Diagnostic performance and inter-reader reliability of bone reporting and data system (Bone-RADS) on computed tomography.

OBJECTIVE: To evaluate the diagnostic performance and inter-reader reliability of the Bone Reporting and Data System (Bone-RADS) for solitary bone lesions on CT. MATERIALS AND METHODS: This retrospective analysis included 179 patients (mean age, 56 ± 18 years; 94 men) who underwent bone biopsies between March 2005 and September 2021. Patients with solitary bone lesions on CT and sufficient histopathology results were included. Two radiologists categorized the bone lesions using the Bone-RADS (1, benign; 4, malignant). The diagnostic performance of the Bone-RADS was calculated using histopathology results as a standard reference. Inter-reader reliability was calculated. RESULTS: Bone lesions were categorized into two groups: 103 lucent (pathology: 34 benign, 12 intermediate, 54 malignant, and 3 osteomyelitis) and 76 sclerotic/mixed (pathology: 46 benign, 2 intermediate, 26 malignant, and 2 osteomyelitis) lesions. The Bone-RADS for lucent lesions had sensitivities of 95% and 82%, specificities of 11% and 11%, and accuracies of 57% and 50% for readers 1 and 2, respectively. The Bone-RADS for sclerotic/mixed lesions had sensitivities of 75% and 68%, specificities of 27% and 27%, and accuracies of 45% and 42% for readers 1 and 2, respectively. Inter-reader reliability was moderate to very good (κ = 0.744, overall; 0.565, lucent lesions; and 0.851, sclerotic/mixed lesions). CONCLUSION: Bone-RADS has a high sensitivity for evaluating malignancy in lucent bone lesions and good inter-reader reliability. However, it has poor specificity and accuracy for both lucent and sclerotic/mixed lesions. A possible explanation is that proposed algorithms heavily depend on clinical features such as pain and history of malignancy.

Clinical characteristics of tetralogy of Fallot with left pulmonary artery acute angulation on computed tomography.

Left pulmonary artery (LPA) stenosis with acute angulation commonly necessitates surgical revision in the treatment of tetralogy of Fallot (TOF). We investigated the clinical characteristics of acute angulation of the LPA in patients with TOF via computed tomography (CT). Between 2011 and 2022, 160 patients were diagnosed with TOF using CT. After excluding 28 patients due to insufficient records or age, 132 patients were included in the present analysis. The patients were divided into two groups according to the presence or absence of acute angulation of the LPA: group I (TOF with an acute angle of the LPA, n = 53) and group II (TOF without an acute angle of the LPA, n = 79). We retrospectively collected clinical data from electronic medical records. T-tests were used to analyze continuous variables (i.e., age, sex, weight, right ventricular outflow tract [RVOT] pressure on echocardiography, and distance to bifurcation), and Fisher's exact and chi-square tests were used to analyze categorical data (i.e., presence of the right aortic arch, pulmonary arterial atresia/hypoplasia, atrial septal defect [ASD], patent ductus arteriosus [PDA], and pulmonary valve). The groups did not differ significantly in terms of sex, weight, presence of the right aortic arch, pulmonary arterial atresia/hypoplasia, ASD, or distance to the bifurcation. Moreover, there was no significant difference between the presence and absence of PDA (P = 0.057); however, patients in group I were significantly older (143.2 ± 97.8 days) than those in group II (91.1 ± 76.0 days, P = 0.002). Furthermore, RVOT pressure was significantly higher among patients in group I (45.1 ± 22.5 mmHg) than in group II (25.0 ± 12.4 mmHg, P < 0.001). In the current study, acute angulation of the LPA in patients with TOF, as observed on CT, was associated with older age and higher RVOT pressure on echocardiography.

Osteomyelitis on MR imaging as a key predictor of recurrent septic arthritis of the shoulder.

