Search

BACKGROUND: Meniscal root tears can lead to early knee osteoarthritis and pain. This study aimed (1) to compare clinical and radiological outcomes between patients who underwent arthroscopic meniscal root repair after meniscal root tears and those who received non-surgical treatment, and (2) to identify whether baseline MRI findings could be potential predictors for future treatment strategies. METHODS: Patients with meniscal root tears were identified from our picture archiving and communication system from 2016 to 2020. Two radiologists reviewed radiographs and MRI studies using Kellgren-Lawrence (KL) grading and a modified Whole Organ MRI Scoring (WORMS) at baseline and follow-up. The median (interquartile range [IQR]) of follow-up radiographs and MRI studies were 134 (44-443) days and 502 (260-1176) days, respectively. MR images were assessed for root tear-related findings. Pain scores using visual analogue scale (VAS) and management strategies (non-surgical vs. arthroscopic root repair) were also collected. Chi-squared tests and independent t-tests were used to assess differences regarding clinical and imaging variables between treatment groups. Logistic regression analyses were performed to evaluate the associations between baseline MRI findings and each future treatment. RESULTS: Ninety patients were included. VAS pain scores were significantly (p

STUDY DESIGN: In vivo retrospective study of fully automatic quantitative imaging feature extraction from clinically acquired lumbar spine magnetic resonance imaging (MRI). OBJECTIVE: To demonstrate the feasibility of substituting automatic for human-demarcated segmentation of major anatomic structures in clinical lumbar spine MRI to generate quantitative image-based features and biomechanical models. SETTING: Previous studies have demonstrated the viability of automatic segmentation applied to medical images; however, the feasibility of these networks to segment clinically acquired images has not yet been demonstrated, as they largely rely on specialized sequences or strict quality of imaging data to achieve good performance. METHODS: Convolutional neural networks were trained to demarcate vertebral bodies, intervertebral disc, and paraspinous muscles from sagittal and axial T1-weighted MRIs. Intervertebral disc height, muscle cross-sectional area, and subject-specific musculoskeletal models of tissue loading in the lumbar spine were then computed from these segmentations and compared against those computed from human-demarcated masks. RESULTS: Segmentation masks, as well as the morphological metrics and biomechanical models computed from those masks, were highly similar between human- and computer-generated methods. Segmentations were similar, with Dice similarity coefficients of 0.77 or greater across networks, and morphological metrics and biomechanical models were similar, with Pearson R correlation coefficients of 0.69 or greater when significant. CONCLUSIONS: This study demonstrates the feasibility of substituting computer-generated for human-generated segmentations of major anatomic structures in lumbar spine MRI to compute quantitative image-based morphological metrics and subject-specific musculoskeletal models of tissue loading quickly, efficiently, and at scale without interrupting routine clinical care.

OBJECTIVES: To evaluate a deep learning model for automated and interpretable classification of central canal stenosis, neural foraminal stenosis, and facet arthropathy from lumbar spine MRI. METHODS: T2-weighted axial MRI studies of the lumbar spine acquired between 2008 and 2019 were retrospectively selected (n = 200) and graded for central canal stenosis, neural foraminal stenosis, and facet arthropathy. Studies were partitioned into patient-level train (n = 150), validation (n = 20), and test (n = 30) splits. V-Net models were first trained to segment the dural sac and the intervertebral disk, and localize facet and foramen using geometric rules. Subsequently, Big Transfer (BiT) models were trained for downstream classification tasks. An interpretable model for central canal stenosis was also trained using a decision tree classifier. Evaluation metrics included linearly weighted Cohens kappa score for multi-grade classification and area under the receiver operator characteristic curve (AUROC) for binarized classification. RESULTS: Segmentation of the dural sac and intervertebral disk achieved Dice scores of 0.93 and 0.94. Localization of foramen and facet achieved intersection over union of 0.72 and 0.83. Multi-class grading of central canal stenosis achieved a kappa score of 0.54. The interpretable decision tree classifier had a kappa score of 0.80. Pairwise agreement between readers (R1, R2), (R1, R3), and (R2, R3) was 0.86, 0.80, and 0.74. Binary classification of neural foraminal stenosis and facet arthropathy achieved AUROCs of 0.92 and 0.93. CONCLUSION: Deep learning systems can be performant as well as interpretable for automated evaluation of lumbar spine MRI including classification of central canal stenosis, neural foraminal stenosis, and facet arthropathy. KEY POINTS: • Interpretable deep-learning systems can be developed for the evaluation of clinical lumbar spine MRI. Multi-grade classification of central canal stenosis with a kappa of 0.80 was comparable to inter-reader agreement scores (0.74, 0.80, 0.86). Binary classification of neural foraminal stenosis and facet arthropathy achieved favorable and accurate AUROCs of 0.92 and 0.93, respectively. • While existing deep-learning systems are opaque, leading to clinical deployment challenges, the proposed system is accurate as well as interpretable, providing valuable information to a radiologist in clinical practice.

