Automatic detection, segmentation, and classification of primary bone tumors and bone infections using an ensemble multi-task deep learning framework on multi-parametric MRIs: a multi-center study.

OBJECTIVES: To develop an ensemble multi-task deep learning (DL) framework for automatic and simultaneous detection, segmentation, and classification of primary bone tumors (PBTs) and bone infections based on multi-parametric MRI from multi-center. METHODS: This retrospective study divided 749 patients with PBTs or bone infections from two hospitals into a training set (N = 557), an internal validation set (N = 139), and an external validation set (N = 53). The ensemble framework was constructed using T1-weighted image (T1WI), T2-weighted image (T2WI), and clinical characteristics for binary (PBTs/bone infections) and three-category (benign/intermediate/malignant PBTs) classification. The detection and segmentation performances were evaluated using Intersection over Union (IoU) and Dice score. The classification performance was evaluated using the receiver operating characteristic (ROC) curve and compared with radiologist interpretations. RESULT: On the external validation set, the single T1WI-based and T2WI-based multi-task models obtained IoUs of 0.71 ± 0.25/0.65 ± 0.30 for detection and Dice scores of 0.75 ± 0.26/0.70 ± 0.33 for segmentation. The framework achieved AUCs of 0.959 (95%CI, 0.955-1.000)/0.900 (95%CI, 0.773-0.100) and accuracies of 90.6% (95%CI, 79.7-95.9%)/78.3% (95%CI, 58.1-90.3%) for the binary/three-category classification. Meanwhile, for the three-category classification, the performance of the framework was superior to that of three junior radiologists (accuracy: 65.2%, 69.6%, and 69.6%, respectively) and comparable to that of two senior radiologists (accuracy: 78.3% and 78.3%). CONCLUSION: The MRI-based ensemble multi-task framework shows promising performance in automatically and simultaneously detecting, segmenting, and classifying PBTs and bone infections, which was preferable to junior radiologists. CLINICAL RELEVANCE STATEMENT: Compared with junior radiologists, the ensemble multi-task deep learning framework effectively improves differential diagnosis for patients with primary bone tumors or bone infections. This finding may help physicians make treatment decisions and enable timely treatment of patients. KEY POINTS: • The ensemble framework fusing multi-parametric MRI and clinical characteristics effectively improves the classification ability of single-modality models. • The ensemble multi-task deep learning framework performed well in detecting, segmenting, and classifying primary bone tumors and bone infections. • The ensemble framework achieves an optimal classification performance superior to junior radiologists' interpretations, assisting the clinical differential diagnosis of primary bone tumors and bone infections.

Occipital condyle syndrome due to small-cell lung carcinoma: A case report.

Skull base metastases, including those from small-cell lung carcinoma (SCLC), can present with various syndromes depending on the site of involvement, such as orbital syndrome, parasellar syndrome, middle fossa syndrome, jugular foramen syndrome, and occipital condyle syndrome (OCS). One such example is OCS, which consists of unilateral occipital headache accompanied with ipsilateral hypoglossal palsy. This case report describes a 51-year-old man initially diagnosed with OCS, which led to the discovery of systemic bone metastases from SCLC. Magnetic resonance imaging showed lesions in the occipital condyle and hypoglossal canal, while positron emission tomography-computed tomography identified a lung mass and widespread metastases. SCLC is highly aggressive and metastatic, with the bone being a common site of spread. In this case, the OCS preceded the diagnosis of the underlying malignancy. Prompt diagnosis and treatment are crucial, as patients with OCS often have advanced disease. This case highlights the importance of considering SCLC as a potential etiology for OCS, given the propensity for bone metastases. Early recognition and evaluation of OCS is essential to initiate appropriate management.

Analysis of the initial orbital MRI in aquaporin-4 antibody-positive optic neuritis (AQP4-ON): lesion location and lesion length can be predictive of visual prognosis.

