Novel 3D printable PEEK-HA-Mg2SiO4 composite material for spine implants: biocompatibility and imaging compatibility assessments.

PURPOSE: To develop a novel 3D printable polyether ether ketone (PEEK)-hydroxyapatite (HA)-magnesium orthosilicate (Mg2SiO4) composite material with enhanced properties for potential use in tumour, osteoporosis and other spinal conditions. We aim to evaluate biocompatibility and imaging compatibility of the material. METHODS: Materials were prepared in three different compositions, namely composite A: 75 weight % PEEK, 20 weight % HA, 5 weight % Mg2SiO4; composite B: 70 weight% PEEK, 25 weight % HA, 5 weight % Mg2SiO4; and composite C: 65 weight % PEEK, 30 weight % HA, 5 weight % Mg2SiO4. The materials were processed to obtain 3D printable filament. Biomechanical properties were analysed as per ASTM standards and biocompatibility of the novel material was evaluated using indirect and direct cell cytotoxicity tests. Cell viability of the novel material was compared to PEEK and PEEK-HA materials. The novel material was used to 3D print a standard spine cage. Furthermore, the CT and MR imaging compatibility of the novel material cage vs PEEK and PEEK-HA cages were evaluated using a phantom setup. RESULTS: Composite A resulted in optimal material processing to obtain a 3D printable filament, while composite B and C resulted in non-optimal processing. Composite A enhanced cell viability up to ~ 20% compared to PEEK and PEEK-HA materials. Composite A cage generated minimal/no artefacts on CT and MR imaging and the images were comparable to that of PEEK and PEEK-HA cages. CONCLUSION: Composite A demonstrated superior bioactivity vs PEEK and PEEK-HA materials and comparable imaging compatibility vs PEEK and PEEK-HA. Therefore, our material displays an excellent potential to manufacture spine implants with enhanced mechanical and bioactive property.

A deep learning-based technique for the diagnosis of epidural spinal cord compression on thoracolumbar CT.

PURPOSE: To develop a deep learning (DL) model for epidural spinal cord compression (ESCC) on CT, which will aid earlier ESCC diagnosis for less experienced clinicians. METHODS: We retrospectively collected CT and MRI data from adult patients with suspected ESCC at a tertiary referral institute from 2007 till 2020. A total of 183 patients were used for training/validation of the DL model. A separate test set of 40 patients was used for DL model evaluation and comprised 60 staging CT and matched MRI scans performed with an interval of up to 2 months. DL model performance was compared to eight readers: one musculoskeletal radiologist, two body radiologists, one spine surgeon, and four trainee spine surgeons. Diagnostic performance was evaluated using inter-rater agreement, sensitivity, specificity and AUC. RESULTS: Overall, 3115 axial CT slices were assessed. The DL model showed high kappa of 0.872 for normal, low and high-grade ESCC (trichotomous), which was superior compared to a body radiologist (R4, κ = 0.667) and all four trainee spine surgeons (κ range = 0.625-0.838)(all p < 0.001). In addition, for dichotomous normal versus any grade of ESCC detection, the DL model showed high kappa (κ = 0.879), sensitivity (91.82), specificity (92.01) and AUC (0.919), with the latter AUC superior to all readers (AUC range = 0.732-0.859, all p < 0.001). CONCLUSION: A deep learning model for the objective assessment of ESCC on CT had comparable or superior performance to radiologists and spine surgeons. Earlier diagnosis of ESCC on CT could reduce treatment delays, which are associated with poor outcomes, increased costs, and reduced survival.

Improved Productivity Using Deep Learning-assisted Reporting for Lumbar Spine MRI.