Magnetic Resonance Imaging (MRI) offers strong soft tissue contrast but suffers from long acquisition times and requires tedious annotation from radiologists. Traditionally, these challenges have been addressed separately with reconstruction and image analysis algorithms. To see if performance could be improved by treating both as end-to-end, we hosted the K2S challenge, in which challenge participants segmented knee bones and cartilage from 8× undersampled k-space. We curated the 300-patient K2S dataset of multicoil raw k-space and radiologist quality-checked segmentations. 87 teams registered for the challenge and there were 12 submissions, varying in methodologies from serial reconstruction and segmentation to end-to-end networks to another that eschewed a reconstruction algorithm altogether. Four teams produced strong submissions, with the winner having a weighted Dice Similarity Coefficient of 0.910 ± 0.021 across knee bones and cartilage. Interestingly, there was no correlation between reconstruction and segmentation metrics. Further analysis showed the top four submissions were suitable for downstream biomarker analysis, largely preserving cartilage thicknesses and key bone shape features with respect to ground truth. K2S thus showed the value in considering reconstruction and image analysis as end-to-end tasks, as this leaves room for optimization while more realistically reflecting the long-term use case of tools being developed by the MR community.

ObjectiveTo assess the prevalence and clinical implications of variant sciatic nerve anatomy in relation to the piriformis muscle on magnetic resonance neurography (MRN), in patients with lumbosacral neuropathic symptoms.Materials and methodsIn this retrospective single-center study, 254 sciatic nerves, from 127 patients with clinical and imaging findings compatible with extra-spinal sciatica on MRN between 2003 and 2013, were evaluated for the presence and type of variant sciatic nerves, split sciatic nerve, abnormal T2-signal hyperintensity, asymmetric piriformis size and increased nerve caliber, and summarized using descriptive statistics. Two-tailed chi-square tests were performed to compare the anatomical variant type and clinical symptoms between imaging and clinical characteristics.ResultsSixty-four variant sciatic nerves were identified with an equal number of right and left variants. Bilateral variants were noted in 15 cases. Abnormal T2-signal hyperintensity was seen significantly more often in variant compared to conventional anatomy (40/64 vs. 82/190; p = 0.01). A sciatic nerve split was seen significantly more often in variant compared to conventional anatomy (56/64 vs. 20/190; p < 0.0001). Increased nerve caliber, abnormal T2-signal hyperintensity, and asymmetric piriformis size were significantly associated with the clinically symptomatic side compared to the asymptomatic side (98:2, 98:2, and 97:3, respectively; p < 0.0001 for all). Clinical symptoms were correlated with variant compared to conventional sciatic nerve anatomy (64% vs. 46%; p = 0.01).ConclusionVariant sciatic nerve anatomy, in relation to the piriformis muscle, is frequently identified with MRN and is more likely to be associated with nerve signal changes and symptomatology.