PURPOSE: Despite mounting evidence indicating that aquaporin-4 antibody-positive optic neuritis (AQP4-ON) presents a less favorable prognosis than other types of optic neuritis, there exists substantial heterogeneity in the prognostic outcomes within the AQP4-ON cohort. Considering the persistent debate over the role of MRI in assessing the prognosis of optic neuritis, we aim to investigate the correlation between the MRI appearance and long-term visual prognosis in AQP4-ON patients. METHODS: We retrospectively reviewed the ophthalmological and imaging data of AQP4-ON patients admitted to our Neuro-ophthalmology Department from January 2015 to March 2018, with consecutive follow-up visits for a minimum of 3 years. RESULTS: A total of 51 AQP4-ON patients (59 eyes) meeting the criteria were enrolled in this research. After assessing the initial orbital MR images of each patient at the first onset, we observed the involvement of the canalicular segment (p < 0.001), intracranial segment (p = 0.004), optic chiasm (p = 0.009), and the presence of LEON (p = 0.002) were significantly different between recovery group and impairment group. For quantitative measurement, the length of the lesions is significantly higher in the impairment group (20.1 ± 9.3 mm) than in the recovery group (12.5 ± 5.3 mm) (p = 0.001). CONCLUSION: AQP4-ON patients with involvement of canalicular, intracranial segment and optic chiasm of the optic nerve, and the longer range of lesions threaten worse vision prognoses. Timely MR examination during the initial acute phase can not only exclude the intracranial or orbital mass lesions but also indicate visual prognosis in the long term.

Clinical features and prognostic factors of patients with cancer-associated stroke.

BACKGROUND: Cerebrovascular diseases in cancer patients significantly aggravate their condition and prognosis; therefore, prompt and accurate diagnosis and treatment are important. The purpose of this study was to investigate patient demographics, laboratory data, brain magnetic resonance imaging (MRI) findings, and prognosis among patients with stroke and cancer, especially cancer-associated ischemic stroke (CAIS). METHODS: We performed a retrospective, single-center study. We enrolled consecutive patients who had acute stroke and were admitted to our hospital between January 2011 and December 2021. We collected general demographic characteristics, cancer histopathological type, laboratory data, brain MRI findings, and prognosis data. RESULTS: Among 2040 patients with acute stroke, a total of 160 patients (7.8%) had active cancer. The types of strokes were cerebral infarction, cerebral hemorrhage, subarachnoid hemorrhage, and transient ischemic attack in 124, 25, 5, and 6 patients, respectively. Among the patients with ischemic stroke, there were 69 cases of CAIS. Pancreas and adenocarcinoma were the most frequent types of primary tumor and histopathology. Patients with adenocarcinoma and those with cerebral infarctions in both bilateral anterior and posterior cerebral circulation areas showed higher D-dimer levels. Pancreatic cancer and high plasma D-dimer levels were associated with poor survival rate. CONCLUSION: CAIS was seen more frequently in patients with pancreatic cancer and adenocarcinoma. Pancreatic cancer and high plasma D-dimer levels were potential factors of poor prognosis in patients with CAIS.

Neurocognitive Function with Conventional Hemodialysis versus Post-Dilution Hemofiltration as Initial Treatment in ESKD Patients: A Randomized Controlled Trial - The DA-VINCI Study.

INTRODUCTION: The ideal modality choice and dialysis prescription during the first renal replacement therapy (RRT) session remain unclear. We conducted a pilot study to determine the safety risk for hemodialysis (HD) versus hemofiltration (HF) and its relationship with neurocognitive assessment on incident RRT patients. METHODS: Twenty-four incident RRT patients were included. Patients were randomized to the conventional HD group or post-dilution HF group. Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MOCA) tests were applied in all patients before and after session, and brain magnetic resonance image (MRI) was performed in 7 patients from the conventional HD group and 8 patients from the post-dilution HF group before and after the intervention. RESULTS: Baseline characteristics were similar between groups. Compared to conventional HD, post-dilution HF had longer treatment time. There were no significant changes in blood pressure after RRT in both groups. The MMSE test showed no significant differences between groups or within groups. The MOCA test showed an increase in the total score for the post-dilution HF group with no significant changes between groups. The MRI evaluation showed no differences between or within groups. CONCLUSION: Post-dilution HF is a safe alternative for the first HD session in incident RRT; it allows longer treatment time if ultrafiltration is required and has a similar neurological risk than conventional HD. This is a pilot study and that larger studies are needed to confirm the findings.