OBJECTIVES: To investigate the clinical and radiologic predictors of postoperative recurrent septic arthritis of the shoulder (SAS) using multivariable analysis. METHODS: Forty-three patients (mean age, 65 years; 24 women) who underwent surgery for SAS between January 2011 and October 2019 were retrospectively enrolled. An orthopedic surgeon assessed clinical (age, sex, comorbidity, duration from symptom onset to MR imaging and surgery, surgical method, antibiotic usage), laboratory (serum white blood cell [WBC] count, C-reactive protein [CRP] level, synovial cell count), and surgical findings (culture/biopsy results). Two musculoskeletal radiologists evaluated MR imaging findings (bone marrow [reactive bone marrow edema, osteomyelitis, osteochondral erosion] and soft tissue [synovitis, bursitis, muscle edema, abscess] abnormalities). Recurrent SAS was evaluated at ≥ 12 months postoperatively. Univariable and multivariable analyses were performed to determine the best predictor of recurrent SAS. RESULTS: The overall recurrent SAS rate was 33% (14/43). On univariable analysis, mean age (without recurrence vs. recurrence: 68 vs. 59 years, p = .04), mean duration from symptom onset to surgery (18 vs. 25 days, p = .02), serum WBC count (12,000 vs. 9,000 cells/mL3, p = .04), CRP level (13 vs. 6 mg/L, p = .01), and osteomyelitis on MR imaging (p < .01 for both readers) significantly differed between patients with and without recurrence; on multivariable analysis, only osteomyelitis on MR imaging was significantly associated with recurrent SAS for both readers (p = .02 and .01 for each reader respectively). The inter-reader agreement was good (κ = .62-1.0) for all MR imaging findings, except for muscle edema (fair, κ = .37). CONCLUSION: Osteomyelitis on MR imaging was the best predictor of recurrent SAS. KEY POINTS: • Osteomyelitis on preoperative MR imaging was the best predictor associated with recurrent septic arthritis of the shoulder on multivariable analysis including clinical, laboratory, and MR findings. • In multivariable analyses focused on each bone marrow abnormality, with adjustment for clinical and laboratory parameters, the presence of reactive bone marrow edema and osteochondral erosion on MR imaging showed no significant association with recurrent septic arthritis of the shoulder.

Factors affecting the occurrence of osseous lesions in septic shoulder arthritis and the recurrence rate after arthroscopic surgery.

HYPOTHESIS: The purpose of this study was to determine the incidence of osseous lesions and the recurrence rate after arthroscopic surgery in shoulder septic arthritis patients and evaluate the influencing factors. MATERIALS AND METHODS: We retrospectively reviewed 44 patients who underwent arthroscopic surgery for septic arthritis of the shoulder between January 2012 and September 2019. The average age of the patients was 65.57 ± 14.2 years, and 56.8% were female patients. The minimum follow-up period was 12 months (average, 32.8 ± 14.2 months; range, 12-72 months). We assessed variables including sex, age, underlying diseases, duration from symptom onset to magnetic resonance imaging (MRI), duration from symptom onset to surgery, radiologic results (radiography and MRI), history of injection therapy, and postoperative infection. The incidence of osseous lesions and the recurrence rate were calculated according to independent variables. In addition, multivariate logistic regression was performed to identify the risk factors for osseous lesions and recurrent infection after adjustment for other variables. RESULTS: Twenty-one patients had an osseous lesion on MRI, and 12 patients had evidence of bone erosion on radiographs. In univariate analyses, significant (P < .05) risk factors for the presence of osseous lesions were female sex, lower C-reactive protein level, and longer duration from symptom onset to MRI. The overall infection recurrence rate was 22.7% (10 of 44 patients). Culture results and the duration from symptom onset to surgery were significant risk factors for recurrent infection (P < .05). As the duration from symptom onset to MRI increased by 1 day, the probability of osseous lesions increased 1.31-fold (95% confidence interval, 1.08- to 1.59-fold; P = .007), and this probability was significantly higher after correction for other risk factors. CONCLUSIONS: To reduce the severity of septic shoulder infection, timely diagnosis and treatment are essential. Even if osseous lesions are present, good results can be obtained if meticulous débridement is performed through arthroscopic surgery. However, functional and radiologic long-term follow-up studies are needed in patients with osseous lesions.

