Magnetic Resonance Imaging Interpretation Research Articles

Prostate magnetic resonance imaging (MRI) in patients with hip implants-presetting a protocol using a phantom.

Metal structures are a source of artifacts that significantly complicate the interpretation of magnetic resonance imaging (MRI). The use of prostate MRI as a preliminary test in men with a suspicion on prostate cancer leads to an increased use of the test. The aim of this study was to solve a clinically significant problem: to ensure the reduction of artifacts from metal hip implants during prostate MRI. Another goal was to evaluate the impact of artifact reduction methods on quantitative measurements. The prostate gland (PG) phantom model was a cylinder filled with an aqueous solution of polyvinylpyrrolidone at the concentrations of 40%, 30%, and 20% [central zone (CZ), peripheral zone (PZ), and "lesion", respectively]. Phantom MRI study was conducted on Philips Ingenia 1.5T and Philips Ingenia 3T scanners. For 1.5 T, the reduction in the influence of artifacts inside region of interest (ROI) was observed, expressed in a decrease in the average apparent diffusion coefficient (ADC) (CZ, PZ, "lesion") for the manual artifact reduction (MAR) and ZOOM (title of software artifact reduction) techniques compared to the standard method. For 3T this effect was not detected. The same ADC results were obtained for Standard and MAR techniques, and increased ADC values for ZOOM. Despite the fact that the spread of ADC values on 3.0T scanners was minimal, there was a significant deviation of ADC values from the reference ones (up to 30.4%). Therefore, it is necessary to use a correction coefficient in the ADC calculation for the 3.0 T device. In the presented clinical case, high-quality tomograms were obtained without any artifacts, despite the presence of two hip replacement devices in the scanning area. The accurate prostate MRI in the presence of implants is essential for an accurate diagnosis. This approach allows to reduce artifacts from hip implants, to visualize PG and periprostatic tissue in the best way, and to detect malignant and benign changes.

The "Ins and Outs" of Dynamic Magnetic Resonance Imaging for Female Pelvic Organ Prolapse.

Concurrent pelvic organ and rectal prolapse have an incidence of 38%. Dynamic pelvic magnetic resonance imaging (MRI) is the modality of choice for workup. We discuss dynamic pelvic MRI indications, interpretation, and clinical application to pelvic floor disorders. The pubococcygeal line (PCL) extends from the pubic symphysis to the last coccygeal joint. The "H line" demonstrates the levator hiatus size, drawn from the inferior pubic symphysis to the posterior rectal wall at the anorectal junction. The "M line" represents vertical descent of the levator hiatus and extends perpendicularly from the PCL to the posterior aspect of the H line. With rectovaginal fascial defects, the small bowel, the peritoneum, and the sigmoid colon can prolapse. Posterior compartment abnormalities include rectocele, rectal prolapse, and descending perineal syndrome. Pelvic MRI can evaluate functional disorders such as anismus, where the anorectal angle is narrowed and associated with lack of pelvic floor descent and incomplete evacuation. Particularly for patients with concurrent urogynecological and colorectal complaints, previous pelvic reconstructive surgery, or when clinical symptomatology does not correlate with physical examination, dynamic pelvic MRI can impact management. It is critical for pelvic reconstructive surgeons to be familiar with this imaging modality to counsel patients and interpret radiographic findings.

Artificial intelligence improves the diagnosis of human leukocyte antigen (HLA)-B27-negative axial spondyloarthritis based on multi-sequence magnetic resonance imaging and clinical features.