The difference of functional MR imaging in evaluating outcome of patients with diffuse and compact brain arteriovenous malformation.

Microsurgical resection is an effective method to treat brain arteriovenous malformations (BAVMs). Functional magnetic resonance imaging (fMRI) can evaluate the spatial relationship of nidus and eloquent. Diffuse BAVMs are related to poor outcomes postoperatively. The role of fMRI in evaluating outcomes in patients with different nidus types remains unclear. BAVM patients received microsurgical resection were included from a prospective, multicenter cohort study. All patients underwent fMRI evaluation preoperatively and were regularly followed up postoperatively. Diffuse BAVM is radiologically identified as nidus containing normal brain tissue interspersing between malformed vessels. Lesion-to-eloquent distance (LED) was calculated based on the relationship between nidus and eloquent. The primary outcome was 180-day unfavorable neurological status postoperatively. The risk of primary outcome was investigated within different BAVM nidus types. The LED's performance to predict poor outcome was evaluated using area under curve (AUC). 346 BAVM patients were included in this study. 93 (26.9%) patients were found to have a 180-day unfavorable outcome. Multivariate logistic analysis demonstrated LED (odd ratio [OR], 0.44; 0.34-0.57; P < 0.001) and mRS at admission (OR, 2.59; 1.90-3.54; P < 0.001) as factors of unfavorable outcome. Subgroup analysis showed LED and mRS at admission as factors of unfavorable outcome for patients with compact BAVMs (all P < 0.05), but not for patients with diffuse BAVMs. Subsequent analysis showed that LED performed poorly to predict the unfavorable outcome for patients with diffuse BAVMs, compared with patients with compact BAVMs (AUC as 0.69 vs. 0.86, P < 0.05). A larger cutoff value of LED to unfavorable outcome was found in patients with diffuse BAVMs (15 mm) compared with patients with compact BAVMs (4.7 mm). Usage of LED to evaluate postoperative outcome of patients with diffuse BAVMs differs from its use in patients with compact BAVMs. Specific assessment strategy considering BAVM nidus types could help improve patients' outcome. MITASREAVM cohort (unique identifier: NCT02868008, https://clinicaltrials.gov/study/NCT02868008?term=NCT02868008&rank=1 ).

Evaluation and mitigation of deformable image registration uncertainties for MRI-guided adaptive radiotherapy.

PURPOSE: We evaluate the performance of a deformable image registration (DIR) software package in registering abdominal magnetic resonance images (MRIs) and then develop a mechanical modeling method to mitigate detected DIR uncertainties. MATERIALS AND METHODS: Three evaluation metrics, namely mean displacement to agreement (MDA), DICE similarity coefficient (DSC), and standard deviation of Jacobian determinants (STD-JD), are used to assess the multi-modality (MM), contour-consistency (CC), and image-intensity (II)-based DIR algorithms in the MIM software package, as well as an in-house developed, contour matching-based finite element method (CM-FEM). Furthermore, we develop a hybrid FEM registration technique to modify the displacement vector field of each MIM registration. The MIM and FEM registrations were evaluated on MRIs obtained from 10 abdominal cancer patients. One-tailed Wilcoxon-Mann-Whitney (WMW) tests were conducted to compare the MIM registrations with their FEM modifications. RESULTS: For the registrations performed with the MIM-CC, MIM-MM, MIM-II, and CM-FEM algorithms, their average MDAs are 0.62 ± 0.27, 2.39 ± 1.30, 3.07 ± 2.42, 1.04 ± 0.72 mm, and average DSCs are 0.94 ± 0.03, 0.80 ± 0.12, 0.77 ± 0.15, 0.90 ± 0.11, respectively. The p-values of the WMW tests between the MIM registrations and their FEM modifications are less than 0.0084 for STD-JDs and greater than 0.87 for MDA and DSC. CONCLUSIONS: Among the three MIM DIR algorithms, MIM-CC shows the smallest errors in terms of MDA and DSC but exhibits significant Jacobian uncertainties in the interior regions of abdominal organs. The hybrid FEM technique effectively mitigates the Jacobian uncertainties in these regions.

