T2-weighted Magnetic Resonance Imaging Images Research Articles

RU-Net: skull stripping in rat brain MR images after ischemic stroke with rat U-Net

BackgroundExperimental ischemic stroke models play a fundamental role in interpreting the mechanism of cerebral ischemia and appraising the development of pathological extent. An accurate and automatic skull stripping tool for rat brain image volumes with magnetic resonance imaging (MRI) are crucial in experimental stroke analysis. Due to the deficiency of reliable rat brain segmentation methods and motivated by the demand for preclinical studies, this paper develops a new skull stripping algorithm to extract the rat brain region in MR images after stroke, which is named Rat U-Net (RU-Net).MethodsBased on a U-shape like deep learning architecture, the proposed framework integrates batch normalization with the residual network to achieve efficient end-to-end segmentation. A pooling index transmission mechanism between the encoder and decoder is exploited to reinforce the spatial correlation. Two different modalities of diffusion-weighted imaging (DWI) and T2-weighted MRI (T2WI) corresponding to two in-house datasets with each consisting of 55 subjects were employed to evaluate the performance of the proposed RU-Net.ResultsExtensive experiments indicated great segmentation accuracy across diversified rat brain MR images. It was suggested that our rat skull stripping network outperformed several state-of-the-art methods and achieved the highest average Dice scores of 98.04% (p < 0.001) and 97.67% (p < 0.001) in the DWI and T2WI image datasets, respectively.ConclusionThe proposed RU-Net is believed to be potential for advancing preclinical stroke investigation and providing an efficient tool for pathological rat brain image extraction, where accurate segmentation of the rat brain region is fundamental.

A new biomarker combining multimodal MRI radiomics and clinical indicators for differentiating inverted papilloma from nasal polyp invaded the olfactory nerve possibly.

Inverted papilloma (IP) and nasal polyp (NP), as two benign lesions, are difficult to distinguish on MRI imaging and clinically, especially in predicting whether the olfactory nerve is damaged, which is an important aspect of treatment and prognosis. We plan to establish a new biomarker to distinguish IP and NP that may invade the olfactory nerve, and to analyze its diagnostic efficacy. A total of 74 cases of IP and 55 cases of NP were collected. A total of 80% of 129 patients were used as the training set (59 IP and 44 NP); the remaining were used as the testing set. As a multimodal study (two MRI sequences and clinical indicators), preoperative MR images including T2-weighted magnetic resonance imaging (T2-WI) and contrast-enhanced T1-weighted magnetic resonance imaging (CE-T1WI) were collected. Radiomic features were extracted from MR images. Then, the least absolute shrinkage and selection operator (LASSO) regression method was used to decrease the high degree of redundancy and irrelevance. Subsequently, the radiomics model is constructed by the rad scoring formula. The area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the model have been calculated. Finally, the decision curve analysis (DCA) is used to evaluate the clinical practicability of the model. There were significant differences in age, nasal bleeding, and hyposmia between the two lesions (p < 0.05). In total, 1,906 radiomic features were extracted from T2-WI and CE-T1WI images. After feature selection, using 12 key features to bulid model. AUC, sensitivity, specificity, and accuracy on the testing cohort of the optimal model were, respectively, 0.9121, 0.828, 0.9091, and 0.899. AUC on the testing cohort of the optimal model was 0.9121; in addition, sensitivity, specificity, and accuracy were, respectively, 0.828, 0.9091, and 0.899. A new biomarker combining multimodal MRI radiomics and clinical indicators can effectively distinguish between IP and NP that may invade the olfactory nerve, which can provide a valuable decision basis for individualized treatment.

Examination of brain area volumes based on voxel-based morphometry and multidomain cognitive impairment in asymptomatic unilateral carotid artery stenosis.