Axial spondyloarthritis (axSpA) is frequently diagnosed late, particularly in human leukocyte antigen (HLA)-B27-negative patients, resulting in a missed opportunity for optimal treatment. This study aimed to develop an artificial intelligence (AI) tool, termed NegSpA-AI, using sacroiliac joint (SIJ) magnetic resonance imaging (MRI) and clinical SpA features to improve the diagnosis of axSpA in HLA-B27-negative patients. We retrospectively included 454 HLA-B27-negative patients with rheumatologist-diagnosed axSpA or other diseases (non-axSpA) from the Third Affiliated Hospital of Southern Medical University and Nanhai Hospital between January 2010 and August 2021. They were divided into a training set (n=328) for 5-fold cross-validation, an internal test set (n=72), and an independent external test set (n=54). To construct a prospective test set, we further enrolled 87 patients between September 2021 and August 2023 from the Third Affiliated Hospital of Southern Medical University. MRI techniques employed included T1-weighted (T1W), T2-weighted (T2W), and fat-suppressed (FS) sequences. We developed NegSpA-AI using a deep learning (DL) network to differentiate between axSpA and non-axSpA at admission. Furthermore, we conducted a reader study involving 4 radiologists and 2 rheumatologists to evaluate and compare the performance of independent and AI-assisted clinicians. NegSpA-AI demonstrated superior performance compared to the independent junior rheumatologist (≤5 years of experience), achieving areas under the curve (AUCs) of 0.878 [95% confidence interval (CI): 0.786-0.971], 0.870 (95% CI: 0.771-0.970), and 0.815 (95% CI: 0.714-0.915) on the internal, external, and prospective test sets, respectively. The assistance of NegSpA-AI promoted discriminating accuracy, sensitivity, and specificity of independent junior radiologists by 7.4-11.5%, 1.0-13.3%, and 7.4-20.6% across the 3 test sets (all P<0.05). On the prospective test set, AI assistance also improved the diagnostic accuracy, sensitivity, and specificity of independent junior rheumatologists by 7.7%, 7.7%, and 6.9%, respectively (all P<0.01). The proposed NegSpA-AI effectively improves radiologists' interpretations of SIJ MRI and rheumatologists' diagnoses of HLA-B27-negative axSpA.

PI-RADS 2.1: A Practical Overview

AbstractThe Prostate Imaging Reporting and Data System (PI-RADS) is an essential tool for standardizing the interpretation of multiparametric magnetic resonance imaging (mp-MRI) of the prostate for detecting prostate cancer. This article provides a comprehensive overview of the latest version, PI-RADS 2.1, with clear pictures to guide radiologists in its practical application for improved diagnostic accuracy. The article explores the key modifications, emphasizing the changes in scoring criteria and their impact on clinical decision-making. It discusses the importance of mp-MRI sequences, such as T2-weighted imaging, diffusion-weighted imaging, and dynamic contrast-enhanced imaging, and clarifies their significance in the PI-RADS 2.1 framework. The article highlights practical insights to help radiologists integrate this updated system into their day-to-day practice, promoting consistency and reliability in reporting.

IMG-28. AUTOMATIC BRAIN TUMOR VOLUMETRIC ANALYSIS IN MAGNETIC RESONANCE IMAGING GENERALIZABLE TO PEDIATRIC NEURO-ONCOLOGY

Abstract BACKGROUND The prognosis of brain tumors is variable in clinical practice if it only relies on human interpretation of magnetic resonance imaging (MRI). The automatic segmentation of brain tumors in MRI enables quantitative analysis in support of clinical trials and personalized patient care. We developed benchmarked deep learning-based tools that are generalizable to the volumetric quantification of various tumor types across diverse populations. METHODS We participated in the well-established international brain tumor segmentation challenge (BraTS 2023) and benchmarking competition. The challenge made available 4,500 multi-national brain tumor cases with multi-sequence MRIs, including pediatric high-grade gliomas (PED), i.e., high-grade astrocytoma and diffuse midline glioma, and adult gliomas, brain metastases (MET) and intracranial meningiomas (MEN). Each case comprises four MRI volumes: T1, contrast-enhanced T1, T2, and T2-FLAIR. Manual segmentations were provided to establish ground truth for enhancing tumor (ET), tumor core (TC), and whole tumor (WT). Our framework used a model ensemble strategy based on two state-of-the-art deep learning models: a convolutional neural network (nnU-Net) and a vision transformer (Swin UNETR) and was tested for broader applicability across multiple tumor types. The framework was trained on 99, 1,000, and 165 cases and validated on 45, 141, and 31 unseen cases for PED, MEN, and MET, respectively. Automatic segmentations were evaluated by lesion-wise volume overlap (Dice similarity score, DSC) and Hausdorff distance (HD). RESULTS In the evaluation on independent unseen test datasets, our automatic tool was ranked first for PED, third for MEN, and fourth for MET volumetric analysis. Our method resulted in PED lesion-wise DSC of 0.733, 0.782, 0.817 and HD (mm) of 75.93, 25.54, 24.18 for ET, TC, and WT, respectively. CONCLUSIONS These brain tumor volumetric analysis tools are readily available to be efficiently tested on diverse datasets. Automatic MRI analysis provides consistent quantitative data for multi-institutional protocols and clinical trials.

