Deep learning-based reconstruction for 3-dimensional heavily T2-weighted fat-saturated magnetic resonance (MR) myelography in epidural fluid detection: image quality and diagnostic performance.

Background: Heavily T2-weighted fat-saturated (HT2W-FS) magnetic resonance myelography (MRM) is useful for diagnosing the cause of intracranial hypotension. Recently, deep learning-based reconstruction (DLR) has been utilized to improve image signal-to-noise ratios and sharpness while reducing artifacts, all without lengthening acquisition times. This study aimed to compare the diagnostic performance and image quality of conventional reconstruction (CR) and DLR of 3-dimensional (3D) HT2W-FS MRM applied to detecting epidural fluid in patients with clinically suspected intracranial hypotension. Methods: This retrospective study included 21 magnetic resonance myelography examinations using both CR and DLR in 21 patients who experienced orthostatic headache between April 2021 and September 2022. Quantitative image quality evaluation was performed by comparing signal-to-noise ratios at the lower thoracic levels. The image quality and artifacts were graded by three readers. The presence of epidural fluid was reported with a confidence score by two readers, and the area under the receiver operating curve, interobserver agreement, and inter-image-set agreement were evaluated. The conspicuity of the dura mater where the epidural fluid was detected was also investigated. Results: Quantitative and subjective image quality, and artifacts significantly improved with DLR (all P<0.001). Diagnostic performance of DLR was higher for both readers [reader 1: area under the curve (AUC) of CR =0.929; 95% confidence interval (CI): 0.902-0.950, AUC of DLR =0.965 (95% CI: 0.944-0.979), P=0.007; reader 2: AUC of CR =0.834 (95% CI: 0.798-0.866), AUC of DLR =0.877 (0.844-0.905), P=0.040]. Correlation with standard care of MRM in CR and DLR were both strong in reader 1 (rho =0.868-0.919, P<0.001), but was respectively strong and moderate in reader 2 (rho =0.734-0.805, P<0.001). Interobserver agreement was substantial (κ=0.708-0.762). The inter-image-set agreement was almost perfect for reader 1 (κ=0.907) and was substantial for reader 2 (κ=0.750). Dura mater conspicuity significantly improved with DLR (P<0.014, reader 1; P<0.001, readers 2 and 3). Conclusions: HT2W-FS magnetic resonance myelography with DLR demonstrates substantial improvements in image quality and may improve confidence in detecting epidural fluid.

Thin-slice elbow MRI with deep learning reconstruction: Superior diagnostic performance of elbow ligament pathologies.

PURPOSE: With the slice thickness routinely used in elbow MRI, small or subtle lesions may be overlooked or misinterpreted as insignificant. To compare 1 mm slice thickness MRI (1 mm MRI) with deep learning reconstruction (DLR) to 3 mm slice thickness MRI (3 mm MRI) without/with DLR, and 1 mm MRI without DLR regarding image quality and diagnostic performance for elbow tendons and ligaments. METHODS: This retrospective study included 53 patients between February 2021 and January 2022, who underwent 3 T elbow MRI, including T2-weighted fat-saturated coronal 3 mm and 1 mm MRI without/with DLR. Two radiologists independently assessed four MRI scans for image quality and artefacts, and identified the pathologies of the five elbow tendons and ligaments. In 19 patients underwent elbow surgery after elbow MRI, diagnostic performance was evaluated using surgical records as a reference standard. RESULTS: For both readers, 3 mm MRI with DLR had significant higher image quality scores than 3 mm MRI without DLR and 1 mm MRI with DLR (all P < 0.01). For common extensor tendon and elbow ligament pathologies, 1 mm MRI with DLR showed the highest number of pathologies for both readers. The 1 mm MRI with DLR had the highest kappa values for all tendons and ligaments. For reader 1, 1 mm MRI with DLR showed superior diagnostic performance than 3 mm MRI without/with DLR. For reader 2, 1 mm MRI with DLR showed the highest diagnostic performance; however, there was no significant difference. CONCLUSIONS: One mm MRI with DLR showed the highest diagnostic performance for evaluating elbow tendon and ligament pathologies, with similar subjective image qualities and artefacts.

Sarcopenia prediction using shear-wave elastography, grayscale ultrasonography, and clinical information with machine learning fusion techniques: feature-level fusion vs. score-level fusion.