Recent evidence has demonstrated that unilateral carotid artery stenosis (CAS) can contribute to the development of cognitive impairment. However, the features of cognitive dysfunction induced by unilateral CAS remain unclear. Sixty asymptomatic patients with unilateral CAS were divided into mild, moderate and severe stenosis groups. These patients and 20 healthy controls provided clinical data and serum, which was used to assess the levels of certain vascular risk factors. Then, they participated in a battery of neuropsychological tests. Additionally, all participants underwent a 3.0 T magnetic resonance imaging (MRI) scan of the brain. Chi-square tests and one-way ANOVA were used to determine significant differences in the risk factors and cognitive test scores between groups. Multiple logistic regression analysis and the receiver operating characteristic (ROC) curve analysis were performed to identify the independent risk factors for cognitive impairment in patients with CAS. Finally, fluid attenuated inversion recovery (FLAIR) T1-weighted MRI images were processed by voxel-based morphometry (VBM) analysis using the Statistical Parametric Mapping (SPM) 8 software. Compared with healthy controls, the scores of the Mini-Mental State Examination, Digital Span Test backward, and Rapid Verbal Retrieve were significantly reduced in patients with left CAS. The scores in all cognitive scales were significantly lower in patients with right CAS than in controls. Logistic regression analysis demonstrated that the degree of carotid stenosis was an independent risk factor for cognitive impairment in asymptomatic patients with unilateral CAS. Furthermore, VBM analysis showed that, compared with those in healthy controls, gray matter and white matter volumes in specific brain areas were markedly decreased in patients with severe unilateral CAS. However, in patients with moderate right CAS, there was a significant decline in the volume of gray matter in the left parahippocampal gyrus and supplementary motor area. Additionally, the volume of white matter in the left insula was obviously lower in patients with moderate right CAS than in healthy controls. Unilateral asymptomatic CAS, especially on the right side, contributed to cognitive impairment, including memory, language, attention, executive function and visuospatial function. In addition, based on VBM analysis, both gray matter atrophy and white matter lesions were found in patients with unilateral asymptomatic CAS.

MRI-based vertebral bone quality score compared to quantitative computed tomography bone mineral density in patients undergoing cervical spinal surgery.

The vertebral bone quality (VBQ) score based on magnetic resonance imaging (MRI) was introduced as a bone quality marker in the lumbar spine. Prior studies showed that it could be utilized as a predictor of osteoporotic fracture or complications after instrumented spine surgery. The objective of this study was to evaluate the correlation between VBQ scores and bone mineral density (BMD) measured by quantitative computer tomography (QCT) in the cervical spine. Preoperative cervical CT and sagittal T1-weighted MRIs from patients undergoing ACDF were retrospectively reviewed and included. The VBQ score in each cervical level was calculated by dividing the signal intensity of the vertebral body by the signal intensity of the cerebrospinal fluid on midsagittal T1-weighted MRI images and correlated with QCT measurements of the C2-T1 vertebral bodies. A total of 102 patients (37.3% female) were included. VBQ values of C2-T1 vertebrae strongly correlated with each other. C2 showed the highest VBQ value [Median (range) 2.33 (1.33, 4.23)] and T1 showed the lowest VBQ value [Median (range) 1.64 (0.81, 3.88)]. There was significant weak to moderate negative correlations between and VBQ Scores for all levels [C2: p < 0.001; C3: p < 0.001; C4: p < 0.001; C5: p < 0.004; C6: p < 0.001; C7: p < 0.025; T1: p < 0.001]. Our results indicate that cervical VBQ scores may be insufficient in the estimation of BMDs, which might limit their clinical application. Additional studies are recommended to determine the utility of VBQ and QCT BMD to evaluate their potential use as bone status markers.

Transorbital sonography and MRI reliability to assess optic nerve sheath diameter in idiopathic intracranial hypertension.

The purpose of this study was to evaluate the performance of magnetic resonance imaging (MRI) in measuring the optic nerve sheath diameter (ONSD) compared to the established method transorbital sonography (TOS) in patients with idiopathic intracranial hypertension (IIH). Twenty-three patients with IIH were prospectively included applying IIH diagnostic criteria. All patients received a lumbar puncture with assessment of the cerebrospinal fluid (CSF) opening pressure to assure the IIH diagnosis. Measurement of ONSD was performed 3mm posterior to inner sclera surface in B-TOS by an expert examiner, while three independent neuroradiologists took measurements in axial T-weighted MRI examinations. The sella turcica with the pituitary gland (and potential presence of an empty sella) and the trigeminal cavity were also assessed on sagittal and transversal T1-weighted MRI images by one independent neuroradiologist. The means of ONSD between ultrasound and MRI measurements were 6.3mm (standard deviation [SD]=0.6mm) and 6.2mm (SD=0.8mm). The interrater reliability between three neuroradiologists showed a high interclass correlation coefficient (ICC) (confidence interval: .573<ICC<.8;p<.001). In patients with an empty sella, the ONSD evaluated by MRI was 6.6mm, while measuring 6.1mm in patients without empty sella. No correlation between CSF opening pressure and ONSD was found. MRI can reliably measure ONSD and yields similar results compared to TOS in patients with IIH. Moreover, patients with empty sella showed significantly larger ONSD than patients without empty sella.