Improving Brain Tumor Classification: An Approach Integrating Pre-Trained CNN Models and Machine Learning Algorithms

Accurate detection of brain tumors is crucial for enhancing patient outcomes, yet the interpretation of Magnetic Resonance Imaging (MRI) scans poses significant challenges. This study introduces a novel approach to brain tumor classification by exploring three pre-trained convolutional neural network (CNN) models: DenseNet201, EfficientNetB5, and InceptionResNetV2, combined with softmax activation for feature extraction. These features are then subjected to Principal Component Analysis (PCA) for dimensionality reduction. Subsequently, three machine learning models—Support Vector Machine (SVM), Multi-layer Perceptron (MLP), and Gaussian Naive Bayes (GNB)—are employed for classification. The results reveal that DenseNet201, when combined with SVM and MLP, outperforms the other models in terms of accuracy, recall, and precision. Specifically, DenseNet201 achieves 100% accuracy, recall, and precision on Dataset-I and 98% accuracy, recall, and precision on Dataset-II when paired with SVM and MLP. This study provides valuable insights into the interplay between CNN models, feature extraction techniques, and machine learning algorithms for brain tumor classification, highlighting the efficacy of DenseNet201 combined with SVM and MLP.

Effect of Bacillus Calmette–Guerin Exposure on Prostate Cancer Detection Using Magnetic Resonance Imaging: A Cohort Study

BackgroundGranulomatous prostatitis is a medical condition that may mimic prostate cancer. PurposeGranulomatous prostatitis resulting from BCG-exposure can confound the diagnosis of prostate cancer based on prostate imaging and data system (PI-RADS) classification observed on multiparametric prostate magnetic resonance imaging (mpMRI). Study type, Population, Assessment and Statistical TestsA cohort study was conducted, enrolling consecutive males at risk for prostate cancer who underwent an mpMRI-targeted prostate biopsy between February 2016 and August 2023. The focus of the study was on prior BCG-exposure as adjuvant treatment for non–muscle-invasive urothelial carcinoma within the 3 years prior the magnetic resonance imaging (MRI). Exclusion criteria were a prior androgen deprivation therapy, prostate surgery or radiation, and BCG-exposure occurring more than 3 years and less than 3 months before the MRI. Chi-square, logistic-regression, statistical association, and homogeneity tests were used. ResultsTotal 712 patients, 899 biopsied lesions (218 PI-RADS 3, 521 PI-RADS 4 and 160 PI-RADS 5) and 20 patients with 30 lesions within the BCG-exposed cohort. Chi-square and logistic-regression tests showed an association between PI-RADS with malignancy and significant tumor (ST), considering PI-RADS3 as the reference (OR: 4.9 [95% CI, 3.4-7.1] for PI-RADS4 and OR: 21.7 [95% CI, 12.4-37.8] for PI-RADS5 for malignancy, and OR: 5.3 [95% CI, 3.2-8.7] for PI-RADS4 and OR: 16.5 [95% CI, 9.4-28.9] for PI-RADS5 regarding ST). A statistically significant negative association was demonstrated between malignancy and ST with respect to BCG-exposure (OR: 0.15 [95% CI, 0.06-0.39] and OR: 0.39 [95% CI, 0.15-1.0], respectively). Statistically significant risk-difference for malignancy in patients nonexposed to BCG regarding those exposed was 45% (61.6% vs. 16.7%) for PI-RADS4, and 68.5% (90.7% vs. 22.2%) and 42.7% (64.9% vs. 22.2%) concerning malignancy and ST for PI-RADS5, respectively. Data ConclusionsGranulomatous prostate reaction caused by BCG-exposure acts as confounding factor for prostate MRI interpretation. The risk of malignancy and significant tumor on targeted biopsy to PI-RADS 3, 4 and 5 is notably lower in exposed patients.

