Using Deep Learning to Accelerate Knee MRI at 3T: Results of an Interchangeability Study
OBJECTIVE:Deep Learning (DL) image reconstruction has the potential to disrupt the current state of MR imaging by significantly decreasing the time required for MR exams. Our goal was to use DL to accelerate MR imaging in order to allow a 5-minute comprehensive examination of the knee, without compromising image quality or diagnostic accuracy. METHODS:A DL model for image reconstruction using a variational network was optimized. The model was trained using dedicated multi-sequence training, in which a single reconstruction model was trained with data from multiple sequences with different contrast and orientations. Following training, data from 108 patients were retrospectively undersampled in a manner that would correspond with a net 3.49-fold acceleration of fully-sampled data acquisition and 1.88-fold acceleration compared to our standard two-fold accelerated parallel acquisition. An interchangeability study was performed, in which the ability of 6 readers to detect internal derangement of the knee was compared for the clinical and DL-accelerated images. RESULTS:The study demonstrated a high degree of interchangeability between standard and DL-accelerated images. In particular, results showed that interchanging the sequences would result in discordant clinical opinions no more than 4% of the time for any feature evaluated. Moreover, the accelerated sequence was judged by all six readers to have better quality than the clinical sequence. CONCLUSIONS:An optimized DL model allowed for acceleration of knee images which performed interchangeably with standard images for the detection of internal derangement of the knee. Importantly, readers preferred the quality of accelerated images to that of standard clinical images.
The combination of an inflammatory peripheral blood gene expression and imaging biomarkers enhance prediction of radiographic progression in knee osteoarthritis
OBJECTIVE:Predictive biomarkers of progression in knee osteoarthritis are sought to enable clinical trials of structure-modifying drugs. A peripheral blood leukocyte (PBL) inflammatory gene signature, MRI-based bone marrow lesions (BML) and meniscus extrusion scores, meniscal lesions, and osteophytes on X-ray each have been shown separately to predict radiographic joint space narrowing (JSN) in subjects with symptomatic knee osteoarthritis (SKOA). In these studies, we determined whether the combination of the PBL inflammatory gene expression and these imaging findings at baseline enhanced the prognostic value of either alone. METHODS:PBL inflammatory gene expression (increased mRNA for IL-1Î², TNFÎ±, and COX-2), routine radiographs, and 3T knee MRI were assessed in two independent populations with SKOA: an NYU cohort and the Osteoarthritis Initiative (OAI). At baseline and 24â€‰months, subjects underwent standardized fixed-flexion knee radiographs and knee MRI. Medial JSN (mJSN) was determined as the change in medial JSW. Progressors were defined by an mJSN cut-point (â‰¥â€‰0.5â€‰mm/24â€‰months). Models were evaluated by odds ratios (OR) and area under the receiver operating characteristic curve (AUC). RESULTS:We validated our prior finding in these two independent (NYU and OAI) cohorts, individually and combined, that an inflammatory PBL inflammatory gene expression predicted radiographic progression of SKOA after adjustment for age, sex, and BMI. Similarly, the presence of baseline BML and meniscal lesions by MRI or semiquantitative osteophyte score on X-ray each predicted radiographic medial JSN at 24â€‰months. The combination of the PBL inflammatory gene expression and medial BML increased the AUC from 0.66 (pâ€‰=â€‰0.004) to 0.75 (pâ€‰<â€‰0.0001) and the odds ratio from 6.31 to 19.10 (pâ€‰<â€‰0.0001) in the combined cohort of 473 subjects. The addition of osteophyte score to BML and PBL inflammatory gene expression further increased the predictive value of any single biomarker. A causal analysis demonstrated that the PBL inflammatory gene expression and BML independently influenced mJSN. CONCLUSION/CONCLUSIONS:The use of the PBL inflammatory gene expression together with imaging biomarkers as combinatorial predictive biomarkers, markedly enhances the identification of radiographic progressors. The identification of the SKOA population at risk for progression will help in the future design of disease-modifying OA drug trials and personalized medicine strategies.