BACKGROUND AND PURPOSE: Quantitative metrics of the dural sac such as the cross-sectional area are commonly used to evaluate central canal stenosis. The aim of this study was to analyze 2 new metrics to measure spinal stenosis on the basis of the ratio between the dural sac and disc cross-sectional areas (DDRCA) and the dural sac and disc anterior-posterior diameters (DDRDIA) and compare them with established quantitative metrics of the dural sac. MATERIALS AND METHODS: T2-weighted axial MR images (n = 260 patients) were retrospectively evaluated, graded for central canal stenosis as normal (no stenosis), mild, moderate, or severe from L1/L2 through L5/S1 with 1 grade per spinal level and annotated to measure the DDRCA and DDRDIA. Thresholds were obtained using a decision tree classifier on a subset of patients (n = 130) and evaluated on the remaining patients (n = 130) for accuracy and consistency across demographics, anatomic variation, and clinical outcomes. RESULTS: DDRCA and DDRDIA had areas under the receiver operating characteristic curve of 98.6 (97.4-99.3) and 98.0 (96.7-98.9) compared with dural sac cross-sectional area at 96.5 (95.0-97.7) for binary classification. DDRDIA and DDRCA had κ scores of 0.75 (0.71-0.79) and 0.80 (0.75-0.83) compared with dural sac cross-sectional area at 0.62 (0.57-0.66) for multigrade classification. No significant differences (P > .1) in the area under the receiver operating characteristic curve were observed for the DDRDIA across variations in the body mass index. The DDRDIA also had the highest area under the receiver operating characteristic curve among symptomatic patients (visual analog scale ≥ 7) or patients who underwent surgery. CONCLUSIONS: Ratio-based metrics (DDRDIA and DDRCA) are accurate and robust to anatomic and demographic variability compared with quantitative metrics of the dural sac and better correlated with symptomatology and surgical outcomes.

OBJECTIVE: To qualitatively evaluate the utility of zero echo-time (ZTE) MRI sequences in identifying osseous findings and to compare ZTE with optimized spoiled gradient echo (SPGR) sequences in detecting knee osseous abnormalities. MATERIALS AND METHODS: ZTE and standard knee MRI sequences were acquired at 3T in 100 consecutive patients. Three radiologists rated confidence in evaluating osseous abnormalities and image quality on a 5-grade Likert scale in ZTE compared to standard sequences. In a subset of knees (n = 57) SPGR sequences were also obtained, and diagnostic confidence in identifying osseous structures was assessed, comparing ZTE and SPGR sequences. Statistical significance of using ZTE over SPGR was characterized with a paired t-test. RESULTS: Image quality of the ZTE sequences was rated high by all reviewers with 278 out of 299 (100 studies, 3 radiologists) scores ≥ 4 on the Likert scale. Diagnostic confidence in using ZTE sequences was rated very high confidence in 97%, 85%, 71%, and 73% of the cases for osteophytosis, subchondral cysts, fractures, and soft tissue calcifications/ossifications, respectively. In 74% of cases with osseous findings, reviewer scores indicated confidence levels (score ≥ 3) that ZTE sequences improved diagnostic certainty over standard sequences. The diagnostic confidence in using ZTE over SPGR sequences for osseous structures as well as abnormalities was favorable and statistically significant (p