Automatic Detection, Classification, and Grading of Lumbar Intervertebral Disc Degeneration Using an Artificial Neural Network Model.

Intervertebral disc degeneration (IDD) is a common cause of symptomatic axial low back pain. Magnetic resonance imaging (MRI) is currently the standard for the investigation and diagnosis of IDD. Deep learning artificial intelligence models represent a potential tool for rapidly and automatically detecting and visualizing IDD. This study investigated the use of deep convolutional neural networks (CNNs) for the detection, classification, and grading of IDD. Sagittal images of 1000 IDD T2-weighted MRI images from 515 adult patients with symptomatic low back pain were separated into 800 MRI images using annotation techniques to create a training dataset (80%) and 200 MRI images to create a test dataset (20%). The training dataset was cleaned, labeled, and annotated by a radiologist. All lumbar discs were classified for disc degeneration based on the Pfirrmann grading system. The deep learning CNN model was used for training in detecting and grading IDD. The results of the training with the CNN model were verified by testing the grading of the dataset using an automatic model. The training dataset of the sagittal intervertebral disc lumbar MRI images found 220 IDDs of grade I, 530 of grade II, 170 of grade III, 160 of grade IV, and 20 of grade V. The deep CNN model was able to detect and classify lumbar IDD with an accuracy of more than 95%. The deep CNN model can reliably automatically grade routine T2-weighted MRIs using the Pfirrmann grading system, providing a quick and efficient method for lumbar IDD classification.

Advanced Ultrasound Imaging of the Brain in Neonatal Hydrocephalus Using a Rat Hydrocephalus Model

Background and Objective:Preterm infants are at risk of developing post-hemorrhagic hydrocephalus. Earlier detection and intervention via placement of an intraventricular catheter improves neurologic and cognitive outcomes. However, intraventricular catheter placement is invasive with many risks, including hemorrhage and infection. Magnetic resonance imaging (MRI) can be used to detect ventriculomegaly and suggest hydrocephalus but cannot diagnose it. MRI is also difficult to obtain in preterm infants. Shear wave elastography (SWE) is an ultrasound technique measuring tissue stiffness and could serve as a rapid, bedside technique to detect hydrocephalus. We hypothesize that SWE stiffness measurements of the brain will positively correlate to MRI-derived ventricular volumes.Methods:Rats were bred to be wild-type, heterozygous, or homozygous for a Wpk mutation causing obstructive hydrocephalus. SWE and MRI of the rats were obtained at 7 and 14 days postnatally. Stiffness was measured using the Aixplorer SWE analysis tool. Lateral ventricular volumes were measured from the T2-weighted MRI images by planimetric analysis.Results:Stiffness and ventricular volumes were compared between genotypes via Kruskal-Wallis test at 7 and 14 days. SWE comparisons were 12.75 (p &lt; 0.05) and 16.54 (p &lt; 0.001), respectively. MRI volume comparisons were 18.78 (p &lt; 0.001) and 19.01 (p &lt; 0.001), respectively. MRI volume measurements were correlated to the SWE measurements via Spearman’s Rank correlation at 7 days (r = 0.539, p &lt; 0.05), 14 days (r = 0.488, p &lt;.05), and combined (r = 0.534, p &lt; 0.001), consistent with a moderate correlation.Conclusion and Potential Impact:This study supports the hypothesis that SWE-derived brain stiffness does significantly correlate with ventricular volumes. SWE has the potential to be used as a bedside tool for detecting neonatal hydrocephalus. SWE could reduce the need for MRI exams and invasive intraventricular catheter placements, allowing for more rapid diagnosis and thus intervention for newborns with hydrocephalus.

MRI anatomical investigation of rabbit prostate gland

The aim of the study was to provide diagnostic imaging data on rabbit prostate complex by means of magnetic resonance imaging (MRI) with regard to their use in morphological investigations of the gland. Six anaesthetised sexually mature clinically healthy New Zealand White rabbits, 8 months of age, with body weight 2.8-3.2 kg were used. The pelvic cavity and pelvic organs were investigated in the sagittal, dorsal and transversal planes from the 7th lumbar (L7) to the 1st coccygeal (C1) vertebra with 2-mm slice thickness using a Magnetom Essenza 1.5T tunnel MRI scanner. The prostate complex (proprostate, prostate and paraprostate parts) demonstrated signal hyperintensity on T1-weighted MRI images. The prostate capsule was hyperintense compared to glandular parenchyma. The prostate complex was visualised on slices between the caudal part of the first sacral and cranial part of the third sacral vertebrae. The prostate part signal was hyperintense on T2-weighted MRI images. The shape of the prostate complex in dorsal view was craniocaudally elongated, oval and localised bilate­rally from prostatic urethra in ventrodorsal view. The transverse T2 image of the pelvis through the 2nd sacral vertebra delineated the prostate part of the gland as a bilateral bilobed oval structure with hypointense parenchyma, embraced by a hyperintense capsule. The intrapelvic localisation and the shape of MRI image of healthy rabbit prostate complex is a species-specific morphological feature. The signal hyperintensity of rabbit prostate complex was higher on T2-weighted images. The T2 hyperintensity of prostate part vs hypointensity of proprostate and paraprostate parts provide evidence for the presence of substantial amount of glandular elements, presuming a dominating role of prostate part in the secretory function of the glandular complex.