CT iterative vs deep learning reconstruction: comparison of noise and sharpness.

OBJECTIVES: To compare image noise and sharpness of vessels, liver, and muscle in lower extremity CT angiography between "adaptive statistical iterative reconstruction-V" (ASIR-V) and deep learning reconstruction "TrueFidelity" (TFI). METHODS: Thirty-seven patients (mean age, 65.2 years; 32 men) with lower extremity CT angiography were enrolled between November and December 2019. Images were reconstructed with two ASIR-V (blending factor of 80% and 100% (AV-100)) and three TFI (low-, medium-, and high-strength-level (TF-H) settings). Two radiologists evaluated these images for vessels (aorta, femoral artery, and popliteal artery), liver, and psoas muscle. For quantitative analyses, conventional indicators (CT number, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR)) and blur metric values (indicating the degree of image sharpness) of selected regions of interest were determined. For qualitative analyses, the degrees of quantum mottle and blurring were assessed. RESULTS: The higher the blending factor in ASIR-V or the strength in TFI, the lower the noise, the higher the SNR and CNR values, and the higher the blur metric values in all structures. The SNR and CNR values of TF-H images were significantly higher than those of AV-80 images and similar to those of AV-100 images. The blur metric values in TFI images were significantly lower than those in ASIR-V images (p < 0.001), indicating increased sharpness. Among all the investigated image procedures, the overall qualitative image quality was best in TF-H images. CONCLUSION: TF-H was the most balanced image in terms of image noise and sharpness among the examined image combinations. KEY POINTS: • Deep learning image reconstruction "TrueFidelity" is superior to iterative reconstruction "ASIR-V" regarding image noise and sharpness. • The high-strength "TrueFidelity" approach generated the best image quality among the examined image reconstruction procedures. • In iterative and deep learning CT image reconstruction, the higher the blending and strength factors, the lower the image noise and the poorer the image sharpness.

Bone marrow edema in non-traumatic hip: high accuracy of dual-energy CT with water-hydroxyapatite decomposition imaging.

OBJECTIVES: To evaluate the diagnostic performance of dual-energy CT with water-hydroxyapatite (HAP) imaging for bone marrow edema in patients with non-traumatic hip pain. METHODS: Forty patients (mean age, 58 years; 16 male and 24 female) who underwent rapid kVp-switching dual-energy CT and MRI within 1 month between April 2018 and February 2019 with hip pain but no trauma were enrolled. Two radiologists retrospectively evaluated 80 hip joints for the presence, extent (femoral head involved, head and neck, and head to intertrochanter), and severity (mild edema, moderate, severe) of bone marrow edema on dual-energy water-HAP images. Water mass density (mg/cm3) on water-HAP images was determined with region of interest-based quantitative analysis. MRI served as the standard of reference. RESULTS: Sensitivity, specificity, and accuracy of readers 1 and 2 for the identification of bone marrow edema in water-HAP images were 85% and 85%, 93% and 73%, and 89% and 79%, respectively. The area under the receiver operating characteristic curve was 0.96 for reader 1 and 0.91 for reader 2 for differentiation of the presence of edema from no edema. The optimal water mass density to classify the presence of edema for reader 1 was 951 mg/cm3 with 93% sensitivity and 93% specificity and for reader 2 was 957 mg/cm3 with 80% sensitivity and 80% specificity. The more severe the edema, the higher was the mean water density value (p < 0.035). CONCLUSION: Dual-energy water-HAP images showed good diagnostic performance for bone marrow edema in patients with non-traumatic hip pain. KEY POINTS: • Dual-energy water-HAP imaging depicts bone marrow edema in patients with non-traumatic hip pain and may serve as an alternative to MRI in select patients. • A cutoff value of 951 mg/cm3mean water mass density results in 93% sensitivity and 93% specificity for the detection of bone marrow edema. • The more severe the bone marrow edema, the higher the mean water density value.