Pre-and post-surgical non-functional pituitary adenomas and their relationship with high levels of serum glucose.

OBJECTIVES: This clinical, analytical, retro-prospective, auto-controlled, not randomized, and not blinded study, aimed to investigate the association of changes in the serum glucose levels with the pre-and-post changes in the size tumor in mm3 in the Non-Functional pituitary adenomas. METHODS: Pre-and post-surgical MRI, as well as the measurements in the serum glucose levels and immunohistochemical techniques were performed in all the patients in the study, with a mean followed-up until 208.57 days. A comparison was made between the reductions in tumor size of hormonally active pituitary adenomas (HSPAs) vs NFPAs. RESULTS: Seventy-four patients were included in this study, of whom, 46 were NFPAs. The decrease in the NFPAs tumor size after surgery was statistically significant (P ≤ 0.0001). The Mean of the differences of both type of tumors in mm3 were -9552 ± 10287. Pre-surgery, the mean of the HSPAs were 8.923 ± 2.078; and the NFPAs were 14.161 ± 1.912. The differences in the tumor size were statistically significant (p = 0.039). Post-surgical, the mean of the HSPAs were 2.079 ± 971, with a (p = 0.14): and the NFPAs were 4.609 ± 1.205. After surgery of the NFPAs, most of the patients-maintained serum levels ≤ 100 mg/dL, with a statistical significance (P ≤ 0.0003). CONCLUSION: This study demonstrates for the first time the correlation between the presence of pre-and post- surgical changes in the NFPAs, with modifications in the levels of serum glucose, and the comparison, pre- and post-surgical between the tumor size of HSPAs and NFPAs.

A Semi-Automatic Magnetic Resonance Imaging Annotation Algorithm Based on Semi-Weakly Supervised Learning.

The annotation of magnetic resonance imaging (MRI) images plays an important role in deep learning-based MRI segmentation tasks. Semi-automatic annotation algorithms are helpful for improving the efficiency and reducing the difficulty of MRI image annotation. However, the existing semi-automatic annotation algorithms based on deep learning have poor pre-annotation performance in the case of insufficient segmentation labels. In this paper, we propose a semi-automatic MRI annotation algorithm based on semi-weakly supervised learning. In order to achieve a better pre-annotation performance in the case of insufficient segmentation labels, semi-supervised and weakly supervised learning were introduced, and a semi-weakly supervised learning segmentation algorithm based on sparse labels was proposed. In addition, in order to improve the contribution rate of a single segmentation label to the performance of the pre-annotation model, an iterative annotation strategy based on active learning was designed. The experimental results on public MRI datasets show that the proposed algorithm achieved an equivalent pre-annotation performance when the number of segmentation labels was much less than that of the fully supervised learning algorithm, which proves the effectiveness of the proposed algorithm.

Magnetic resonance imaging findings in patients with dropped head syndrome.

BACKGROUND: Dropped head syndrome (DHS) is difficult to diagnose only by clinical examination. Although characteristic images on X-rays of DHS have been studied, changes in soft tissue of the disease have remained largely unknown. Magnetic resonance imaging (MRI) is useful for evaluating soft tissue, and we therefore performed this study with the purpose of investigating the characteristic signal changes of DHS on MRI by a comparison with those of cervical spondylosis. METHODS: The study involved 35 patients diagnosed with DHS within 6 months after the onset and 32 patients with cervical spondylosis as control. The signal changes in cervical extensor muscles, interspinous tissue, anterior longitudinal ligament (ALL) and Modic change on MRI were analyzed. RESULTS: Signal changes of cervical extensor muscles were 51.4% in DHS and 6.3% in the control group, those of interspinous tissue were 85.7% and 18.8%, and those of ALL were 80.0% and 21.9%, respectively, suggesting that the frequency of signal changes of cervical extensor muscles, interspinous tissue and ALL was significantly higher in the DHS group (p < 0.05). The presence of Modic change of acute phase (Modic type I) was also significantly higher in the DHS group than in the control group (p < 0.001). CONCLUSION: MRI findings of DHS within 6 months after the onset presented the characteristic signal changes in cervical extensor muscles, interspinous tissue, ALL and Modic change. Evaluation of MRI signal changes is useful for an objective evaluation of DHS.