Background Lumbar spine MRI studies are widely used for back pain assessment. Interpretation involves grading lumbar spinal stenosis, which is repetitive and time consuming. Deep learning (DL) could provide faster and more consistent interpretation. Purpose To assess the speed and interobserver agreement of radiologists for reporting lumbar spinal stenosis with and without DL assistance. Materials and Methods In this retrospective study, a DL model designed to assist radiologists in the interpretation of spinal canal, lateral recess, and neural foraminal stenoses on lumbar spine MRI scans was used. Randomly selected lumbar spine MRI studies obtained in patients with back pain who were 18 years and older over a 3-year period, from September 2015 to September 2018, were included in an internal test data set. Studies with instrumentation and scoliosis were excluded. Eight radiologists, each with 2-13 years of experience in spine MRI interpretation, reviewed studies with and without DL model assistance with a 1-month washout period. Time to diagnosis (in seconds) and interobserver agreement (using Gwet κ) were assessed for stenosis grading for each radiologist with and without the DL model and compared with test data set labels provided by an external musculoskeletal radiologist (with 32 years of experience) as the reference standard. Results Overall, 444 images in 25 patients (mean age, 51 years ± 20 [SD]; 14 women) were evaluated in a test data set. DL-assisted radiologists had a reduced interpretation time per spine MRI study, from a mean of 124-274 seconds (SD, 25-88 seconds) to 47-71 seconds (SD, 24-29 seconds) (P < .001). DL-assisted radiologists had either superior or equivalent interobserver agreement for all stenosis gradings compared with unassisted radiologists. DL-assisted general and in-training radiologists improved their interobserver agreement for four-class neural foraminal stenosis, with κ values of 0.71 and 0.70 (with DL) versus 0.39 and 0.39 (without DL), respectively (both P < .001). Conclusion Radiologists who were assisted by deep learning for interpretation of lumbar spinal stenosis on MRI scans showed a marked reduction in reporting time and superior or equivalent interobserver agreement for all stenosis gradings compared with radiologists who were unassisted by deep learning. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Hayashi in this issue.

Effect of Training Data Volume on Performance of Convolutional Neural Network Pneumothorax Classifiers.

Large datasets with high-quality labels required to train deep neural networks are challenging to obtain in the radiology domain. This work investigates the effect of training dataset size on the performance of deep learning classifiers, focusing on chest radiograph pneumothorax detection as a proxy visual task in the radiology domain. Two open-source datasets (ChestX-ray14 and CheXpert) comprising 291,454 images were merged and convolutional neural networks trained with stepwise increase in training dataset sizes. Model iterations at each dataset volume were evaluated on an external test set of 525 emergency department chest radiographs. Learning curve analysis was performed to fit the observed AUCs for all models generated. For all three network architectures tested, model AUCs and accuracy increased rapidly from 2 × 103 to 20 × 103 training samples, with more gradual increase until the maximum training dataset size of 291 × 103 images. AUCs for models trained with the maximum tested dataset size of 291 × 103 images were significantly higher than models trained with 20 × 103 images: ResNet-50: AUC20k = 0.86, AUC291k = 0.95, p < 0.001; DenseNet-121 AUC20k = 0.85, AUC291k = 0.93, p < 0.001; EfficientNet AUC20k = 0.92, AUC 291 k = 0.98, p < 0.001. Our study established learning curves describing the relationship between dataset training size and model performance of deep learning convolutional neural networks applied to a typical radiology binary classification task. These curves suggest a point of diminishing performance returns for increasing training data volumes, which algorithm developers should consider given the high costs of obtaining and labelling radiology data.

Deep Learning Model for Automated Detection and Classification of Central Canal, Lateral Recess, and Neural Foraminal Stenosis at Lumbar Spine MRI.

Background Assessment of lumbar spinal stenosis at MRI is repetitive and time consuming. Deep learning (DL) could improve -productivity and the consistency of reporting. Purpose To develop a DL model for automated detection and classification of lumbar central canal, lateral recess, and neural -foraminal stenosis. Materials and Methods In this retrospective study, lumbar spine MRI scans obtained from September 2015 to September 2018 were included. Studies of patients with spinal instrumentation or studies with suboptimal image quality, as well as postgadolinium studies and studies of patients with scoliosis, were excluded. Axial T2-weighted and sagittal T1-weighted images were used. Studies were split into an internal training set (80%), validation set (9%), and test set (11%). Training data were labeled by four radiologists using predefined gradings (normal, mild, moderate, and severe). A two-component DL model was developed. First, a convolutional neural network (CNN) was trained to detect the region of interest (ROI), with a second CNN for classification. An internal test set was labeled by a musculoskeletal radiologist with 31 years of experience (reference standard) and two subspecialist radiologists (radiologist 1: A.M., 5 years of experience; radiologist 2: J.T.P.D.H., 9 years of experience). DL model performance on an external test set was evaluated. Detection recall (in percentage), interrater agreement (Gwet κ), sensitivity, and specificity were calculated. Results Overall, 446 MRI lumbar spine studies were analyzed (446 patients; mean age ± standard deviation, 52 years ± 19; 240 women), with 396 patients in the training (80%) and validation (9%) sets and 50 (11%) in the internal test set. For internal testing, DL model and radiologist central canal recall were greater than 99%, with reduced neural foramina recall for the DL model (84.5%) and radiologist 1 (83.9%) compared with radiologist 2 (97.1%) (P < .001). For internal testing, dichotomous classification (normal or mild vs moderate or severe) showed almost-perfect agreement for both radiologists and the DL model, with respective κ values of 0.98, 0.98, and 0.96 for the central canal; 0.92, 0.95, and 0.92 for lateral recesses; and 0.94, 0.95, and 0.89 for neural foramina (P < .001). External testing with 100 MRI scans of lumbar spines showed almost perfect agreement for the DL model for dichotomous classification of all ROIs (κ, 0.95-0.96; P < .001). Conclusion A deep learning model showed comparable agreement with subspecialist radiologists for detection and classification of central canal and lateral recess stenosis, with slightly lower agreement for neural foraminal stenosis at lumbar spine MRI. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Hayashi in this issue.