P.109 Machine learning based approach to improving the prediction of neurological deterioration in mild Degenerative Cervical Myelopathy

Background: Degenerative cervical myelopathy (DCM) is the most common form of atraumatic spinal cord injury globally, yet clinical guidelines remain unclear on surgical recommendations for patients with mild forms of DCM. This is in part due to limitations in current MR imaging interpretation and complex mechanisms of neurological deterioration. Supervised machine learning (ML) models can help to identify clinical and imaging indicators of deterioration within mild DCM patients. Methods: 127 MRI scans (T2w, Diffusion Tensor Imaging, and Magnetization transfer scans) accompanied by a series of clinical tests underwent a semi-automated analysis to derive quantitative metrics. Random forest classifier, Support Vector Machine, and Logistic Regression models were trained and tested to predict 6-month neurological deterioration within patients. Results: The ML models performed, on average, better than previous studies with a balanced accuracy ranging between 70-75%. “Advanced” imaging metrics such as diffusion tensor imaging and magnetization transfer scans played an important role in improving model accuracy but only when used near the maximally compressed disc level, suggesting that limited yet targetted imaging metrics support ML model performance. Conclusions: The inclusion of specific, targeted imaging and clinical metrics support ML model performance in predicting neurological deterioration within mild DCM patients.

Comparing MRI and arthroscopic appearances of common knee pathologies: A pictorial review.

Knee pathology, including anterior cruciate ligament (ACL) tears, meniscal tears, articular cartilage lesions, and intra-articular masses or cysts are common clinical entities treated by orthopedic surgeons with arthroscopic surgery. Preoperatively, magnetic resonance imaging (MRI) is now standard in confirming knee pathology, particularly detecting pathology less evident with history and physical examination alone. The radiologist's MRI interpretation becomes essential in evaluating intra-articular knee structures. Typically, the radiologist that interprets the MRI does not have the opportunity to view the same pathology arthroscopically. Thus, the purpose of this article is to illustratively reconcile what the orthopedic surgeon sees arthroscopically with what the radiologist sees on magnetic resonance imaging when viewing the same pathology. Correlating virtual and actual images can help better understand pathology, resulting in more accurate MRI interpretations. In this article, we present and review a series of MR and correlating arthroscopic images of ACL tears, meniscal tears, chondral lesions, and intra-articular masses and cysts. Short teaching points are included to highlight the importance of radiological signs and pathological MRI appearance with significant clinical and arthroscopic findings.

Deep learning-based platform performs high detection sensitivity of intracranial aneurysms in 3D brain TOF-MRA: An external clinical validation study