Abnormal brain activities of cognitive processes in cerebral small vessel disease: A systematic review of task fMRI studies.

Cerebral small vessel disease (CSVD) is characterized by widespread functional changes in the brain, as evident from abnormal brain activations during cognitive tasks. However, the existing findings in this area are not yet conclusive. We systematically reviewed 25 studies reporting task-related fMRI in five cognitive domains in CSVD, namely executive function, working memory, processing speed, motor, and affective processing. The findings highlighted: (1) CSVD affects cognitive processes in a domain-specific manner; (2) Compensatory and regulatory effects were observed simultaneously in CSVD, which may reflect the interplay between the negative impact of brain lesion and the positive impact of cognitive reserve. Combined with behavioral and functional findings in CSVD, we proposed an integrated model to illustrate the relationship between altered activations and behavioral performance in different stages of CSVD: functional brain changes may precede and be more sensitive than behavioral impairments in the early pre-symptomatic stage; Meanwhile, compensatory and regulatory mechanisms often occur in the early stages of the disease, while dysfunction/decompensation and dysregulation often occur in the late stages. Overall, abnormal hyper-/hypo-activations are crucial for understanding the mechanisms of small vessel lesion-induced behavioral dysfunction, identifying potential neuromarker and developing interventions to mitigate the impact of CSVD on cognitive function.

Multimodal hybrid convolutional neural network based brain tumor grade classification.

An abnormal growth or fatty mass of cells in the brain is called a tumor. They can be either healthy (normal) or become cancerous, depending on the structure of their cells. This can result in increased pressure within the cranium, potentially causing damage to the brain or even death. As a result, diagnostic procedures such as computed tomography, magnetic resonance imaging, and positron emission tomography, as well as blood and urine tests, are used to identify brain tumors. However, these methods can be labor-intensive and sometimes yield inaccurate results. Instead of these time-consuming methods, deep learning models are employed because they are less time-consuming, require less expensive equipment, produce more accurate results, and are easy to set up. In this study, we propose a method based on transfer learning, utilizing the pre-trained VGG-19 model. This approach has been enhanced by applying a customized convolutional neural network framework and combining it with pre-processing methods, including normalization and data augmentation. For training and testing, our proposed model used 80% and 20% of the images from the dataset, respectively. Our proposed method achieved remarkable success, with an accuracy rate of 99.43%, a sensitivity of 98.73%, and a specificity of 97.21%. The dataset, sourced from Kaggle for training purposes, consists of 407 images, including 257 depicting brain tumors and 150 without tumors. These models could be utilized to develop clinically useful solutions for identifying brain tumors in CT images based on these outcomes.

In vivo characterization of magnetic resonance imaging-based T1w/T2w ratios reveals myelin-related changes in temporal lobe epilepsy.

Temporal lobe epilepsy (TLE) is the most common type of intractable epilepsy in adults. Although brain myelination alterations have been observed in TLE, it remains unclear how the myelination network changes in TLE. This study developed a novel method in characterization of myelination structural covariance network (mSCN) by T1-weighted and T2-weighted magnetic resonance imaging (MRI). The mSCNs were estimated in 42 left TLE (LTLE), 42 right TLE (RTLE) patients, and 41 healthy controls (HCs). The topology of mSCN was analyzed by graph theory. Voxel-wise comparisons of myelination laterality were also examined among the three groups. Compared to HC, both patient groups showed decreased myelination in frontotemporal regions, amygdala, and thalamus; however, the LTLE showed lower myelination in left medial temporal regions than RTLE. Moreover, the LTLE exhibited decreased global efficiency compared with HC and more increased connections than RTLE. The laterality in putamen was differently altered between the two patient groups: higher laterality at posterior putamen in LTLE and higher laterality at anterior putamen in RTLE. The putamen may play a transfer station role in damage spreading induced by epileptic seizures from the hippocampus. This study provided a novel workflow by combination of T1-weighted and T2-weighted MRI to investigate in vivo the myelin-related microstructural feature in epileptic patients first time. Disconnections of mSCN implicate that TLE is a system disorder with widespread disruptions at regional and network levels.