The peroneus longus muscle and tendon: a review of its anatomy and pathology.

This article will review the anatomy and common pathologies affecting the peroneus longus muscle and tendon. The anatomy of the peroneus longus is complex and its long course can result in symptomatology referable to the lower leg, ankle, hindfoot, and plantar foot. Proximally, the peroneus longus muscle lies within the lateral compartment of the lower leg with its distal myotendinous junction arising just above the level of the ankle. The distal peroneus longus tendon has a long course and makes two sharp turns at the lateral ankle and hindfoot before inserting at the medial plantar foot. A spectrum of pathology can occur in these regions. At the lower leg, peroneus longus muscle injuries (e.g., denervation) along with retromalleolar tendon instability/subluxation will be discussed. More distally, along the lateral calcaneus and cuboid tunnel, peroneus longus tendinosis and tears, tenosynovitis, and painful os peroneum syndrome (POPS) will be covered. Pathology of the peroneus longus will be illustrated using clinical case examples along its entire length; these will help the radiologist understand and interpret common peroneus longus disorders.

Comparison of magnetic resonance elastography and diffusion-weighted imaging for differentiating benign and malignant liver lesions.

OBJECTIVES: Comparison of magnetic resonance elastography (MRE) and diffusion-weighted imaging (DWI) for differentiating malignant and benign focal liver lesions (FLLs). METHODS: Seventy-nine subjects with 124 FLLs (44 benign and 80 malignant) underwent both MRE and DWI. MRE was performed with a modified gradient-echo sequence and DWI with a free breathing technique (b = 0.500). Apparent diffusion coefficient (ADC) maps and stiffness maps were generated. FLL mean stiffness and ADC values were obtained by placing regions of interest over the FLLs on stiffness and ADC maps. The accuracy of MRE and DWI for differentiation of benign and malignant FLL was compared using receiver operating curve (ROC) analysis. RESULTS: There was a significant negative correlation between stiffness and ADC (r = -0.54, p < 0.0001) of FLLs. Malignant FLLs had significantly higher mean stiffness (7.9kPa vs. 3.1kPa, p < 0.001) and lower mean ADC (129 vs. 200 × 10(-3)mm(2)/s, p < 0.001) than benign FLLs. The sensitivity/specificity/positive predictive value/negative predictive value for differentiating malignant from benign FLLs with MRE (cut-off, >4.54kPa) and DWI (cut-off, <151 × 10(-3)mm(2)/s) were 96.3/95.5/97.5/93.3% (p < 0.001) and 85/81.8/88.3/75% (p < 0.001), respectively. ROC analysis showed significantly higher accuracy for MRE than DWI (0.986 vs. 0.82, p = 0.0016). CONCLUSION: MRE is significantly more accurate than DWI for differentiating benign and malignant FLLs. KEY POINTS: • MRE is superior to DWI for differentiating benign and malignant focal liver lesions. • Benign lesions with large fibrous components may have higher stiffness with MRE. • Cholangiocarcinomas tend to have higher stiffness than hepatocellular carcinomas. • Hepatocellular adenomas tend to have lower stiffness than focal nodular hyperplasia. • MRE is superior to conventional MRI in differentiating benign and malignant liver lesions.

Magnetic resonance elastography of liver: influence of intravenous gadolinium administration on measured liver stiffness.

