Deep Learning Algorithm for Simultaneous Noise Reduction and Edge Sharpening in Low-Dose CT Images: A Pilot Study Using Lumbar Spine CT.

OBJECTIVE: The purpose of this study was to assess whether a deep learning (DL) algorithm could enable simultaneous noise reduction and edge sharpening in low-dose lumbar spine CT. MATERIALS AND METHODS: This retrospective study included 52 patients (26 male and 26 female; median age, 60.5 years) who had undergone CT-guided lumbar bone biopsy between October 2015 and April 2020. Initial 100-mAs survey images and 50-mAs intraprocedural images were reconstructed by filtered back projection. Denoising was performed using a vendor-agnostic DL model (ClariCT.AI™, ClariPI) for the 50-mAS images, and the 50-mAs, denoised 50-mAs, and 100-mAs CT images were compared. Noise, signal-to-noise ratio (SNR), and edge rise distance (ERD) for image sharpness were measured. The data were summarized as the mean ± standard deviation for these parameters. Two musculoskeletal radiologists assessed the visibility of the normal anatomical structures. RESULTS: Noise was lower in the denoised 50-mAs images (36.38 ± 7.03 Hounsfield unit [HU]) than the 50-mAs (93.33 ± 25.36 HU) and 100-mAs (63.33 ± 16.09 HU) images (p < 0.001). The SNRs for the images in descending order were as follows: denoised 50-mAs (1.46 ± 0.54), 100-mAs (0.99 ± 0.34), and 50-mAs (0.58 ± 0.18) images (p < 0.001). The denoised 50-mAs images had better edge sharpness than the 100-mAs images at the vertebral body (ERD; 0.94 ± 0.2 mm vs. 1.05 ± 0.24 mm, p = 0.036) and the psoas (ERD; 0.42 ± 0.09 mm vs. 0.50 ± 0.12 mm, p = 0.002). The denoised 50-mAs images significantly improved the visualization of the normal anatomical structures (p < 0.001). CONCLUSION: DL-based reconstruction may enable simultaneous noise reduction and improvement in image quality with the preservation of edge sharpness on low-dose lumbar spine CT. Investigations on further radiation dose reduction and the clinical applicability of this technique are warranted.

Dual-Layer Detector CT With Virtual Noncalcium Imaging: Diagnostic Performance in Patients With Suspected Wrist Fractures.

OBJECTIVE. The purpose of this study was to evaluate the diagnostic utility of dual-layer CT (DLCT) for evaluating wrist injuries and to compare it with MRI. MATERIALS AND METHODS. The cases of 62 patients with suspected wrist fractures who underwent imaging with both DLCT and MRI from January 2018 through February 2019 were retrospectively reviewed. By means of a calcium suppression algorithm, virtual noncalcium (VNCa) image reconstruction was performed, and the images were reviewed by two readers to identify fractures, bone contusions, and nontraumatic lesions in the radius, ulna, and carpal bones. Sensitivity, specificity, PPV, and NPV were calculated and compared between standard CT and VNCa images with a combination of standard CT and MRI as the reference standard. RESULTS. Use of DLCT with VNCa reconstruction increased the sensitivity of diagnosis of fractures in the radius and carpal bones over that of standard CT alone; occult fractures were detected that were not seen with standard CT. The sensitivity and specificity for detecting radius fracture were 98.1% and 93.8% for DLCT and 96.3% and 93.8% for standard CT. For detecting carpal bone fracture, sensitivity and specificity were 100% and 98.9% for DLCT and 93.8% and 100% for standard CT. VNCa reconstruction also had good diagnostic accuracy with regard to diagnosing nonfracture bone contusions in carpal bones. The accuracy was comparable to that of MRI with sensitivity of 92.9% and specificity of 94.5%. Interreader agreement in interpreting VNCa images was generally good to excellent. CONCLUSION. DLCT with VNCa reconstruction is a promising tool for identifying occult wrist fractures and nonfracture contusion injuries in patients with wrist trauma.

Relationship between cerebral microbleeds and white matter MR hyperintensities in systemic lupus erythematosus: a retrospective observational study.

PURPOSE: White matter hyperintensities (WMH) and cerebral microbleeds (CMBs) are known to be associated with small vessel diseases (SVD) and neuroinflammation. The purpose was to investigate the relationship between CMBs and WMH in patients with systemic lupus erythematosus (SLE). METHODS: Thirty-one SLE patients with WMH and 27 SLE patients with normal brain MRI were compared. The presence, location, and grading of CMBs were assessed using susceptibility-weighted images. WMH volume was quantitatively measured. Clinical characteristics and serologic markers were compared. We also performed two separate subgroup analyses after (1) dividing WMH into inflammatory lesion vs. SVD subgroups and (2) dividing WMH into those with vs. without CMB subgroups. RESULTS: The WMH group showed more frequent CMBs than the normal MR group (p < 0.001). The WMH group showed higher SLE disease activity index, longer disease duration, and a higher incidence of antiphospholipid syndrome than the normal MR group (p = 0.02, 0.04, and 0.04, respectively). There was a moderate correlation between WMH volume and CMB grading (r = 0.49, p = 0.006). Within the WMH group, the inflammatory lesion subgroup showed more frequent CMBs and larger WMH volume than the SVD subgroup (p < 0.001 and 0.02, respectively). The WMH with CMB subgroup had larger WMH volume than the WMH without CMB subgroup (p = 0.004). CONCLUSION: In patients with SLE, CMBs could be related to large-volume WMH and inflammatory lesions. CMBs along with severe WMH could be used as an imaging biomarker of vasculitis in patients with SLE.