Dynamic contrast-enhanced MRI radiomics model predicts epidermal growth factor receptor amplification in glioblastoma, IDH-wildtype.

PURPOSE: To develop and validate a dynamic contrast-enhanced (DCE) MRI-based radiomics model to predict epidermal growth factor receptor (EGFR) amplification in patients with glioblastoma, isocitrate dehydrogenase (IDH) wildtype. METHODS: Patients with pathologically confirmed glioblastoma, IDH wildtype, from January 2015 to December 2020, with an EGFR amplification status, were included. Patients who did not undergo DCE or conventional brain MRI were excluded. Patients were categorized into training and test sets by a ratio of 7:3. DCE MRI data were used to generate volume transfer constant (Ktrans) and extracellular volume fraction (Ve) maps. Ktrans, Ve, and conventional MRI were then used to extract the radiomics features, from which the prediction models for EGFR amplification status were developed and validated. RESULTS: A total of 190 patients (mean age, 59.9; male, 55.3%), divided into training (n = 133) and test (n = 57) sets, were enrolled. In the test set, the radiomics model using the Ktrans map exhibited the highest area under the receiver operating characteristic curve (AUROC), 0.80 (95% confidence interval [CI], 0.65-0.95). The AUROC for the Ve map-based and conventional MRI-based models were 0.74 (95% CI, 0.58-0.90) and 0.76 (95% CI, 0.61-0.91). CONCLUSION: The DCE MRI-based radiomics model that predicts EGFR amplification in glioblastoma, IDH wildtype, was developed and validated. The MRI-based radiomics model using the Ktrans map has higher AUROC than conventional MRI.

Ablative 5-Fraction Stereotactic MRI-Guided Adaptive Radiotherapy for Oligometastatic Pancreatic Adenocarcinoma.

INTRODUCTION: Metastatic pancreatic ductal adenocarcinoma (PDAC) carries a poor prognosis and significant morbidity from local tumor progression. We investigated outcomes among oligometastatic PDAC patients treated with stereotactic magnetic resonance image-guided ablative radiotherapy (SMART) to primary disease. METHODS: We performed a retrospective multi-institutional analysis of oligometastatic PDAC at diagnosis or with metachronous oligoprogression during induction chemotherapy treated with primary tumor SMART. Outcomes of interest included overall survival (OS), progression-free survival (PFS), freedom from locoregional failure (FFLRF), and freedom from distant failure (FFDF). Acute and late toxicity were reported and in exploratory analyses patients were stratified by the number of metastases, SMART indication, and addition of metastasis-directed therapy. RESULTS: From 2019 to 2021, 22 patients with oligometastatic PDAC (range: 1-6 metastases) received SMART to the primary tumor with a median follow-up of 11.2 months from SMART. Nineteen patients had de novo synchronous metastatic disease and three had metachronous oligoprogression. Metastasis location most commonly was liver only (40.9%), multiple organs (27.3%), lungs only (13.6%), or abdominal/pelvic nodes (13.6%). All patients received either FOLFIRINOX (64%) or gemcitabine/nab-paclitaxel (36%) followed by SMART (median 50 Gy, 5 fractions) for local control (77%), pain control (14%), or local progression (9%). Additionally, 41% of patients received other metastasis-directed treatments. The median OS from diagnosis and SMART was 23.9 months and 11.6 months, respectively. Calculated from SMART, the median PFS was 2.4 months with 91% of patients having distant progression, and 1-year local control was 68. Two patients (9%) experienced grade 3 toxicities, gastric outlet obstruction, and gastrointestinal bleed without grade 4 or 5 toxicity. CONCLUSION: There was minimal morbidity of local disease progression after SMART in this cohort of oligometastatic PDAC. As systemic therapy options improve, additional strategies to identify patients who may derive benefits from local consolidation or metastasis-directed therapy are needed.

Dual parallel net: A novel deep learning model for rectal tumor segmentation via CNN and transformer with Gaussian Mixture prior.