PURPOSE: To evaluate the effect of intravenous gadolinium-diethylenetriamine penta-acetic acid (Gd-DTPA) on estimation of liver stiffness using magnetic resonance elastography (MRE) for detection of liver fibrosis. MATERIALS AND METHODS: Liver MRI with MRE was performed in 210 subjects on a single 1.5 Tesla clinical MRI scanner. Liver MRE was performed before intravenous Gd-DTPA injection (NC-MRE) and 5 minutes post injection (PC-MRE) using a modified phase-contrast gradient-echo sequence (TR/TE=100/27 ms, FOV = 30-46 cm, 4 x 10 mm slices, gap 5 mm) which automatically generated stiffness maps. Two readers' blinded to clinical details independently performed liver stiffness measurements (LSM) by drawing 2 or more regions of interest (ROI) on the stiffness maps on each of the four slices of NC-MRE and PC-MRE obtained for each patient. The mean LSM in kilopascals (kPa) for NC-MRE and PC-MRE was calculated. The correlation between NC-MRE and PC-MRE LSM was evaluated with a paired t test and Pearson's correlation analysis, and the inter-observer correlation was evaluated using intra class coefficient (ICC) analysis. A receiver operating curve analysis (ROC) was performed to compare accuracies for detection and staging of liver fibrosis in a subgroup of 72 subjects with histological confirmation of liver fibrosis. RESULTS: There was an excellent correlation between NC-MRE and PC-MRE LSM (R(2)=0.98, p<0.001) with no significant differences. The interobserver agreement was also excellent (ICC, 0.94-0.99). There were no significant differences in the cut-off LSM value/accuracy/sensitivity/specificity for detection of significant liver fibrosis with NC-MRE and PC-MRE (2.98 kPa/98.5%/100%/88%, p<0.001 and 3.1 kPa/98.2%/98%/88%, p<0.001 respectively). CONCLUSION: Intravenous Gd-DTPA had no significant influence on LSM with MRE and does not significantly affect the diagnostic performance of MRE for fibrosis detection.

Pedicle Screw Pseudofracture on Computed Tomography Secondary to Metal Artifact Reduction.

Metal artifact reduction (MAR) algorithms are commonly used in computed tomography (CT) scans where metal implants are involved. However, MAR algorithms also have the potential to create new artifacts in reconstructed images. We present a case of a screw pseudofracture due to MAR on CT.

Reducing Delays in MRIs under Sedation and General Anesthesia Using Quality Improvement Tools.

OBJECTIVE: Develop structured, quality improvement (QI) interventions to achieve a 15%-point reduction in MRIs performed under sedation or general anesthesia (GA) delayed over 15 minutes within a 6-month period. METHODS: A prospective audit of MRIs under sedation or GA from January 2022 to June 2023 was conducted. A multidisciplinary team performed process mapping and root cause analysis for delays. Interventions were developed and implemented over four 'Plan, Do, Study, Act' (PDSA) cycles, targeting workflow standardization, pre-admission patient counselling, reinforcing adherence to scheduled scan times and written consent respectively. Delay times (compared with Kruskal-Wallis and Dunn's tests), delays over 15 minutes and delays of 60 or more minutes at baseline and after each PDSA cycle were recorded. RESULTS: 627 MRIs under sedation or GA were analyzed, comprising 443 at baseline and 184 post-implementation. 556/627 (88.7%) scans were performed under sedation, 22/627 (3.5%) under monitored anesthesia care and 49/627 (7.8%) under GA. At baseline, 71.6% (317/443) scans were delayed over 15 minutes and 28.2% (125/443) scans by 60 or more minutes, with a median delay of 30 minutes. Post-implementation, there was a 34.7%-point reduction in scans delayed over 15 minutes, 17.5%-point reduction in scans delayed by 60 or more minutes and reduced median delay time by 15 minutes (p <0.001). DISCUSSION: Structured interventions significantly reduced delays in MRIs under sedation and GA, potentially improving outcomes for both patients and providers. Key factors included a diversity of perspectives in the study team, continued stakeholder engagement and structured QI tools including PDSA cycles.

Application of Artificial Intelligence Methods on Osteoporosis Classification with Radiographs-A Systematic Review.

Osteoporosis is a complex endocrine disease characterized by a decline in bone mass and microstructural integrity. It constitutes a major global health problem. Recent progress in the field of artificial intelligence (AI) has opened new avenues for the effective diagnosis of osteoporosis via radiographs. This review investigates the application of AI classification of osteoporosis in radiographs. A comprehensive exploration of electronic repositories (ClinicalTrials.gov, Web of Science, PubMed, MEDLINE) was carried out in adherence to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 statement (PRISMA). A collection of 31 articles was extracted from these repositories and their significant outcomes were consolidated and outlined. This encompassed insights into anatomical regions, the specific machine learning methods employed, the effectiveness in predicting BMD, and categorizing osteoporosis. Through analyzing the respective studies, we evaluated the effectiveness and limitations of AI osteoporosis classification in radiographs. The pooled reported accuracy, sensitivity, and specificity of osteoporosis classification ranges from 66.1% to 97.9%, 67.4% to 100.0%, and 60.0% to 97.5% respectively. This review underscores the potential of AI osteoporosis classification and offers valuable insights for future research endeavors, which should focus on addressing the challenges in technical and clinical integration to facilitate practical implementation of this technology.