This study aimed to develop and evaluate a sarcopenia prediction model by fusing numerical features from shear-wave elastography (SWE) and gray-scale ultrasonography (GSU) examinations, using the rectus femoris muscle (RF) and categorical/numerical features related to clinical information. Both cohorts (development, 70 healthy subjects; evaluation, 81 patients) underwent ultrasonography (SWE and GSU) and computed tomography. Sarcopenia was determined using skeletal muscle index calculated from the computed tomography. Clinical and ultrasonography measurements were used to predict sarcopenia based on a linear regression model with the least absolute shrinkage and selection operator (LASSO) regularization. Furthermore, clinical and ultrasonography features were combined at the feature and score levels to improve sarcopenia prediction performance. The accuracies of LASSO were 70.57 ± 5.00-81.54 ± 4.83 (clinical) and 69.00 ± 4.52-69.73 ± 5.47 (ultrasonography). Feature-level fusion of clinical and ultrasonography (accuracy, 70.29 ± 6.63 and 83.55 ± 4.32) showed similar performance with clinical features. Score-level fusion by AdaBoost showed the best performance (accuracy, 73.43 ± 6.57-83.17 ± 5.51) in the development and evaluation cohorts, respectively. This study might suggest the potential of machine learning fusion techniques to enhance the accuracy of sarcopenia prediction models and improve clinical decision-making in patients with sarcopenia.

Impact of deep learning-based reconstruction and anti-peristaltic agent on the image quality and diagnostic performance of magnetic resonance enterography comparing single breath-hold single-shot fast spin echo with and without anti-peristaltic agent.

Background: While anti-peristaltic agents are beneficial for high quality magnetic resonance enterography (MRE), their use is constrained by potential side effects and increased examination complexity. We explored the potential of deep learning-based reconstruction (DLR) to compensate for the absence of anti-peristaltic agent, improve image quality and reduce artifact. This study aimed to evaluate the need for an anti-peristaltic agent in single breath-hold single-shot fast spin-echo (SSFSE) MRE and compare the image quality and artifacts between conventional reconstruction (CR) and DLR. Methods: We included 45 patients who underwent MRE for Crohn's disease between October 2021 and September 2022. Coronal SSFSE images without fat saturation were acquired before and after anti-peristaltic agent administration. Four sets of data were generated: SSFSE CR with and without an anti-peristaltic agent (CR-A and CR-NA, respectively) and SSFSE DLR with and without an anti-peristaltic agent (DLR-A and DLR-NA, respectively). Two radiologists independently reviewed the images for overall quality and artifacts, and compared the three images with DLR-A. The degree of distension and inflammatory parameters were scored on a 5-point scale in the jejunum and ileum, respectively. Signal-to-noise ratio (SNR) levels were calculated in superior mesenteric artery (SMA) and iliac bifurcation level. Results: In terms of overall quality, DLR-NA demonstrated no significant difference compared to DLR-A, whereas CR-NA and CR-A demonstrated significant differences (P<0.05, both readers). Regarding overall artifacts, reader 1 rated DLR-A slightly better than DLR-NA in four cases and rated them as identical in 41 cases (P=0.046), whereas reader 2 demonstrated no difference. Bowel distension was significantly different in the jejunum (Reader 1: P=0.046; Reader 2: P=0.008) but not in the ileum. Agreements between the images (Reader 1: ĸ=0.73-1.00; Reader 2: ĸ=1.00) and readers (ĸ=0.66 for all comparisons) on inflammation were considered good to excellent. The sensitivity, specificity, and accuracy in diagnosing inflammation in the terminal ileum were the same among DLR-NA, DLR-A, CR-NA and CR-A (94.42%, 81.83%, and 89.69 %; and 83.33%, 90.91%, and 86.21% for Readers 1 and 2, respectively). In both SMA and iliac bifurcation levels, SNR of DLR images exhibited no significant differences. CR images showed significantly lower SNR compared with DLR images (P<0.001). Conclusions: SSFSE without anti-peristaltic agents demonstrated nearly equivalent quality to that with anti-peristaltic agents. Omitting anti-peristaltic agents before SSFSE and adding DLR could improve the scanning outcomes and reduce time.

Hounsfield Units Predict Survival of Patients With Estrogen Receptor-Positive and Human Epithelial Growth Factor Receptor 2-Negative Breast Cancer.