BackgroundModic changes (MCs) are the most prevalent classification system for describing magnetic resonance imaging (MRI) signal intensity changes in the vertebrae. However, there is a growing need for novel quantitative and standardized methods of characterizing these anomalies, particularly for lesions of transitional or mixed nature, due to the lack of conclusive evidence of their associations with low back pain. This retrospective imaging study aims to develop an interpretable deep learning-based detection tool for voxel-wise mapping of MCs.MethodsSeventy-five lumbar spine MRI exams that presented with acute-to-chronic low back pain, radiculopathy, and other symptoms of the lumbar spine were enrolled. The pipeline consists of two deep convolutional neural networks to generate an interpretable voxel-wise Modic map. First, an autoencoder was trained to segment vertebral bodies from T1-weighted sagittal lumbar spine images. Next, two radiologists segmented and labeled MCs from a combined T1- and T2-weighted assessment to serve as ground truth for training a second autoencoder that performs segmentation of MCs. The voxels in the detected regions were then categorized to the appropriate Modic type using a rule-based signal intensity algorithm. Post hoc, three radiologists independently graded a second dataset with the aid of the model predictions in an artificial (AI)-assisted experiment.ResultsThe model successfully identified the presence of changes in 85.7% of samples in the unseen test set with a sensitivity of 0.71 (±0.072), specificity of 0.95 (±0.022), and Cohen's kappa score of 0.63. In the AI-assisted experiment, the agreement between the junior radiologist and the senior neuroradiologist significantly improved from Cohen's kappa score of 0.52 to 0.58 (p

[This corrects the article DOI: 10.1148/ryai.2021200165.].

PurposeTo test the hypothesis that artificial intelligence (AI) techniques can aid in identifying and assessing lesion severity in the cartilage, bone marrow, meniscus, and anterior cruciate ligament (ACL) in the knee, improving overall MRI interreader agreement.Materials and methodsThis retrospective study was conducted on 1435 knee MRI studies (n = 294 patients; mean age, 43 years ± 15 [standard deviation]; 153 women) collected within three previous studies (from 2011 to 2014). All MRI studies were acquired using high-spatial-resolution three-dimensional fast-spin-echo CUBE sequence. Three-dimensional convolutional neural networks were developed to detect the regions of interest within MRI studies and grade abnormalities of the cartilage, bone marrow, menisci, and ACL. Evaluation included sensitivity, specificity, and Cohen linear-weighted ĸ. The impact of AI-aided grading in intergrader agreement was assessed on an external dataset.ResultsBinary lesion sensitivity reported for all tissues was between 70% and 88%. Specificity ranged from 85% to 89%. The area under the receiver operating characteristic curve for all tissues ranged from 0.83 to 0.93. Deep learning-assisted intergrader Cohen ĸ agreement significantly improved in 10 of 16 comparisons among two attending physicians and two trainees for all tissues.ConclusionThe three-dimensional convolutional neural network had high sensitivity, specificity, and accuracy for knee-lesion-severity scoring and also increased intergrader agreement when used on an external dataset.Supplemental material is available for this article. Keywords: Bone Marrow, Cartilage, Computer Aided Diagnosis (CAD), Computer Applications-3D, Computer Applications-Detection/Diagnosis, Knee, Ligaments, MR-Imaging, Neural Networks, Observer Performance, Segmentation, Statistics © RSNA, 2021See also the commentary by Li and Chang in this issue.: An earlier incorrect version of this article appeared online. This article was corrected on April 16, 2021.

The spine is an articulated, 3D structure with 6 degrees of translational and rotational freedom. Clinical studies have shown spinal deformities are associated with pain and functional disability in both adult and pediatric populations. Clinical decision making relies on accurate characterization of the spinal deformity and monitoring of its progression over time. However, Cobb angle measurements are time‐consuming, are limited by interobserver variability, and represent a simplified 2D view of a 3D structure. Instead, spine deformities can be described by 3D shape parameters, addressing the limitations of current measurement methods. To this end, we develop and validate a deep learning algorithm to automatically extract the vertebral midline (from the upper endplate of S1 to the lower endplate of C7) for frontal and lateral radiographs. Our results demonstrate robust performance across datasets and patient populations. Approximations of 3D spines are reconstructed from the unit normalized midline curves of 20,118 pairs of full spine radiographs belonging to 15,378 patients acquired at our institution between 2008 and 2020. The resulting 3D dataset is used to describe global imbalance parameters in the patient population and to build a statistical shape model to describe global spine shape variations in preoperative deformity patients via eight interpretable shape parameters. The developed method can identify patient subgroups with similar shape characteristics without relying on an existing shape classification system. [ABSTRACT FROM AUTHOR]