Vascular Adhesion Protein-1 (VAP-1) as Predictor of Radiographic Severity in Symptomatic Knee Osteoarthritis in the New York University Cohort
BACKGROUND:To investigate the expression of vascular adhesion protein-1 (VAP-1) in joint tissues and serum in symptomatic knee osteoarthritis (SKOA) patients and examine whether VAP-1 levels predict increased risk of disease severity in a cross-sectional study. METHODS:Baseline VAP-1 expression and soluble VAP-1 (sVAP-1) levels were assessed in the synovium synovial fluid and in the serum in cohorts of patients with tibiofemoral medial knee OA and healthy subjects. Standardized fixed-flexion poster anterior knee radiographs scored for Kellgren-Lawrence (KL) grade (0-4) and medial joint space width (JSW). KL1/2 vs. KL3/4 scores defined early and advanced radiographic severity, respectively. Biochemical markers assessed in serum or synovial fluids (SF) comprised sVAP-1, interleukin 1 receptor antagonist (IL-1Ra), interleukin 6 (IL-6), soluble receptor for advanced glycation end-products (sRAGE), C-C motif chemokine ligand 2 (CCL2), C-C motif chemokine ligand 4 (CCL4), cluster of differentiation 163 (CD163), high sensitivity C-reactive protein (hsCRP), and matrix metalloproteinases (MMPs)-1,-3,-9. Associations between biomarkers and radiographic severity KL1/2 vs. KL3/4 (logistic regression controlling for covariates) and pain (Spearman correlation) were evaluated. RESULTS:Elevated levels of sVAP-1 observed in OA synovial fluid and VAP-1 expression in synovium based on immunohistochemical, microarray, and real-time quantitative polymerase chain reaction (qRT-PCR) analyses. However, serum sVAP-1 levels in OA patients were lower than in controls and inversely correlated with pain and inflammation markers (hsCRP and soluble RAGE). Soluble VAP-1 levels in serum were also lower in radiographically advanced (KL3/4) compared with early KL1/2 knee SKOA patients. CONCLUSION/CONCLUSIONS:Local (synovial fluid) semicarbazide-sensitive amine oxidase (SSAO)/sVAP-1 levels were elevated in OA and correlated with radiographic severity. However, systemic (serum) sVAP-1 levels were lower in SKOA patients than normal and inversely correlated with pain and inflammation markers. Serum sVAP-1 levels were higher in early (KL1/2) compared with advanced (KL3/4) SKOA patients.
The Use of an Emergency Department Expeditor to Improve Emergency Department CTÂ Workflow: Initial Experiences
Vascular adhesion protein-1 (VAP-1) as predictor of radiographic severity in symptomatic knee osteoarthritis [Meeting Abstract]
Background/Purpose: To investigate the expression of vascular adhesion protein -1 (VAP-1) in joint tissues and serum in knee osteoarthritis (OA) patients and examine whether VAP-1 levels predict increased risk of disease severity or progression of knee OA.
Method(s): Baseline serum and synovial fluid VAP-1/semicarbazide-sensitive amine oxidase (SSAO) levels were assessed in cohorts of patients with tibiofemoral medial knee OA and healthy subjects. Standardized fixed-flexion posteroanterior knee radiographs were scored for Kellgren Lawrence (KL) grade (0-4) and medial joint space width (JSW) at the mid-portion of the joint space. Radiographic severity was defined by KL2/3 vs. KL4. Biochemical markers assessed comprised VAP-1/ SSAO, IL-1Ra, IL-6, sRAGE, CCL2, CCL4, CD163, hsCRP and MMPs-1,-3,-9. Associations between biomarkers and radiographic severity (logistic regression controlling for covariates) and pain (Spearman correlation) were evaluated.
Result(s): VAP-1 was locally overexpressed at least 2 fold in the OA synovium based on immunohistochemical, microarray and qRT-PCR analyses compared to controls. Synovial fluid SSAO levels was also significantly higher in OA (107.94+41.42) compared to normals (38.12 + 22.98 ng/ml; p=0.0001) and inversely associated with radiographic severity. We observed a positive correlation with the levels of SSAO in the synovial fluid and serum of OA patients (r=0.47; p=0.014). However, serum SSAO levels in OA patients were lower than in controls, and inversely correlated with pain and inflammation markers (CRP and soluble RAGE). Serum SSAO levels were also lower in radiographically severe (KL4) OA patients compared to KL2/3. Serum SSAO did not correlate with other markers of inflammation or radiographic joint space narrowing (JSN) over 24 months.