Radiologic findings for prediction of rehabilitation outcomes in patients with chronic symptomatic os subfibulare.

OBJECTIVE: To retrospectively evaluate the radiologic findings for predicting rehabilitation outcomes in patients with chronic symptomatic os subfibulare. MATERIALS AND METHODS: 38 patients with chronic lateral ankle pain and os subfibulare underwent a standardized rehabilitation program. Rehabilitation outcome was evaluated after ≥3 months of intervention as the following: good response group (n = 20) without the need for further treatment and poor response group (n = 18) who underwent surgery after rehabilitation. Size, shape and location of os subfibulare, anterior talofibular ligament abnormality and attachment to the os subfibulare, interposition of fluid signal intensity between the os subfibulare and the fibula, and bone marrow edema in the os subfibulare on radiographs and MRI were evaluated by two radiologists blinded to rehabilitation outcomes and were compared between the two groups. RESULTS: The mean size of os subfibulare was significantly different between good and poor response groups: 7 versus 12 mm (p < 0.01), respectively. The prevalence of interposition of fluid signal intensity between the os subfibulare and the fibula and bone marrow edema in the os subfibulare on MRI was significantly different between the two groups (p < 0.01). However, other radiologic findings were not significantly different (p ≥ 0.08). The optimal size to classify the two groups was 9 mm with sensitivity of 83.3% and specificity of 80%. Inter-observer reliability was good to excellent (ICC = 0.991 for size and kappa = 0.735-0.923 for others). CONCLUSIONS: MRI may be used to predict rehabilitation outcome in patients with chronic symptomatic os subfibulare.

W-DRAG: A joint framework of WGAN with data random augmentation optimized for generative networks for bone marrow edema detection in dual energy CT.

Dual-energy computed tomography (CT) is an excellent substitute for identifying bone marrow edema in magnetic resonance imaging. However, it is rarely used in practice owing to its low contrast. To overcome this problem, we constructed a framework based on deep learning techniques to screen for diseases using axial bone images and to identify the local positions of bone lesions. To address the limited availability of labeled samples, we developed a new generative adversarial network (GAN) that extends expressions beyond conventional augmentation (CA) methods based on geometric transformations. We theoretically and experimentally determined that combining the concepts of data augmentation optimized for GAN training (DAG) and Wasserstein GAN yields a considerably stable generation of synthetic images and effectively aligns their distribution with that of real images, thereby achieving a high degree of similarity. The classification model was trained using real and synthetic samples. Consequently, the GAN technique used in the diagnostic test had an improved F1 score of approximately 7.8% compared with CA. The final F1 score was 80.24%, and the recall and precision were 84.3% and 88.7%, respectively. The results obtained using the augmented samples outperformed those obtained using pure real samples without augmentation. In addition, we adopted explainable AI techniques that leverage a class activation map (CAM) and principal component analysis to facilitate visual analysis of the network's results. The framework was designed to suggest an attention map and scattering plot to visually explain the disease predictions of the network.