Intravascular Large B-Cell Lymphoma (IVLBCL) Presenting with CNS Involvement in Patient with Chronic Lymphocytic Leukemia

Background: Intravascular lymphoma (IVL) is a rare, often fatal disease characterized by intraluminal proliferation of lymphoid cells within blood vessels. Intravascular Large B-Cell Lymphoma (IVLBCL) is frequently found in different organs, but the skin and central nervous system (CNS) are the most affected ones. Routine radiological examination and additional diagnostic tools, such as high-resolution, three-dimensional (3D) FLAIR sequence, T2-weighted (T2W) and T1-weighted (T1W) contrast-enhanced magnetic resonance imaging (MRI), MRI diffusion and perfusion imaging studies are useful to confirm the diagnosis in cases where it is not easy to perform histopathological examinations [1]. 1.2. Case Report: We present a clinical case of a 63-year-old female patient who presented with progressive cognitive deficit as mild motor aphasia, acalculia, agraphia, dyslexia and mild ideomotor apraxia without any signs of other organ involvement. It was known that the patient had 5-year anamnesis of chronic lymphocytic leukemia, stage II and splanchnic venous thrombosis 2 months ago. Blood biochemical analysis without any significant change, except leukocytosis in full blood count (leukocytes 15,2 x 10*9/L, where neutrophils were in normal range, but lymphocytes were increased). Head magnetic resonance imaging (MRI) showed pathological signal areas on the left parietal lobe with hypointense signal on contrast-enhanced T1-weighted, hyperintense signal on contrast-enhanced T2-weighted and FLAIR-weighted MRI images with restricted diffusion and slightly lower ADC value. Magnetic resonance perfusion imaging studies did not show an increase in cerebral blood flow or in cerebral blood volume. In this case, the patient’s diagnosis was made based on typical radiological features. 1.3. Conclusions: Intravascular large B-cell lymphoma is a rare subtype of extra nodal large B-cell lymphoma with an aggressive clinical course. The diagnosis is challenging, as it presents without any obvious tumour mass or lymphadenopathy. [2] As the clinical presentation is variable without specific characteristics, mostly biopsy or even autopsy confirms the definite diagnosis. But in some cases, the diagnosis can be made by typical radiological findings in the restricted availability of brain biopsy.

Prediction of Pedicle Screw Loosening Using an MRI-Based Vertebral Bone Quality Score in Patients with Lumbar Degenerative Disease

To explore a new magnetic resonance imaging (MRI)-based bone quality assessment method for predicting pedicle screw loosening in the lumbar spine. We reviewed 174 patients aged ≥50years who were treated for lumbar degenerative diseases using posterior lumbar interbody fusion. All patients were followed-up for ≥12months. Based on the presence of radiolucent areas on follow-up MRI images, the patients were divided into loosening and nonloosening groups. The vertebral bone quality (VBQ) score was calculated using T1-weighted MRI images. Demographic data, health history, and radiological parameters were also recorded and compared between the 2 groups. Logistic regression analysis was used to predict the independent risk factors affecting screw loosening. Screw loosening occurred in 29.88% (52/174) of patients. A total of 83 screws (9.18%, 83/904) were loosened. There were differences in the age, fixation length, fixation at S1, preoperative and postoperative PI-LL, PT, preoperative LL, lowest bone mineral density (BMD), and VBQ scores (P<0.05) between the nonloosening and loosening groups. In the logistic regression, the VBQ score (OR=1.02 per point; 95% CI: 1.01-1.03; P < 0.001) was identified as an independent factor influencing screw loosening. As an independent risk factor for screw loosening after lumbar spine fusion, the VBQ score provides a new noninvasive protocol for assessing bone quality during surgical planning.