PurposeTo evaluate the diagnostic efficacy of a developed artificial intelligence (AI) platform incorporating deep learning algorithms for the automated detection of intracranial aneurysms in time-of-flight (TOF) magnetic resonance angiography (MRA). MethodThis retrospective study encompassed 3D TOF MRA images acquired between January 2023 and June 2023, aiming to validate the presence of intracranial aneurysms via our developed AI platform. The manual segmentation results by experienced neuroradiologists served as the “gold standard”. Following annotation of MRA images by neuroradiologists using InferScholar software, the AI platform conducted automatic segmentation of intracranial aneurysms. Various metrics including accuracy (ACC), balanced ACC, area under the curve (AUC), sensitivity (SE), specificity (SP), F1 score, Brier Score, and Net Benefit were utilized to evaluate the generalization of AI platform. Comparison of intracranial aneurysm identification performance was conducted between the AI platform and six radiologists with experience ranging from 3 to 12 years in interpreting MR images. Additionally, a comparative analysis was carried out between radiologists’ detection performance based on independent visual diagnosis and AI-assisted diagnosis. Subgroup analyses were also performed based on the size and location of the aneurysms to explore factors impacting aneurysm detectability. Results510 patients were enrolled including 215 patients (42.16 %) with intracranial aneurysms and 295 patients (57.84 %) without aneurysms. Compared with six radiologists, the AI platform showed competitive discrimination power (AUC, 0.96), acceptable calibration (Brier Score loss, 0.08), and clinical utility (Net Benefit, 86.96 %). The AI platform demonstrated superior performance in detecting aneurysms with an overall SE, SP, ACC, balanced ACC, and F1 score of 91.63 %, 92.20 %, 91.96 %, 91.92 %, and 90.57 % respectively, outperforming the detectability of the two resident radiologists. For subgroup analysis based on aneurysm size and location, we observed that the SE of the AI platform for identifying tiny (diameter＜3mm), small (3 mm ≤ diameter＜5mm), medium (5 mm ≤ diameter＜7mm) and large aneurysms (diameter ≥ 7 mm) was 87.80 %, 93.14 %, 95.45 %, and 100 %, respectively. Furthermore, the SE for detecting aneurysms in the anterior circulation was higher than that in the posterior circulation. Utilizing the AI assistance, six radiologists (i.e., two residents, two attendings and two professors) achieved statistically significant improvements in mean SE (residents: 71.40 % vs. 88.37 %; attendings: 82.79 % vs. 93.26 %; professors: 90.07 % vs. 97.44 %; P < 0.05) and ACC (residents: 85.29 % vs. 94.12 %; attendings: 91.76 % vs. 97.06 %; professors: 95.29 % vs. 98.82 %; P < 0.05) while no statistically significant change was observed in SP. Overall, radiologists’ mean SE increased by 11.40 %, mean SP increased by 1.86 %, and mean ACC increased by 5.88 %, mean balanced ACC promoted by 6.63 %, mean F1 score grew by 7.89 %, and Net Benefit rose by 12.52 %, with a concurrent decrease in mean Brier score declined by 0.06. ConclusionsThe deep learning algorithms implemented in the AI platform effectively detected intracranial aneurysms on TOF-MRA and notably enhanced the diagnostic capabilities of radiologists. This indicates that the AI-based auxiliary diagnosis model can provide dependable and precise prediction to improve the diagnostic capacity of radiologists.

A pictorial essay of PI-RADS pearls and pitfalls: toward less ambiguity and better practice.

Prostate Imaging Reporting and Data System (PI-RADS) was designed to standardize the interpretation of multiparametric magnetic resonance imaging (MRI) of the prostate, aiding in assessing the probability of clinically significant prostate cancer. By providing a structured scoring system, it enables better risk stratification, guiding decisions regarding the need for biopsy and subsequent treatment options. In this article, we explore both the strengths and weaknesses of PI-RADS, offering insights into its updated diagnostic performance and clinical applications, while also addressing potential pitfalls using diverse, representative MRI cases.

Predictive factors of disc perforation of the temporomandibular joint (TMJ)

Magnetic resonance imaging (MRI) is the gold standard for diagnosing internal derangement, but its accuracy in detecting disc perforations varies. This cohort study included 92 patients who underwent arthroscopic surgery to identify associated variables. The presence of arthroscopically verified disc perforation served as the primary predictor variable, while the radiologist's diagnosis of disc perforation served as the primary outcome variable. Demographic data, patient signs, and symptoms were considered as covariates. Statistical tests were performed with a significance level of 5%. Logistic regression models were used to predict disc perforation, and diagnostic accuracy was assessed using ROC curves. Among 155 joints analyzed (92 patients: 135 females, 20 male), radiologists' MRI interpretations did not significantly correlate with verified disc perforation (p&lt;0.218), showing 14.3% sensitivity and 92.5% specificity. Age, symptomatic onset, joint noise type, osteoarthritis diagnosis, altered disc shape, and Wilkes classification showed significant associations (p&lt;0.05). A logistic regression model used associated variables for disc perforation. Risk factors: symptom onset time, joint crepitus, osteoarthrosis diagnosis by MRI; joint click as protective. The model demonstrated high diagnostic accuracy (AUC of 0.836, 95% CI_0.76–0.91). Radiologists' MRI interpretation was unreliable, but our model accurately predicted disc perforation. Identifying these factors could guide surgical decisions.

AI Applications to Breast MRI: Today and Tomorrow.