Deep Learning-reconstructed Parallel Accelerated Imaging for Knee MRI

BACKGROUND: Deep learning (DL) can improve image quality by removing noise from accelerated MRI. OBJECTIVE: To compare the quality of various accelerated imaging applications in knee MRI with and without DL. METHOD: We analyzed 44 knee MRI scans from 38 adult patients using the DL-reconstructed parallel acquisition technique (PAT) between May 2021 and April 2022. The participants underwent sagittal fat-saturated T2-weighted turbo-spin-echo accelerated imaging without DL (PAT-2 [2-fold parallel accelerated imaging], PAT-3, and PAT-4) and with DL (DL with PAT-3 [PAT-3DL] and PAT-4 [PAT-4DL]). Two readers independently evaluated subjective image quality (diagnostic confidence of knee joint abnormalities, subjective noise and sharpness, and overall image quality) using a 4-point grading system (1-4, 4=best). Objective image quality was assessed based on noise (noise power) and sharpness (edge rise distance). RESULTS: The mean acquisition times for PAT-2, PAT-3, PAT-4, PAT-3DL, and PAT-4DL sequences were 2:55, 2:04, 1:33, 2:04, and 1:33 min, respectively. Regarding subjective image quality, PAT-3DL and PAT-4DL scored higher than PAT-2. Objectively, DL-reconstructed imaging had significantly lower noise than PAT-3 and PAT-4 (P <0.001), but the results were not significantly different from those for PAT-2 (P >0.988). Objective image sharpness did not differ significantly among the imaging combinations (P =0.470). The inter-reader reliability ranged from good to excellent (κ = 0.761­0.832). CONCLUSION: PAT-4DL imaging in knee MRI exhibits similar subjective image quality, objective noise, and sharpness levels compared with conventional PAT-2 imaging, with an acquisition time reduction of 47%.

Metal Artifact Reduction Dual-Energy CT as an Accurate and Reliable Method for Measuring Total Knee Arthroplasty Femoral Component Rotation Compared to Conventional CT.

This article determines the accuracy and reliability of dual-energy computed tomography (DECT) with metal artifact reduction (MAR) in the evaluation of femoral component rotation after total knee arthroplasty (TKA), in comparison with conventional CT images. A total of 49 patients (mean age, 69 years; 42 women) who underwent TKA between January 2019 and March 2020 were retrospectively enrolled. Femoral component rotation, including the anatomic and surgical transepicondylar axes, was evaluated with preoperative conventional CT and postoperative conventional CT and DECT with MAR. Surgical femoral component rotation was also assessed as a reference standard. Accuracy was assessed using paired t-test, and inter- and intraobserver reliability using intraclass correlation coefficients (ICCs) based on postoperative conventional CT and DECT with MAR. Clinical outcomes were evaluated using the Knee Society objective and functional scores. Accuracy of femoral component rotation was not significantly different from that of surgical rotation with both conventional CT and DECT with MAR. However, inter- and intraobserver reliability were better for DECT with MAR (ICC: 0.953-0.966) than for conventional CT (ICC: 0.641-0.749). The Knee Society objective and functional scores improved 1 year postoperatively. CONCLUSION: DECT with MAR showed accurate and more reliable results than did conventional CT in the evaluation of femoral component rotation after TKA.

Recovery After Hip Arthroscopy in Patients With Combined Femoroacetabular Impingement and Labral Tears Compared With Isolated Pathology.

Background: Outcomes for patients with femoroacetabular impingement (FAI) treated with hip arthroscopy can differ depending on whether there is underlying intra-articular pathology. Purpose: To evaluate the outcomes of patients after undergoing hip arthroscopy depending on their underlying pathology (isolated FAI, isolated labral tear, or combined FAI and a labral tear) using the 12-Item International Hip Outcome Tool (iHOT-12). Study Design: Cohort study; Level of evidence, 3. Methods: A total of 75 patients diagnosed with FAI with or without labral tears and isolated labral tears who underwent hip arthroscopy performed by the same surgeon at a single institution from January 2014 to December 2019 were included in this study. All patients had at least 2 years of follow-up data. Patients were divided into 3 groups as follows: patients with FAI and an intact labrum; patients with an isolated labral tear; and patients with combined FAI and a labral tear. The iHOT-12 scores at 1.5, 3, 6, 12, 18, and >24 months postoperatively were compared and analyzed. Outcome scores were also evaluated in terms of the substantial clinical benefit (SCB) and the patient-acceptable symptomatic state (PASS). Results: Of 75 patients who underwent hip arthroscopy, 14 had FAI, 23 had labral tears, and 38 had both. All groups showed significant improvements on the iHOT-12 from preoperative to the final follow-up (FAI, from 37.64 ± 3.77 to 93.64 ± 1.50; labral tear, from 33.70 ± 3.55 to 93 ± 1.24; combined, from 28.55 ± 3.15 to 93.03 ± 0.88) (P < .001 for all). However, compared with other groups, the patients with FAI and a labral tear had lower scores at 1.5, 3, 6, and 12 months postoperatively (P < .001), highlighting a slower rate of recovery. For all groups, recovery to normal function according to the SCB was 100% at 12 months, and satisfaction according to the PASS was 100% at 18 months postoperatively. Conclusion: The final iHOT-12 scores were similar at 18 months regardless of the pathology treated; however, patients with FAI and a labral tear took longer to reach their plateau.

