosis of benign and malignant liver lesions and also in the diagnosis and differentiation of hemangiomas. When dynamic examination cannot distinguish cases with vascular metastasis and lesions from hemangioma, DWI and ADC values can be useful in the primary diagnosis and differential diagnosis. The technique does not require contrast material, so it can safely be used in patients with renal failure.

Diffusion-Weighted imaging and diffusion tensor imaging detect delayed graft function and correlate with allograft fibrosis in patients early after kidney transplantation

PURPOSE

To combine diffusion-weighted imaging (DWI) and diffusion tensor imaging (DTI) for detection of allograft dysfunction in patients early after kidney transplantation and to correlate diffusion parameters with renal function and renal histology of allograft biopsies.

MATERIALS AND METHODS

Between day 4 and 11 after kidney transplantation 33 patients with initial graft function and 31 patients with delayed graft function (DGF) were examined with a 1.5T magnetic resonance imaging (MRI) scanner. DTI and DWI sequences were acquired and fractional anisotropy (FA), apparent diffusion coefficient (ADCmono), pure diffusion (ADCdiff ), and the perfusion fraction (Fp) were calculated. Kidney biopsies in 26 patients were analyzed for allograft pathology, ie, acute tubular injury, inflammation, edema, renal fibrosis, and rejection. Histological results were correlated with MRI parameters.

RESULTS

In the renal medulla FA (0.25 ± 0.06 vs. 0.29 ± 0.06, P < 0.01) and ADCmono (1.73 ± 0.13*10(-3) vs. 1.93 ± 0.16*10(-3) mm(2) /s, P < 0.001) were significantly reduced in DGF patients compared with patients with initial function. For ADCdiff and Fp similar reductions were observed. FA and ADCmono significantly correlated with renal function (r = 0.53 and r = 0.57, P < 0.001) and were inversely correlated with the amount of renal fibrosis (r = -0.63 and r = -0.65, P < 0.05).

CONCLUSION

Combined DTI and DWI detected allograft dysfunction early after kidney transplantation and correlated with allograft fibrosis. J. Magn. Reson. Imaging 2016;44:112-121.

Diffusion Weighted Magnetic Resonance Imaging for the Characterization of Solitary Pulmonary Lesions

BACKGROUND

We evaluated the differential diagnosis of solitary pulmonary lesions on magnetic resonance imaging.

AIMS

To investigate the value of diffusion weighted imaging on the differential diagnosis of solitary pulmonary lesions.

STUDY DESIGN

Randomized prospective study.

METHODS

This prospective study included 48 solitary pulmonary nodules and masses (18 benign, 30 malignant). Single shot echo planar spin echo diffusion weighted imaging (DWI) was performed with two b factors (0 and 1000 s/mm(2)). Apparent diffusion coefficients (ADCs) were calculated. On diffusion weighted (DW) trace images, the signal intensities (SI) of the lesions were visually compared to the SI of the thoracic spinal cord using a 5-point scale: 1: hypointense, 2: moderately hypointense, 3: isointense, 4: moderately hyperintense, 5: significantly hyperintense. For the quantitative evaluation, the lesion to thoracic spinal signal intensity ratios and the ADCs of the lesions were compared between groups.

RESULTS

On visual evaluation, taking the density of the spinal cord as a reference, most benign lesions were found to be hypointense, while most of the malignant lesions were evaluated as hyperintense on DWI with a b factor of 1000 s/mm(2). In contrast, on T2 weighted images, it was seen that the distinction of malignant lesions from benign lesions was not statistically significant. The ADCs of the malignant lesions were significantly lower than those of benign lesions (mean ADC was 2.02×10(-3) mm(2)/s for malignant lesions, and 1.195×10(-3)±0.3 mm(2)/s for benign lesions). Setting the cut-off value at 1.5×10(-3), ADC had a sensitivity of 86.7% and a specificity of 88.9% for the differentiation of benign lesions from malignant lesions.

CONCLUSION

DWI may aid in the differential diagnosis of solitary pulmonary lesions. (ClinicalTrials.gov Identifier: NCT02482181).

Osmotic myelinolysis: Does extrapontine myelinolysis precede central pontine myelinolysis? Report of two cases and review of literature

Osmotic myelinolysis is an acute, rare, demyelinating process. After the initial description of the condition by Adam and colleagues in 1959, many case series have been published describing the central and extrapontine myelinolysis. Imaging has a definitive role in establishing the diagnosis of osmotic myelinolysis in vivo and diffusion-weighted imaging reveals earliest changes in affected brain parenchyma. We report two cases of patients with proven malignancy who developed extrapontine myelinolysis after treatment for hyponatremia and progressed to central pontine myelinolysis within a week. This was confirmed with magnetic resonance (MR) imaging and clinical assessment. This temporal progression of MR features, especially on diffusion-weighted imaging, from extrapontine to central pontine myelinolysis in osmotic injury has not been described in literature to the best of our knowledge. An early MRI of the brain in suspected/high-risk cases of osmotic myelinolysis may show features of extrapontine myelinolysis in the form of restricted diffusion in bilateral basal ganglia and may serve as a guide for predicting progression, prognosticating and deciding further treatment of pontine myelinolysis. We propose that in a significant number of cases, central pontine myelinolysis may be predicted by doing an early MRI of the brain with diffusion-weighted imaging, when extrapontine symptoms start to develop. This can potentially increase the window period and possibilities for therapeutic intervention and may even help in prevention.