In breast imaging, there is an unrelenting increase in the demand for breast imaging services, partly explained by continuous expanding imaging indications in breast diagnosis and treatment. As the human workforce providing these services is not growing at the same rate, the implementation of artificial intelligence (AI) in breast imaging has gained significant momentum to maximize workflow efficiency and increase productivity while concurrently improving diagnostic accuracy and patient outcomes. Thus far, the implementation of AI in breast imaging is at the most advanced stage with mammography and digital breast tomosynthesis techniques, followed by ultrasound, whereas the implementation of AI in breast magnetic resonance imaging (MRI) is not moving along as rapidly due to the complexity of MRI examinations and fewer available dataset. Nevertheless, there is persisting interest in AI-enhanced breast MRI applications, even as the use of and indications of breast MRI continue to expand. This review presents an overview of the basic concepts of AI imaging analysis and subsequently reviews the use cases for AI-enhanced MRI interpretation, that is, breast MRI triaging and lesion detection, lesion classification, prediction of treatment response, risk assessment, and image quality. Finally, it provides an outlook on the barriers and facilitators for the adoption of AI in breast MRI. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 6.

Fixation technique of biodegradable magnesium alloy suture anchor in the rotator cuff repair of the shoulder in a goat model: a technical note

BackgroundShoulder disorders, particularly rotator cuff tears, are prevalent musculoskeletal conditions related to aging. Although the widely used suture anchor technique provides strong mechanical support to the tendon, it is associated with a risk of postoperative tendon retearing. The conventionally used titanium alloys can affect the interpretation of magnetic resonance imaging. Degradable magnesium alloys possess excellent biocompatibility, similar mechanical property to the bone, and stimulating bone formation ability from Mg2+. The purpose of this experiment was to develop innovative magnesium-based suture anchors to enhance rotator cuff repair by improving fixation materials, and to evaluate their feasibility in a goat model.MethodsWe developed fluoridized ZK60 suture anchors as the implantation material for two goats, who underwent rotator cuff repair surgery on both shoulders. Computed tomography (CT) and histological analysis were performed at 12 weeks postoperatively, and the results were compared between the magnesium and titanium alloy groups. Additionally, a hematological examination was conducted, which included assessments of red blood cells, white blood cells, platelets, coagulation function, liver function, kidney function, and magnesium ion concentration.ResultsThe 12-week postoperative CT images showed intact MgF2 ZK60 suture anchors, effectively reconnecting the infraspinatus tendon to the humeral head. The anchors became less visible on CT scans, indicating absorption by surrounding tissues. New bone formation in the MgF2 group surpassed that in the Ti group, demonstrating superior osseointegration. The similarity between cortical bone and magnesium reduced stress-shielding and promoted bone regeneration. Histological analysis revealed successful tendon healing with MgF2 anchors, while the Ti group showed discontinuous interfaces and reduced collagen secretion. Hematological examination showed stable liver, renal function, and magnesium ion levels.ConclusionsThe findings indicate that MgF2-coated suture anchors are feasible for rotator cuff repair and potentially other orthopedic applications. We hope that magnesium alloy anchors can become the solution for rotator cuff tendon repair surgery.

Can an in-person hands-on applicator-based teaching session improve trainee knowledge and comfort with complex gynecologic brachytherapy?

INTRODUCTIONWith the emergence of imaged-based planning and hybrid applicators the complexity of gynecologic brachytherapy has dramatically increased. Despite the known advantages of brachytherapy, notable national declines in utilization of brachytherapy have been documented. Clearly improved education in the sphere of gynecologic brachytherapy is needed. We hypothesize that a hands-on applicator-based training session would improve trainee comfort with gynecologic brachytherapy. METHODS AND MATERIALSAn in-person, applicator-based, hands-on training session was held with trainees from both radiation and gynecologic oncology programs. Trainees practiced assembling and handling applicators while receiving instruction on clinical scenarios in which various applicators are used in gynecologic cancer brachytherapy. Pre- and post-session, participants were administered an objective test of 10 pictorial-based case vignettes to quantify ability to select the correct applicator based on the interpretation of T2-weighted MR images. Participants additionally received a subjective survey to quantify comfort and experience with gynecologic brachytherapy using Likert-type question formatting. RESULTSA total of 14 trainees participated. Most common case volume experience was 0–10 intracavitary (57%), 0–10 hybrid (71%), and 0–10 interstitial (71%). Pre-session, the most common answer to comfort level was “not comfortable still learning” for all brachytherapy types, and most common answer to largest gap in knowledge was all facets of brachytherapy. Average case-based test score was 3.5/10 pre-session versus 5.3/10 post-session (p = 0.028). Post-session, all respondents reported improved comfort level with brachytherapy. Post-session, most common answer to largest gap in knowledge was applicator/patient selection, and applicator/patient selection was also the largest area of identified improvement. 100% of participants felt repeating the session in the future would be helpful. CONCLUSIONSHands-on training with applicators improves both subjective and objective comfort with gynecologic brachytherapy. With 100% of participants requesting to implement this session into resident training, we suggest national opportunities might exist to expand educational processes and improve utilization of complex gynecologic brachytherapy in practice.