BACKGROUNDS: Tumor vascularity plays a fundamental role in cancer progression, including breast cancer. This study aimed to elucidate tumor vascularity and its impact on patient survival in the context of breast cancer subtypes using Hounsfield units (HU) on contrast-enhanced computed tomography (CT). MATERIALS AND METHODS: Patients with early-stage breast cancer who completed planned treatment between 2003 and 2013 were retrospectively assessed. RESULTS: The final cohort comprised 440 patients. Of the 440 patients, 262 had estrogen receptor (ER)-positive disease and 119 had human epidermal growth factor receptor 2 (HER2)-overexpressing disease. The tumor-to-aorta ratio of Hounsfield units (TAR) was related to significantly worse recurrence-free interval (RFI) (P < .001) and overall survival (OS) (P < .001) in patients with TAR > 0.33 for RFI and > 0.35 for OS than their counterparts. In the subgroup analysis, the survival disadvantage was limited only to patients with ER-positive and HER2-negative disease (P < .001 for both RFI and OS). CONCLUSION: This study showed that TAR, which reflects tumor vascularity, was significantly related to patients' RFI and OS, suggesting its capacity as a feasible biomarker. This study also showed that TAR was associated with the survival in patients with ER-positive and HER2-negative disease.

Ultrasound findings of subpial hemorrhage in neonates.

PURPOSE: Subpial hemorrhage (SPH) is a subtype of intracranial hemorrhage characterized by damage to the adjacent brain parenchyma. The aim of this study was to describe the sonographic features of SPH in neonates. METHODS: The cranial ultrasound (US) findings of neonates with SPH confirmed by brain magnetic resonance imaging (MRI) were analyzed retrospectively. Initial and follow-up US and MRI scans were reviewed by two pediatric radiologists who were blinded to both clinical history and outcomes. The US features were compared with the MRI findings. RESULTS: Sixteen patients were included (median gestational age, 38 weeks; range, 26 to 40 weeks; 69% term). SPH was detected most often in the temporal lobe (63%), and multiple SPHs were found in seven of 16 neonates, based on MRI. Acute SPH with an underlying venous infarct (UVI) was detected on US in 15 of 16 patients: small or large fan-shaped hyperechoic lesions (n=7 and 4, respectively) and gyriform hyperechoic lesions (n=4). The sonographic yin-yang sign was observed in three of the four large fan-shaped SPH cases. The accompanying findings on US were intraventricular hemorrhage (four out of six MRI-confirmed cases), and concurrent periventricular venous infarcts (five out of nine MRI-confirmed cases). In five patients, subpial cysts were observed on follow-up US or MRI (n=4 and n=4, respectively). CONCLUSION: Acute SPH with UVI can appear as a peripheral fan-shaped or gyriform hyperechoic lesion on cranial US. SPH can be detected and suspected based on the US features of SPH with the accompanying findings.

Comparison of deep learning-based reconstruction of PROPELLER Shoulder MRI with conventional reconstruction.

OBJECTIVE: To compare the image quality and agreement among conventional and accelerated periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) MRI with both conventional reconstruction (CR) and deep learning-based reconstruction (DLR) methods for evaluation of shoulder. MATERIALS AND METHODS: We included patients who underwent conventional (acquisition time, 8 min) and accelerated (acquisition time, 4 min and 24 s; 45% reduction) PROPELLER shoulder MRI using both CR and DLR methods between February 2021 and February 2022 on a 3 T MRI system. Quantitative evaluation was performed by calculating the signal-to-noise ratio (SNR). Two musculoskeletal radiologists compared the image quality using conventional sequence with CR as the reference standard. Interobserver agreement between image sets for evaluating shoulder was analyzed using weighted/unweighted kappa statistics. RESULTS: Ninety-two patients with 100 shoulder MRI scans were included. Conventional sequence with DLR had the highest SNR (P < .001), followed by accelerated sequence with DLR, conventional sequence with CR, and accelerated sequence with CR. Comparison of image quality by both readers revealed that conventional sequence with DLR (P = .003 and P < .001) and accelerated sequence with DLR (P = .016 and P < .001) had better image quality than the conventional sequence with CR. Interobserver agreement was substantial to almost perfect for detecting shoulder abnormalities (κ = 0.600-0.884). Agreement between the image sets was substantial to almost perfect (κ = 0.691-1). CONCLUSION: Accelerated PROPELLER with DLR showed even better image quality than conventional PROPELLER with CR and interobserver agreement for shoulder pathologies comparable to that of conventional PROPELLER with CR, despite the shorter scan time.