Muscle impairment in MRI affect variability in treatment response to nusinersen in patients with spinal muscular atrophy type 2 and 3: A retrospective cohort study

BackgroundReal-world data have shown variability in treatment responses to nusinersen in spinal muscular atrophy (SMA). We investigated whether the magnitude of muscle impairment assessed by magnetic resonance imaging (MRI) at baseline can predict the treatment response. MethodsWe retrospectively assessed the clinical data in relevance to the thigh and pelvic MRI taken before the nusinersen treatment. A total of 16 patients with SMA types 2 and 3 (age = mean [SD]; 9.2 [4.6] year) receiving nusinersen treatment were enrolled. The T1-weighted MRI images of the pelvis and thigh were scored for muscle fatty infiltration and atrophy. The minimally clinically important difference (MCID) was considered as gaining at least 3 points of Hammersmith Functional Motor Scale-Expanded (HFMSE) from baseline. ResultsOf these 16 individuals, 14 had been treated for at least 15 months with baseline data. At 15 months, seven individuals obtained MCID in HFMSE. Baseline muscle MRI score could not differentiate the two groups; however, individuals who obtained MCID had significantly less severe scoliosis. In addition, there was a significant and negative relationship between baseline MRI score and the change of score in HFMSE after 15 months of treatment. Further, baseline Cobb angle along with MRI score also indicated the correlation to the degree of change in motor function. ConclusionThe degree of muscle damage may confer the variability in response to nusinersen in SMA types 2 and 3. Muscle MRI score along with the severity of scoliosis assessed at baseline may help to predict the motor function change.

Characterization of a fat tissue mimicking material for high intensity focused ultrasound applications.

Tissue-mimicking materials (TMMs) have a prominent role in validating new high intensity focused ultrasound (HIFU) therapies. Agar-based TMMs are often developed mimicking the thermal properties of muscle tissue, while TMMs simulating fat tissue properties are rarely developed. Herein, twelve agar-based TMMs were iteratively developed with varied concentrations of agar, water, glycerol and propan-2-ol, and characterized for their suitability in emulating the thermal conductivity of human fat tissue. Varied agar concentrations (2%, 4%, 6%, 8%, 12%, 16% and 20% w/v) were utilized for developing seven water-based TMMs, while a 20% w/v agar concentration was utilized for developing two water/alcohol-based TMMs (50% v/v water and 50% v/v either glycerol or propan-2-ol) and three alcohol-based TMMs (varied glycerol and propan-2-ol concentrations). Thermal conductivity was measured for all TMMs, and the tissue mimicking material (TMM) exhibiting thermal conductivity closest to human fat was considered the optimum fat TMM and was further characterized using ultrasound (US) and Magnetic Resonance Imaging (MRI). For the seven water-based TMMs an inverse linear trend was observed between thermal conductivity and increased agar concentration, being between 0.524 and 0.445W/mK. Alcohol addition decreased thermal conductivity of the two water/alcohol-based TMMs to about 0.33W/mK, while in the alcohol-based TMMs, increased concentrations of propan-2-ol emerged as a modifier of thermal conductivity. The optimum fat TMM (33.3% v/v glycerol and 66.7% v/v propan-2-ol) exhibited a 0.231W/mK thermal conductivity, and appeared hypoechoic on US images and with increased brightness on T1-Weighted MRI images. The optimum fat TMM emulates the thermal conductivity of human fat tissue and exhibits a fat-like appearance on US and MRI images. The TMM is cost-effective and has a long lifespan and possesses great potential for use in HIFU applications as a fat TMM.

Advantages of CT Versus MRI for Preoperative Assessment of Osteochondritis Dissecans of the Capitellum