Convolutional neural network-based magnetic resonance image differentiation of filum terminale ependymomas from schwannomas

PurposePreoperative diagnosis of filum terminale ependymomas (FTEs) versus schwannomas is difficult but essential for surgical planning and prognostic assessment. With the advancement of deep-learning approaches based on convolutional neural networks (CNNs), the aim of this study was to determine whether CNN-based interpretation of magnetic resonance (MR) images of these two tumours could be achieved.MethodsContrast-enhanced MRI data from 50 patients with primary FTE and 50 schwannomas in the lumbosacral spinal canal were retrospectively collected and used as training and internal validation datasets. The diagnostic accuracy of MRI was determined by consistency with postoperative histopathological examination. T1-weighted (T1-WI), T2-weighted (T2-WI) and contrast-enhanced T1-weighted (CE-T1) MR images of the sagittal plane containing the tumour mass were selected for analysis. For each sequence, patient MRI data were randomly allocated to 5 groups that further underwent fivefold cross-validation to evaluate the diagnostic efficacy of the CNN models. An additional 34 pairs of cases were used as an external test dataset to validate the CNN classifiers.ResultsAfter comparing multiple backbone CNN models, we developed a diagnostic system using Inception-v3. In the external test dataset, the per-examination combined sensitivities were 0.78 (0.71–0.84, 95% CI) based on T1-weighted images, 0.79 (0.72–0.84, 95% CI) for T2-weighted images, 0.88 (0.83–0.92, 95% CI) for CE-T1 images, and 0.88 (0.83–0.92, 95% CI) for all weighted images. The combined specificities were 0.72 based on T1-WI (0.66–0.78, 95% CI), 0.84 (0.78–0.89, 95% CI) based on T2-WI, 0.74 (0.67–0.80, 95% CI) for CE-T1, and 0.81 (0.76–0.86, 95% CI) for all weighted images. After all three MRI modalities were merged, the receiver operating characteristic (ROC) curve was calculated, and the area under the curve (AUC) was 0.93, with an accuracy of 0.87.ConclusionsCNN based MRI analysis has the potential to accurately differentiate ependymomas from schwannomas in the lumbar segment.

Axial Spondyloarthritis: Does Magnetic Resonance Imaging Classification Improve Report Interpretation.

The interpretation of magnetic resonance imaging (MRI) reports is crucial for the diagnosis of axial spondyloarthritis, but the subjective nature of narrative reports can lead to varying interpretations. This study presents a validation of a novel MRI reporting system for the sacroiliac joint in clinical practice. A historical review was conducted on 130 consecutive patients referred by 2 rheumatologists for initial MRI assessment of possible axial spondyloarthritis. The original MRI reports were interpreted by the rheumatologists and the radiologist who originally read the images and then categorized according to the novel system. Two musculoskeletal radiologists then reinterpreted the original MRI scans using the new system, and the resulting reports were interpreted and categorized by the same rheumatologists. The quality of the new framework was assessed by comparing the interpretations of both reports. Ninety-two patients met the study criteria. The rheumatologists disagreed on the categorization of the original MRI reports in 12% of cases. The rheumatologists and original radiologists disagreed on the categorization of the initial report in 23.4% of cases. In contrast, there was 100% agreement between the rheumatologists and radiologists on the categorization of the new MRI report. The new MRI categorization system significantly improved the agreement between the clinician and radiologist in report interpretation. The system provided a standard vocabulary for reporting, reduced variability in report interpretation, and may therefore improve clinical decision-making.