Conclusion(s): Synovial fluid VAP-1/SSAO levels were elevated in OA and correlate with radiographic severity. However, serum or circulating SSAO levels are lower in OA patients and inversely correlate with pain and inflammation. Serum VAP-1 levels could identify patients at increased risk for knee radiographic severity
The Reading Room Coordinator: Reducing Radiologist Burnout in the Digital Age
Combinatorial Peripheral Blood Inflammatory and MRI-Based Biomarkers Predict Radiographic Joint Space Narrowing in Knee OA [Meeting Abstract]
Sparse-SEMAC: rapid and improved SEMAC metal implant imaging using SPARSE-SENSE acceleration
PURPOSE: To develop an accelerated SEMAC metal implant MRI technique (Sparse-SEMAC) with reduced scan time and improved metal distortion correction. METHODS: Sparse-SEMAC jointly exploits the inherent sparsity along the additional phase-encoding dimension and multicoil encoding capabilities to significantly accelerate data acquisition. A prototype pulse sequence with pseudorandom ky -kz undersampling and an inline image reconstruction was developed for integration in clinical studies. Three patients with hip implants were imaged using the proposed Sparse-SEMAC with eight-fold acceleration and compared with the standard-SEMAC technique used in clinical studies (three-fold GRAPPA acceleration). Measurements were performed with SEMAC-encoding steps (SES) = 15 for Sparse-SEMAC and SES = 9 for Standard-SEMAC using high spatial resolution Proton Density (PD) and lower-resolution STIR acquisitions. Two expert musculoskeletal (MSK) radiologists performed a consensus reading to score image-quality parameters. RESULTS: Sparse-SEMAC enables up to eight-fold acceleration of data acquisition that results in two-fold scan time reductions, compared with Standard-SEMAC, with improved metal artifact correction for patients with hip implants without degrading spatial resolution. CONCLUSION: The high acceleration enabled by Sparse-SEMAC would enable clinically feasible examination times with improved correction of metal distortion. Magn Reson Med, 2016. (c) 2016 Wiley Periodicals, Inc.
Serum Urate Levels Predict Joint Space Narrowing in Non-gout Patients with Medial Knee Osteoarthritis
OBJECTIVE: OA pathogenesis includes both mechanical and inflammatory features. Studies have implicated synovial fluid urate (UA) as a potential OA biomarker, possibly reflecting chondrocyte damage. Whether serum urate (sUA) levels reflect/contribute to OA is unknown. We investigated whether sUA predicts OA progression in a non-gout knee OA population. METHODS: Eighty-eight subjects with medial knee OA (BMI <33) but without gout were included. Baseline sUA was measured in previously banked serum. At 0 and 24 months, subjects underwent standardized weight-bearing fixed-flexion posteroanterior knee radiographs to determine joint space width (JSW) and Kellgren-Lawrence (KL) grades. Joint space narrowing (JSN) was determined as JSW change from 0 to 24 months. Twenty-seven subjects underwent baseline contrast-enhanced 3T knee MRI for synovial volume (SV) assessment. RESULTS: sUA correlated with JSN in both univariate (r=0.40, p=0.01) and multivariate analyses (r=0.28, p=0.01). There was a significant difference in mean JSN after dichotomizing at sUA of 6.8 mg/dL, the solubility point for serum urate, even after adjustment (JSN of 0.90 mm for sUA>/=6.8; JSN of 0.31 mm for sUA<6.8, p<0.01). Baseline sUA distinguished progressors (JSN>0.2mm) and fast progressors (JSN>0.5mm) from nonprogressors (JSN=0.0mm) in multivariate analyses (area under the receiver operating characteristic curve 0.63, p=0.03; AUC 0.62, p=0.05, respectively). sUA correlated with SV (r=0.44, p<0.01), a possible marker of JSN, though this correlation did not persist after controlling for age, gender and BMI (r=0.13, p=0.56). CONCLUSIONS: In non-gout patients with knee OA, sUA predicted future JSN and may serve as a biomarker for OA progression
Accelerated knee imaging using a deep learning based reconstruction [Meeting Abstract]