Background: Magnetic resonance imaging (MRI) is considered to be the gold standard for imaging of osteochondritis dissecans (OCD). Purpose/Hypothesis: The purpose was to determine the additional value of a preoperative computed tomography (CT) scan in adolescent patients with capitellar OCD of the elbow. Consistent with the fact that OCD is a lesion involving the subchondral bone, the hypothesis was that CT would be superior to MRI for imaging OCD of the capitellum. Study Design: Cohort study (diagnosis); Level of evidence, 3. Methods: All patients being treated surgically for an OCD of the capitellum between 2006 and 2016 at one institution were reviewed for preoperative imaging. A total of 28 patients met the inclusion criteria. Corresponding MRI and CT scans were compared retrospectively. Multiple parameters were recorded, with special emphasis on OCD lesion size, fragmentation, and tilt as well as joint surface integrity, loose bodies, and osteophytes. Results: The OCD lesions were best seen on CT scans, whereas MRI T1-weighted images overestimated and T2-weighted images underestimated the size of defects. A subchondral fracture nonunion was found on CT scans in 18 patients, whereas this was seen on MRI T1-weighted images in only 2 patients (P < .001) and MRI T2-weighted images in 4 patients (P < .001). Fragmentation of the OCD fragment was found on CT scans in 17 patients but on MRI scans in only 9 patients (P = .05). Osteophytes as a sign of secondary degenerative changes were seen on CT scans in 24 patients and were seen on MRI scans in 15 patients (P = .02). Altogether, only 51 of 89 secondary changes including loose bodies, effects on the radial head and ulnohumeral joint, and osteophytes that were seen on CT scans were also seen on MRI scans (P = .002). Conclusion: OCD fragmentation and secondary changes were more often diagnosed on CT. These factors indicate OCD instability or advanced OCD stages, which are indications for surgery. In an adolescent who is considered at risk for OCD (baseball, gymnastics, weightlifting, tennis) and who has lateral elbow joint pain with axial or valgus load bearing, CT is our imaging modality of choice for diagnosing and staging OCD of the capitellum.

MRI parameters for predicting the effect of ultrasound-guided high-intensity focused ultrasound in the ablation of uterine fibroids

To study the value of magnetic resonance imaging (MRI) parameters in predicting the efficacy of ultrasonic ablation of fibroids. A total of 91 patients were divided into groups based on non-perfused volume (NPV) ratio and blood supply type. The preoperative MRI parameters were measured and analysed. A correlation analysis between the MRI parameters and the NPV ratio was performed. Receiver operating characteristic (ROC) curves were used to analyse and determine the cut-off value of MRI parameters to predict the ablation rate of fibroids. The uterine fibroids group with an NPV ratio <80% and the group with an NPV ratio of ≥80% had significant differences in signal intensity (SI) at MRI T2-weighted imaging (WI), fibroid-to-rectus abdominis SI ratio (SIR) at T2WI, and blood supply type (p<0.05). There were no significant differences in fibroid volume, T2WI signal uniformity, and apparent diffusion coefficient (ADC) values. The ADC value and SI and SIR at MRI T2WI in the group with poor blood supply were lower than those in the group with a rich blood supply (p<0.05). SI at MRI T2WI correlated negatively with the NPV ratio. The cut-off values for SI and SIR at MRI T2WI of fibroids whose NPV ratio exceeds 80% were 220.58 and 1.315, respectively. SI at MRI T2WI and blood supply type could be predictors of the efficacy of ablation. Ultrasonic ablation of fibroids with MRI T2WI hyperintensity and a rich blood supply had poor efficacy.

GUBS: Graph-Based Unsupervised Brain Segmentation in MRI Images.

Brain segmentation in magnetic resonance imaging (MRI) images is the process of isolating the brain from non-brain tissues to simplify the further analysis, such as detecting pathology or calculating volumes. This paper proposes a Graph-based Unsupervised Brain Segmentation (GUBS) that processes 3D MRI images and segments them into brain, non-brain tissues, and backgrounds. GUBS first constructs an adjacency graph from a preprocessed MRI image, weights it by the difference between voxel intensities, and computes its minimum spanning tree (MST). It then uses domain knowledge about the different regions of MRIs to sample representative points from the brain, non-brain, and background regions of the MRI image. The adjacency graph nodes corresponding to sampled points in each region are identified and used as the terminal nodes for paths connecting the regions in the MST. GUBS then computes a subgraph of the MST by first removing the longest edge of the path connecting the terminal nodes in the brain and other regions, followed by removing the longest edge of the path connecting non-brain and background regions. This process results in three labeled, connected components, whose labels are used to segment the brain, non-brain tissues, and the background. GUBS was tested by segmenting 3D T1 weighted MRI images from three publicly available data sets. GUBS shows comparable results to the state-of-the-art methods in terms of performance. However, many competing methods rely on having labeled data available for training. Labeling is a time-intensive and costly process, and a big advantage of GUBS is that it does not require labels.