Noise estimation from averaged diffusion weighted images: Can unbiased quantitative decay parameters assist cancer evaluation?

PURPOSE

Multiexponential decay parameters are estimated from diffusion-weighted-imaging that generally have inherently low signal-to-noise ratio and non-normal noise distributions, especially at high b-values. Conventional nonlinear regression algorithms assume normally distributed noise, introducing bias into the calculated decay parameters and potentially affecting their ability to classify tumors. This study aims to accurately estimate noise of averaged diffusion-weighted-imaging, to correct the noise induced bias, and to assess the effect upon cancer classification.

METHODS

A new adaptation of the median-absolute-deviation technique in the wavelet-domain, using a closed form approximation of convolved probability-distribution-functions, is proposed to estimate noise. Nonlinear regression algorithms that account for the underlying noise (maximum probability) fit the biexponential/stretched exponential decay models to the diffusion-weighted signal. A logistic-regression model was built from the decay parameters to discriminate benign from metastatic neck lymph nodes in 40 patients.

RESULTS

The adapted median-absolute-deviation method accurately predicted the noise of simulated (R(2)  = 0.96) and neck diffusion-weighted-imaging (averaged once or four times). Maximum probability recovers the true apparent-diffusion-coefficient of the simulated data better than nonlinear regression (up to 40%), whereas no apparent differences were found for the other decay parameters.

CONCLUSIONS

Perfusion-related parameters were best at cancer classification. Noise-corrected decay parameters did not significantly improve classification for the clinical data set though simulations show benefit for lower signal-to-noise ratio acquisitions.

Cerebral fat embolism syndrome after long bone fracture due to gunshot injury

Cerebral fat embolism syndrome is a lethal complication of long-bone fractures and clinically manifasted with respiratory distress, altered mental status, and petechial rash. We presented a 20-year-old male admitted with gun-shot wounds to his left leg. Twenty-four hours after the event, he had generalized tonic clonic seizures, decorticate posture and a Glascow Coma Scale of seven with localization of painful stimuli. Subsequent magnetic resonance imaging of the brain showed a star-field pattern defining multiple lesions of restricted diffusion. On a 4-week follow-up, he had returned to normal neurological function. Despite the severity of the neurological condition upon initial presentation, the case cerebral fat embolism illustrates that, cerebral dysfunction associated with cerebral fat embolism illustrates reversible.

Strategies for assessing diffusion anisotropy on the basis of magnetic resonance images: comparison of systematic errors

Diffusion weighted imaging uses the signal loss associated with the random thermal motion of water molecules in the presence of magnetic field gradients to derive a number of parameters that reflect the translational mobility of the water molecules in tissues. With a suitable experimental set-up, it is possible to calculate all the elements of the local diffusion tensor (DT) and derived parameters describing the behavior of the water molecules in each voxel. One of the emerging applications of the information obtained is an interpretation of the diffusion anisotropy in terms of the architecture of the underlying tissue. These interpretations can only be made provided the experimental data which are sufficiently accurate. However, the DT results are susceptible to two systematic error sources: On one hand, the presence of signal noise can lead to artificial divergence of the diffusivities. In contrast, the use of a simplified model for the interaction of the protons with the diffusion weighting and imaging field gradients (b matrix calculation), common in the clinical setting, also leads to deviation in the derived diffusion characteristics. In this paper, we study the importance of these two sources of error on the basis of experimental data obtained on a clinical magnetic resonance imaging system for an isotropic phantom using a state of the art single-shot echo planar imaging sequence. Our results show that optimal diffusion imaging require combining a correct calculation of the b-matrix and a sufficiently large signal to noise ratio.

MRI of Focal Liver Lesions

Magnetic resonance imaging, MRI has more advantages than ultrasound, computed tomography, CT, positron emission tomography, PET, or any other imaging modality in diagnosing focal hepatic masses. With a combination of basic T1 and T2 weighted sequences, diffusion weighted imaging, DWI, and hepatobiliary gadolinium contrast agents, that is gadobenate dimeglumine (Gd-BOPTA) and gadoxetic acid (Gd-EOB), most liver lesions can be adequately diagnosed. Benign lesions, as cyst, hemangioma, focal nodular hyperplasia, FNH or adenoma, can be distinguished from malignant lesions. In a non-cirrhotic liver, the most common malignant lesions are metastases which may be hypovascular or hypervascular. In the cirrhotic liver hepatocellular carcinoma, HCC, is of considerable importance. Besides, intrahepatic cholangiocarcinoma and other less common malignancies has to be assessed. In this review, the techniques and typical MRI features are presented as well as the new algorithm issued by American Association for the Study of the Liver Diseases (AASLD).