Clinical feasibility of simultaneous multislice acceleration in knee MRI.

PURPOSE: To find the best simultaneous multislice (SMS) accelerated setting for clinical application in knee MRI. MATERIAL AND METHODS: Thirty-three patients (mean age, 54 years; 21 women) who underwent knee MRI (conventional/SMS sequences) between June and October 2020 were enrolled. Two radiologists retrospectively evaluated sagittal T1- and T2-weighted conventional (2-fold parallel acquisition technique [PAT-2]) and SMS (SMS-2 [PAT-2 with 2-fold SMS], SMS-3, and SMS-4) images. For qualitative analysis, artifacts (zebra/residual aliasing) and diagnostic confidence for internal derangement of knee (bone marrow, cartilage, meniscus, anterior cruciate ligament, and synovium abnormalities) were evaluated. For quantitative analysis, contrast-to-noise ratios of bone marrow, meniscus, joint effusion, and ligament were evaluated. RESULTS: Compared to PAT-2 (2 min 32 s), mean acquisition time was reduced by 47% in SMS-2; 64%, SMS-3; and 70%, SMS-4. In qualitative analysis, zebra artifacts were only seen on T2-weighted SMS images. The more SMS was applied, the more zebra and residual aliasing artifacts were seen and the lower diagnostic confidence was for internal derangement. However, qualitative analysis showed acceptable image quality in SMS-2 and SMS-3 images, but not in SMS-4 images. In quantitative analysis, SMS-4 images showed the lowest contrast-to-noise ratios and there were no significant differences among PAT-2, SMS-2, and SMS-3 images. CONCLUSION: Applying SMS-3 to knee MRI reduced scan time and showed acceptable image quality compared to conventional (PAT-2). However, when evaluating SMS images, radiologists should know that when more SMS is applied, more zebra and residual aliasing artifacts appear.

Factors Influencing the Size of a Non-Traumatic Full-Thickness Rotator Cuff Tear: Focusing on Socioeconomic Factors.

This study aimed to identify the risk factors for non-traumatic rotator cuff tears in Korean adult patients who underwent surgical treatment, focusing on socioeconomic factors. A retrospective study was conducted with 659 patients who were diagnosed with a full rotator cuff tear and underwent surgical treatment. The outcome variable was the rotator cuff tear size (mm), as indicated by preoperative magnetic resonance imaging. Socioeconomic variables included occupation, education level, insurance type, and residential area. Univariate analyses were used to evaluate the relation between tear size and independent variables, and multivariate regression was used to estimate the effects of socioeconomic factors on tear size after adjusting for other variables. Significant differences were found in mean tear size according to age, occupation, residence area, and symptom duration (p < 0.05) in multivariate regression analysis. Rural residents had a 2.12 mm larger tear size than urban residents. Compared to National Health Insurance patients, the tear size of Medicaid beneficiaries was significantly larger (6.79 mm) in urban areas. The larger the rotator cuff tear, the greater the risk of retear and poor shoulder function. Therefore, policy efforts are required to expand access to medical care for the vulnerable.

Comparison of a Deep Learning-Based Reconstruction Algorithm with Filtered Back Projection and Iterative Reconstruction Algorithms for Pediatric Abdominopelvic CT.