Differentiating Primary Tumors for Brain Metastasis with Integrated Radiomics from Multiple Imaging Modalities

Objectives To differentiate the primary site of brain metastases (BMs) is of high clinical value for the successful management of patients with BM. The purpose of this study is to investigate a combined radiomics model with computer tomography (CT) and magnetic resonance imaging (MRI) images in differentiating BMs originated from lung and breast cancer. Methods Pretreatment cerebral contrast enhanced CT and T1-weighted MRI images of 78 patients with 179 BMs from primary lung and breast cancer were retrospectively analyzed. Radiomic features were extracted from contoured BM lesions and selected using the Mann–Whitney U test and the least absolute shrinkage and selection operator (LASSO) logistic regression. Binary logistic regression (BLR) and support vector machine (SVM) models were built and evaluated based on selected radiomic features from CT alone, MRI alone, and combined images to differentiate BMs originated from lung and breast cancer. Results A total of 10 and 6 optimal radiomic features were screened out of 1288 CT and 1197 MRI features, respectively. The mean area under the curves (AUCs) of the BLR and SVM models using fivefolds cross-validation were 0.703 vs. 0.751, 0.718 vs. 0.754, and 0.781 vs. 0.803 in the training dataset and 0.708 vs. 0.763, 0.715 vs. 0.717, and 0.771 vs. 0.805 in the testing dataset for models with CT alone, MRI alone, and combined CT and MRI radiomic features, respectively. Conclusions Radiomics model based on combined CT and MRI features is feasible and accurate in the differentiation of the primary site of BMs from lung and breast cancer.

Complete Response Evaluation of Locally Advanced Rectal Cancer to Neoadjuvant Chemoradiotherapy Using Textural Features Obtained from T2 Weighted Imaging and ADC Maps.

The prediction of pathological responses for locally advanced rectal cancer using magnetic resonance imaging (MRI) after neoadjuvant chemoradiotherapy (CRT) is a challenging task for radiologists, as residual tumor cells can be mistaken for fibrosis. Texture analysis of MR images has been proposed to understand the underlying pathology. This study aimed to assess the responses of lesions to CRT in patients with locally advanced rectal cancer using the first-order textural features of MRI T2-weighted imaging (T2-WI) and apparent diffusion coefficient (ADC) maps. Forty-four patients with locally advanced rectal cancer (median age: 57 years) who underwent MRI before and after CRT were enrolled in this retrospective study. The first-order textural parameters of tumors on T2-WI and ADC maps were extracted. The textural features of lesions in pathologic complete responders were compared to partial responders using Student's t- or Mann-Whitney U tests. A comparison of textural features before and after CRT for each group was performed using the Wilcoxon rank sum test. Receiver operating characteristic curves were calculated to detect the diagnostic performance of the ADC. Of the 44 patients evaluated, 22 (50%) were placed in a partial response group and 50% were placed in a complete response group. The ADC changes of the complete responders were statistically more significant than those of the partial responders (P = 0.002). Pathologic total response was predicted with an ADC cut-off of 1310 x 10-6 mm2/s, with a sensitivity of 72%, a specificity of 77%, and an accuracy of 78.1% after neoadjuvant CRT. The skewness of the T2-WI before and after neoadjuvant CRT showed a significant difference in the complete response group compared to the partial response group (P = 0.001 for complete responders vs. P = 0.482 for partial responders). Also, relative T2-WI signal intensity in the complete response group was statistically lower than that of the partial response group after neoadjuvant CRT (P = 0.006). As a result of the conversion of tumor cells to fibrosis, the skewness of the T2-WI before and after neoadjuvant CRT was statistically different in the complete response group compared to the partial response group, and the complete response group showed statistically lower relative T2-WI signal intensity than the partial response group after neoadjuvant CRT. Additionally, the ADC cut-off value of 1310 × 10-6 mm2/s could be used as a marker for a complete response along with absolute ADC value changes within this dataset.

Radiomics of locally advanced rectal cancer: machine learning-based prediction of response to neoadjuvant chemoradiotherapy using pre-treatment sagittal T2-weighted MRI.

Variable response to neoadjuvant chemoradiotherapy (nCRT) is observed among individuals with locally advanced rectal cancer (LARC), having a significant impact on patient management. In this work, we aimed to investigate the potential value of machine learning (ML)-based magnetic resonance imaging (MRI) radiomics in predicting therapeutic response to nCRT in patients with LARC. Seventy-six patients with LARC were included in this retrospective study. Radiomic features were extracted from pre-treatment sagittal T2-weighted MRI images, with 3D segmentation. Dimension reduction was performed with a reliability analysis, pair-wise correlation analysis, analysis of variance, recursive feature elimination, Kruskal-Wallis, and Relief methods. Models were created using four different algorithms. In addition to radiomic models, clinical only and different combined models were developed and compared. The reference standard was tumor regression grade (TRG) based on the Modified Ryan Scheme (TRG 0 vs TRG 1-3). Models were compared based on net reclassification index (NRI). Clinical utility was assessed with decision curve analysis (DCA). Number of features with excellent reliability is 106. The best result was achieved with radiomic only model using eight features. The area under the curve (AUC), accuracy, sensitivity,and specificity for validation were 0.753 (standard deviation [SD], 0.082), 81.1%, 83.8%, and 75.0%; for testing, 0.705 (SD, 0.145), 73.9%, 81.2%,and 57.1%, respectively. Based on the clinical only model as reference, NRI for radiomic only model was the best. DCA also showed better clinical utility for radiomic only model. ML-based T2-weighted MRI radiomics might have a potential in predicting response to nCRT in patients with LARC.