Arterial Spin-Labeling and DSC Perfusion Metrics Improve Agreement in Neuroradiologists' Clinical Interpretations of Posttreatment High-Grade Glioma Surveillance MR Imaging-An Institutional Experience.

MR perfusion has shown value in the evaluation of posttreatment high-grade gliomas, but few studies have shown its impact on the consistency and confidence of neuroradiologists' interpretation in routine clinical practice. We evaluated the impact of adding MR perfusion metrics to conventional contrast-enhanced MR imaging in posttreatment high-grade glioma surveillance imaging. This retrospective study included 45 adults with high-grade gliomas who had posttreatment perfusion MR imaging. Four neuroradiologists assigned Brain Tumor Reporting and Data System scores for each examination on the basis of the interpretation of contrast-enhanced MR imaging and then after the addition of arterial spin-labeling-CBF, DSC-relative CBV, and DSC-fractional tumor burden. Interrater agreement and rater agreement with a multidisciplinary consensus group were assessed with κ statistics. Raters used a 5-point Likert scale to report confidence scores. The frequency of clinically meaningful score changes resulting from the addition of each perfusion metric was determined. Interrater agreement was moderate for contrast-enhanced MR imaging alone (κ = 0.63) and higher with perfusion metrics (arterial spin-labeling-CBF, κ = 0.67; DSC-relative CBV, κ = 0.66; DSC-fractional tumor burden, κ = 0.70). Agreement between raters and consensus was highest with DSC-fractional tumor burden (κ = 0.66-0.80). Confidence scores were highest with DSC-fractional tumor burden. Across all raters, the addition of perfusion resulted in clinically meaningful interpretation changes in 2%-20% of patients compared with contrast-enhanced MR imaging alone. Adding perfusion to contrast-enhanced MR imaging improved interrater agreement, rater agreement with consensus, and rater confidence in the interpretation of posttreatment high-grade glioma MR imaging, with the highest agreement and confidence scores seen with DSC-fractional tumor burden. Perfusion MR imaging also resulted in interpretation changes that could change therapeutic management in up to 20% of patients.

Mucinous Rectal Adenocarcinoma-Challenges in Magnetic Resonance Imaging Interpretation.

Mucinous rectal cancer (MRC) is defined by the World Health Organization as an adenocarcinoma with greater than 50% mucin content. Classic teaching suggests that it carries a poorer prognosis than conventional rectal adenocarcinoma. This poorer prognosis is thought to be related to mucin dissecting through tissue planes at a higher rate, thus increasing the stage of disease at presentation. Developments in immunotherapy have bridged much of this prognostic gap in recent years. Magnetic resonance imaging is the leading modality in assessing the locoregional spread of rectal cancer. Mucinous rectal cancer carries unique imaging challenges when using this modality. Much of the difficulty lies in the inherent increased T2-weighted signal of mucin on magnetic resonance imaging. This creates difficulty in differentiating mucin from the adjacent background fat, making the detection of both the primary disease process as well as the locoregional spread challenging. Computed tomography scan can act as a valuable companion modality as mucin tends to be more apparent in the background fat. After therapy, diagnostic challenges remain. Mucin is frequently present, and distinguishing cellular from acellular mucin can be difficult. In this article, we will discuss each of these challenges and present examples of such situations and strategies that can be used to overcome them.

Food for Thought: A Review of Neuroradiographic Signs Inspired by Food

Whetting appetites for both knowledge and lunch, there is a rich history of naming signs in radiology after delectable treats. Built to entice even the most discerning palate, this collection of neuroradiologic signs is presented here to prove that learning neuroradiology can be as sweet as it is stimulating. Imaging signs are important to the practice of neuroradiology, a medical subspecialty whose work can influence the decisions of other clinicians. They can assist the physician by associating complex pathologic processes with common, recognizable images from everyday life. Signs may aid the neuroradiologist in the interpretation of MR imaging and other imaging studies by adding confidence to a diagnosis. Sugary treats apparently offer the most inspiration, causing radiology reads to make the ordering physician’s stomach grumble.Learning Objective: To recognize 9 food-inspired radiologic signs and identify the pathology behind them