Comparison of shear wave elastography with gray-scale USG and CT for quantitative evaluation of rectus femoris muscle.

PURPOSE: The purpose of this study was to compare the measurement of shear wave elastography (SWE) and gray scale ultrasonography (GSU) and CT attenuation of mid-rectus femoris (RF) muscle in healthy adults. METHODS: This prospective study included 70 participants with a healthy body mass index (<25 kg/m2 ) between June 2019 and January 2020. Echo intensity (EI) grading of RF on GSU was performed. SWE was performed for the three levels of the RF. Measurements were repeated 10 min after the first measurement. The mid-RF attenuation on CT was also measured. Interobserver agreement of EI grade among three readers was assessed using weighted-kappa statistics. The reliability of SWE was assessed using intraclass correlation coefficient. The correlations between the SWE and CT/GSU measurements were analyzed. RESULTS: Interobserver agreement of EI grade on GSU by the three radiologists was moderate to substantial (k = 0.562-0.767). The inter-session agreements for SWE were almost perfect for mid RF (k = 0.822-0.829) and substantial for proximal and distal RF (k = 0.767-0.795). There were significant correlations between SWE-EI and SWE-CT attenuation (p < 0.001, respectively) at the mid-RF. CONCLUSIONS: SWE measurements on mid-RF demonstrated the highest reliability. SWE parameters showed a strong correlation with EI on GSU and attenuation on CT.

Comparing the clinical application of strain elastography and shear wave elastography for the evaluation of lateral epicondylitis.

PURPOSE: This study aimed to compare the diagnostic performance of gray-scale ultrasound (GSU), strain elastography (SE), and shear wave elastography (SWE) for lateral epicondylitis (LE). METHODS: From November 2018 to March 2021, 87 common extensor tendons (CETs) of 70 patients were evaluated using elbow ultrasound. Patients were divided into two groups: patients with LE and those without lateral elbow pain. GSU, SE, and SWE findings from the two groups were compared, and the diagnostic performance was compared for significant variables. RESULTS: Among the 70 patients, 41 with 44 CETs were clinically diagnosed with LE, and 29 with 43 CETs had no lateral elbow pain. Hypoechogenicity, calcification, and hyperemia were significantly different. Strain ratio (SR), stiffness, and shear wave velocity (SWV) were also significantly different. The diagnostic accuracy of stiffness and SWV was better than that of other significant findings. Furthermore, there were no superiorities in SR, stiffness, or SWV. CONCLUSION: Both SE and SWE can be valuable diagnostic tools for diagnosing LE. The diagnostic performances of both SE and SWE were similar in the detection of LE.

Shear-wave elastography evaluation of thenar muscle in carpal tunnel syndrome.

PURPOSE: We aim to evaluate the shear wave velocity (SWV) of the thenar muscle as an adjunct diagnostic tool for carpal tunnel syndrome (CTS). METHODS: Ninety-two wrists with CTS and 30 control wrists without CTS underwent ultrasonographic evaluation of thenar muscle and median nerve including shear-wave elastography. Cross sectional area (CSA) of medial nerve and SWV of thenar muscle and median nerve were evaluated. CTS patients were assessed for Boston CTS, Padua CTS, modified Hirani grading scores, and nerve conduction study (NCS). SWVs, CSA, and NCS parameters were compared between two groups. RESULTS: The SWVs of thenar muscle and median nerve (p < 0.001, respectively), and CSA of median nerve (p < 0.001) were more significantly greater in patients with CTS than in controls. The SWV of median nerve was moderately correlated with CSA of median nerve (r = 0.35, p < 0.001) and modified Hirani CTS score (r = 0.35, p < 0.001). The SWV of thenar muscle was inversely correlated with modified Hirani CTS score (r = -0.21, p = 0.04). CONCLUSION: The SWV of thenar muscle and median nerve of CTS were significantly increased compared to that of control, and significantly negatively correlated with NCS parameters (modified Hirani CTS score). SWVs may be used as an adjunct diagnostic tool for CTS.