OBJECTIVE: To compare a deep learning-based reconstruction (DLR) algorithm for pediatric abdominopelvic computed tomography (CT) with filtered back projection (FBP) and iterative reconstruction (IR) algorithms. MATERIALS AND METHODS: Post-contrast abdominopelvic CT scans obtained from 120 pediatric patients (mean age ± standard deviation, 8.7 ± 5.2 years; 60 males) between May 2020 and October 2020 were evaluated in this retrospective study. Images were reconstructed using FBP, a hybrid IR algorithm (ASiR-V) with blending factors of 50% and 100% (AV50 and AV100, respectively), and a DLR algorithm (TrueFidelity) with three strength levels (low, medium, and high). Noise power spectrum (NPS) and edge rise distance (ERD) were used to evaluate noise characteristics and spatial resolution, respectively. Image noise, edge definition, overall image quality, lesion detectability and conspicuity, and artifacts were qualitatively scored by two pediatric radiologists, and the scores of the two reviewers were averaged. A repeated-measures analysis of variance followed by the Bonferroni post-hoc test was used to compare NPS and ERD among the six reconstruction methods. The Friedman rank sum test followed by the Nemenyi-Wilcoxon-Wilcox all-pairs test was used to compare the results of the qualitative visual analysis among the six reconstruction methods. RESULTS: The NPS noise magnitude of AV100 was significantly lower than that of the DLR, whereas the NPS peak of AV100 was significantly higher than that of the high- and medium-strength DLR (p < 0.001). The NPS average spatial frequencies were higher for DLR than for ASiR-V (p < 0.001). ERD was shorter with DLR than with ASiR-V and FBP (p < 0.001). Qualitative visual analysis revealed better overall image quality with high-strength DLR than with ASiR-V (p < 0.001). CONCLUSION: For pediatric abdominopelvic CT, the DLR algorithm may provide improved noise characteristics and better spatial resolution than the hybrid IR algorithm.

Diagnostic performance for detecting bone marrow edema of the hip on dual-energy CT: Deep learning model vs. musculoskeletal physicians and radiologists.

PURPOSE: To compare the diagnostic performance of a deep learning (DL) model with that of musculoskeletal physicians and radiologists for detecting bone marrow edema on dual-energy CT (DECT). METHOD: This retrospective study included adult patients underwent hip DECT and MRI within 1 month between April 2018 and December 2020. A total of 8709 DECT images were divided into training/validation (85%, 7412 augmented images) and test (15%, 1297 images) sets. The images were labeled as present/absent bone marrow edema, with MRI as reference standard. We developed and trained a DL model to detect bone marrow edema from DECT images. Thereafter, DL model, two orthopedic surgeons, and three radiologists evaluated the presence of bone marrow edema on every test image. The diagnostic performance of the DL model and readers was compared. Inter-reader agreement was calculated using Fleiss-kappa statistics. RESULTS: A total of 73 patients (mean age, 59 ± 12 years; 38 female) were included. The DL model had a significantly higher area under the curve (AUC, 0.84 vs. 0.61-0.70, p < 0.001) and sensitivity (79% vs. 29-66%) without loss of specificity (90% vs. 74-93%) than the non- or less-experienced readers and similar to the trained reader (AUC, 0.83, p = 0.402; sensitivity, 71%; specificity, 94%). Additionally, AUCs were strongly dependent on the reader's DECT experience. Inter-reader agreement was fair (κ = 0.303). CONCLUSION: The DL model showed better diagnostic performance than less-experienced physicians in detecting bone marrow edema on DECT and comparable performance to a trained radiologist.

Deep Learning-Enhanced Parallel Imaging and Simultaneous Multislice Acceleration Reconstruction in Knee MRI.