Segmentation of rectal cancerous regions from Magnetic Resonance (MR) images can help doctor define the extent of the rectal cancer and judge the severity of rectal cancer, so rectal tumor segmentation is crucial to improve the accuracy of rectal cancer diagnosis. However, accurate segmentation of rectal cancerous regions remains a challenging task due to the shape of rectal tumor has significant variations and the tumor and surrounding tissue are indistinguishable. In addition, in the early research on rectal tumor segmentation, most deep learning methods were based on convolutional neural networks (CNNs), and traditional CNN have small receptive field, which can only capture local information while ignoring the global information of the image. Nevertheless, the global information plays a crucial role in rectal tumor segmentation, so traditional CNN-based methods usually cannot achieve satisfactory segmentation results. In this paper, we propose an encoder-decoder network named Dual Parallel Net (DuPNet), which fuses transformer and classical CNN for capturing both global and local information. Meanwhile, as for capture features at different scales as well as to avoid accuracy loss and parameters reduction, we design a feature adaptive block (FAB) in skip connection between encoder and decoder. Furthermore, in order to utilize the apriori information of rectal tumor shape effectively, we design a Gaussian Mixture prior and embed it in self-attention mechanism of the transformer, leading to robust feature representation and accurate segmentation results. We have performed extensive ablation experiments to verify the effectiveness of our proposed dual parallel encoder, FAB and Gaussian Mixture prior on the dataset from the Shanxi Cancer Hospital. In the experimental comparison with the state-of-the-art methods, our method achieved a Mean Intersection over Union (MIoU) of 89.34% on the test set. In addition to that, we evaluated the generalizability of our method on the dataset from Xinhua Hospital, the promising results verify the superiority of our method.

Classification of brain tumours from MRI images using deep learning-enabled hybrid optimization algorithm.

Brain tumours are produced by the uncontrolled, and unusual tissue growth of brain. Because of the wide range of brain tumour locations, potential shapes, and image intensities, segmentation of the brain tumour by magnetic resonance imaging (MRI) is challenging. In this research, the deep learning (DL)-enabled brain tumour detection is developed by hybrid optimization method. The pre-processing stage used adaptive Wiener filter for minimizing the noise from input image. After that, the abnormal section of the image is segmented using U-Net. Afterwards, the data augmentation is accomplished to recover the random erasing, brightness, and translation characters. The statistical, shape, and texture features are extracted in feature extraction process. In first-level classification, the abnormal section of the image is sensed as brain tumour or not. Here, the Red Deer Tasmanian Devil Optimization (RDTDO) trained DenseNet is hired for brain tumour detection process. If tumour is identified, then second-level classification provides the brain tumour classification, where deep residual network (DRN)-enabled RDTDO is employed. Furthermore, the system performance is assessed by accuracy, true positive rate (TPR), true negative rate (TNR), positive predictive value (PPV), and negative predictive value (NPV) with the maximum values of 0.947, 0.926, 0.950, 0.937, and 0.926 are attained.

Application of magnetic resonance image compilation (MAGiC) in the diagnosis of middle-aged and elderly women with osteoporosis.

OBJECTIVE: To investigate the feasibility of diagnosing osteoporosis (OP) in women through magnetic resonance image compilation (MAGiC). METHODS: A total of 110 patients who underwent lumbar magnetic resonance imaging and dual X-ray absorptiometry examinations were collected and divided into two groups according bone mineral density: osteoporotic group (OP) and non-osteoporotic group (non-OP). The variation trends of T1 (longitudinal relaxation time), T2 (transverse relaxation time) and BMD (bone mineral density) with the increase of age, and the correlation of T1 and T2 with BMD were examined by establishing a clinical mathematical model. RESULTS: With the increase of age, BMD and T1 value decreased gradually, while T2 value increased. T1 and T2 had statistical significance in diagnosing OP (P < 0.001), and there is moderate positive correlation between T1 and BMD values (R = 0.636, P < 0.001), while moderate negative correlation between T2 and BMD values (R=-0.694, P < 0.001). Receiver characteristic curve test showed that T1 and T2 had high accuracy in diagnosing OP (T1 AUC = 0.982, T2 AUC = 0.978), and the critical values of T1 and T2 for evaluating osteoporosis were 0.625s and 0.095s, respectively. Besides, the combined utilization of T1 and T2 had higher diagnostic efficiency (AUC = 0.985). Combined T1 and T2 had higher diagnostic efficiency (AUC = 0.985). Function fitting results of OP group: BMD=-0.0037* age - 0.0015*T1 + 0.0037*T2 + 0.86, sum of squared error (SSE) = 0.0392, and non-OP group: BMD = 0.0024* age - 0.0071*T1 + 0.0007*T2 + 1.41, SSE = 0.1007. CONCLUSION: T1 and T2 value of MAGiC have high efficiency in diagnosing OP by establishing a function fitting formula of BMD with T1, T2 and age.