Bi-MGAN: Bidirectional T1-to-T2 MRI images prediction using multi-generative multi-adversarial nets

Magnetic Resonance Imaging (MRI) is a widely-used detection technology both in clinical treatment and image-guide therapy, and it can yield multi-modal medical images. However, MRI technique requires expensive costs and long-time process, which causes extra financial burden and low throughput. Therefore, it is necessary to estimate desired modality from acquired modality without real acquisition, which can effectively decrease financial cost and indirectly improve throughput. Considering that current unidirectional methods can only predict one modality fixedly, we propose a bidirectional prediction method by using multi-generative multi-adversarial nets (Bi-MGAN) to predict desired modality from acquired modality in paired and unpaired fashion. Our method simultaneously learns two nonlinear mappings between T1-weighted MRI and T2-weighted MRI image. Within it, pathological label is introduced to ensure invariance of pathological information. Then, deep and handcrafted features are fused to preserve consistency of context feature and anatomical structure. Besides, spectral normalization is utilized to control the performance of discriminator and accelerate learning process, and local regularization is used to reduce generalization error of predictors. Experimental results show that our Bi-MGAN has preserved pathological feature and anatomical structure comparing with other state-of-the-art methods. It gains significant increments on paired dataset and obtains competitive results on unpaired dataset. Specifically, it achieves average increment of 57.1% SSIM, 47.1% FSIM and 50.0% MSIM on paired SPLP dataset, and obtains 9.1% SSIM, 3.6% FSIM and 9.6% MSIM on unpaired Brain dataset comparing with others.

Prognostic effect of increased postoperative MRI T2WI high signal intensity in degenerative cervical myelopathy

BACKGROUND CONTEXTHigh cord signals (HCS) are common in patients undergoing surgery for degenerative cervical myelopathy (DCM). Few studies have investigated the prognostic effects of postoperative HCS changes. PURPOSETo investigate whether different changes in the postoperative magnetic resonance imaging (MRI) T2-weighted images (T2WIs) signal intensity ratio (SIR) is related to patient outcomes, predictors of unsatisfactory recovery, and the factors associated with changes in postoperative SIR. STUDY DESIGNA retrospective chart review at a single institution. PATIENT SAMPLEThe study population included all consecutive patients who showed HCS in preoperative MRI T2WIs and underwent double-door laminoplasty for DCM from December 2017 to December 2020. OUTCOME MEASURESPatient self-reported measures included the Japanese Orthopedic Association (JOA) score, 36-Item Short Form Survey (SF-36) physical component score, and SF-36 mental component score. Imaging measures included SIR, length of HCS, and canal narrowing ratio (CNR). METHODSWe reviewed patient records and analyzed the statistical associations of MRI T2WI measures with the JOA score, SF-36 physical, and mental component scores. RESULTSFifty-three patients were categorized into three groups based on the postoperative HCS changes, identified from MRI T2WI before and after surgery: reduced (Group A, N=26); unchanged (Group B, N=12); and increased (Group C, N=15). The neurological recovery rates 12 months after surgery were 67.72%±17.45% in Group A, 51.53%±16.00% in Group B, and 13.35%±21.35% in Group C (p<.001). Significant differences across the three groups were found in symptom duration, postoperative SIR and length of HCS, pre- and postoperative CNR, recovery rate, JOA, SF-36 scores, with patients in Group C having the worst outcomes. Longer DCM symptom duration, greater preoperative CNR, and increased postoperative HCS were prognostic factors for a recovery rate<50%. Preoperative CNR, with an optimal threshold of 57.303%, was an independent risk factor for increased postoperative HCS. CONCLUSIONSLess than one-third of DCM patients with preoperative HCS exhibited an increase in HCS after double-door laminoplasty and reported worse outcomes at the 12-month follow-up when compared to patients with decreased or unchanged postoperative HCS.