OBJECTIVES: This study aimed to examine various combinations of parallel imaging (PI) and simultaneous multislice (SMS) acceleration imaging using deep learning (DL)-enhanced and conventional reconstruction. The study also aimed at comparing the diagnostic performance of the various combinations in internal knee derangement and provided a quantitative evaluation of image sharpness and noise using edge rise distance (ERD) and noise power (NP), respectively. MATERIALS AND METHODS: The data from adult patients who underwent knee magnetic resonance imaging using various DL-enhanced acquisitions between June 2021 and January 2022 were retrospectively analyzed. The participants underwent conventional 2-fold PI and DL protocols with 4- to 8-fold acceleration imaging (P2S2 [2-fold PI with 2-fold SMS], P3S2, and P4S2). Three readers evaluated the internal knee derangement and the overall image quality. The diagnostic performance was calculated using consensus reading as a standard reference, and we conducted comparative evaluations. We calculated the ERD and NP for quantitative evaluations of image sharpness and noise, respectively. Interreader and intermethod agreements were calculated using Fleiss κ. RESULTS: A total of 33 patients (mean age, 49 ± 19 years; 20 women) were included in this study. The diagnostic performance for internal knee derangement and the overall image quality were similar among the evaluated protocols. The NP values were significantly lower using the DL protocols than with conventional imaging ( P < 0.001), whereas the ERD values were similar among these methods ( P > 0.12). Interreader and intermethod agreements were moderate-to-excellent (κ = 0.574-0.838) and good-to-excellent (κ = 0.755-1.000), respectively. In addition, the mean acquisition time was reduced by 47% when using DL with P2S2, by 62% with P3S2, and by 71% with P4S2, compared with conventional P2 imaging (2 minutes and 55 seconds). CONCLUSIONS: The combined use of DL-enhanced 8-fold acceleration imaging (4-fold PI with 2-fold SMS) showed comparable performance with conventional 2-fold PI for the evaluation of internal knee derangement, with a 71% reduction in acquisition time.

Cardiac CTA image quality of adaptive statistical iterative reconstruction-V versus deep learning reconstruction "TrueFidelity" in children with congenital heart disease.

BACKGROUND: Several recent studies have reported that deep learning reconstruction "TrueFidelity" (TF) improves computed tomography (CT) image quality. However, no study has compared adaptive statistical repeated reconstruction (ASIR-V) using TF in pediatric cardiac CT angiography (CTA) with a low peak kilovoltage. OBJECTIVE: This study aimed to determine whether ASIR-V or TF CTA image quality is superior in children with congenital heart disease (CHD). MATERIALS AND METHODS: Fifty children (median age, 2 months; interquartile range, 0-5 months; 28 men) with CHD who underwent CTA were enrolled between June and September 2020. Images were reconstructed using 2 ASIR-V blending factors (80% and 100% [AV-100]) and 3 TF settings (low, medium, and high [TF-H] strength levels). For the quantitative analyses, 3 objective image qualities (attenuation, noise, and signal-to-noise ratio [SNR]) were measured of the great vessels and heart chambers. The contrast-to-noise ratio (CNR) was also evaluated between the left ventricle and the dial wall. For the qualitative analyses, the degree of quantum mottle and blurring at the upper level to the first branch of the main pulmonary artery was assessed independently by 2 radiologists. RESULTS: When the ASIR-V blending factor level and TF strength were higher, the noise was lower, and the SNR was higher. The image noise and SNR of TF-H were significantly lower and higher than those of AV-100 (P < .01), except for noise in the right atrium and left pulmonary artery and SNR of the right ventricle. Regarding CNR, TF-H was significantly better than AV-100 (P < .01). In addition, in the objective assessment of the degree of quantum mottle and blurring, TF-H had the best score among all examined image sets (P < .01). CONCLUSION: TF-H is superior to AV-100 in terms of objective and subjective image quality. Consequently, TF-H was the best image set for cardiac CTA in children with CHD.

Erratum: Transiting Nerve Rootlet Abnormalities on MRI after Lumbar Laminectomy: Associations with Persistent Postoperative Pain.

This corrects the article on p. 225 in vol. 22, PMID: 32901462.