Damage and recovery of the intrinsic and extrinsic foot muscles from running a full marathon.

PURPOSE: To investigate the effects of full marathon running on intrinsic and extrinsic foot muscle damage and to determine the relationship with the height change of the longitudinal foot arch following full marathon completion. METHODS: Magnetic resonance imaging-measured transverse relaxation time (T2 ) of the abductor hallucis (ABH), flexor digitorum brevis (FDB) and quadratus plantae (QP), flexor digitorum longus (FDL), tibialis posterior (TP), and flexor hallucis longus (FHL) from 22 collegiate runners were assessed before and 1, 3, and 8 days after full marathon running. The three-dimensional foot posture of 10 of the 22 runners was further obtained using a foot scanner system before and 1, 3, and 8 days after the marathon. RESULTS: Marathon-induced increases in T2 were observed in the QP, FDL, TP, and FHL 1 day after the marathon (+7.5%, +4.7%, +6.7%, and +5.9%, respectively), with the increased T2 of TP persisting until 3 days after the marathon (+4.6%). T2 changes of FDL and FHL from pre-marathon to DAY 1 showed direct correlations with the corresponding change in the arch height ratio (r = 0.823, p = 0.003, and r = 0.658, p = 0.038). CONCLUSION: The damage and recovery response from a full marathon differed among muscles; QP, FDL, TP, and FHL increased T2 after the marathon, whereas ABH and FDB did not. In addition, T2 changes in FDL and FHL and the arch height ratio change were correlated. Our results suggest that the extrinsic foot muscles could be more susceptible to damage than the intrinsic during marathon running.

Differential diagnosis of atypical and anaplastic meningiomas based on conventional MRI features and ADC histogram parameters using a logistic regression model nomogram.

The purpose of the study was to determine the value of a logistic regression model nomogram based on conventional magnetic resonance imaging (MRI) features and apparent diffusion coefficient (ADC) histogram parameters in differentiating atypical meningioma (AtM) from anaplastic meningioma (AnM). Clinical and imaging data of 34 AtM and 21 AnM diagnosed by histopathology were retrospectively analyzed. The whole tumor delineation along the tumor edge on ADC images and ADC histogram parameters were automatically generated and comparisons between the two groups using the independent samples t test or Mann-Whitney U test. Univariate and multivariate logistic regression analyses were used to construct the nomogram of the AtM and AnM prediction model, and the model's predictive efficacy was evaluated using calibration and decision curves. Significant differences in the mean, enhancement, perc.01%, and edema were noted between the AtM and AnM groups (P < 0.05). Age, sex, location, necrosis, shape, max-D, variance, skewness, kurtosis, perc.10%, perc.50%, perc.90%, and perc.99% exhibited no significant differences (P > 0.05). The mean and enhancement were independent risk factors for distinguishing AtM from AnM. The area under the curve, accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the nomogram were 0.871 (0.753-0.946), 80.0%, 81.0%, 79.4%, 70.8%, and 87.1%, respectively. The calibration curve demonstrated that the model's probability to predict AtM and AnM was in favorable agreement with the actual probability, and the decision curve revealed that the prediction model possessed satisfactory clinical availability. A logistic regression model nomogram based on conventional MRI features and ADC histogram parameters is potentially useful as an auxiliary tool for the preoperative differential diagnosis of AtM and AnM.