Development of an individual display optimization system based on deep convolutional neural network transition learning for somatostatin receptor scintigraphy.

Somatostatin receptor scintigraphy (SRS) is an essential examination for the diagnosis of neuroendocrine tumors (NETs). This study developed a method to individually optimize the display of whole-body SRS images using a deep convolutional neural network (DCNN) reconstructed by transfer learning of a DCNN constructed using Gallium-67 (67Ga) images. The initial DCNN was constructed using U-Net to optimize the display of 67Ga images (493 cases/986 images), and a DCNN with transposed weight coefficients was reconstructed for the optimization of whole-body SRS images (133 cases/266 images). A DCNN was constructed for each observer using reference display conditions estimated in advance. Furthermore, to eliminate information loss in the original image, a grayscale linear process is performed based on the DCNN output image to obtain the final linearly corrected DCNN (LcDCNN) image. To verify the usefulness of the proposed method, an observer study using a paired-comparison method was conducted on the original, reference, and LcDCNN images of 15 cases with 30 images. The paired comparison method showed that in most cases (29/30), the LcDCNN images were significantly superior to the original images in terms of display conditions. When comparing the LcDCNN and reference images, the number of LcDCNN and reference images that were superior to each other in the display condition was 17 and 13, respectively, and in both cases, 6 of these images showed statistically significant differences. The optimized SRS images obtained using the proposed method, while reflecting the observer's preference, were superior to the conventional manually adjusted images.

Resting-state functional magnetic resonance imaging indices are related to electrophysiological dysfunction in degenerative cervical myelopathy.

The age-related degenerative pathologies of the cervical spinal column that comprise degenerative cervical myelopathy (DCM) cause myelopathy due spinal cord compression. Functional neurological assessment of DCM can potentially reveal the severity and pathological mechanism of DCM. However, functional assessment by conventional MRI remains difficult. This study used resting-state functional MRI (rs-fMRI) to investigate the relationship between functional connectivity (FC) strength and neurophysiological indices and examined the feasibility of functional assessment by FC for DCM. Preoperatively, 34 patients with DCM underwent rs-fMRI scans. Preoperative central motor conduction time (CMCT) reflecting motor functional disability and intraoperative somatosensory evoked potentials (SEP) reflecting sensory functional disability were recorded as electrophysiological indices of severity of the cervical spinal cord impairment. We performed seed-to-voxel FC analysis and correlation analyses between FC strength and the two electrophysiological indices. We found that FC strength between the primary motor cortex and the precuneus correlated significantly positively with CMCT, and that between the lateral part of the sensorimotor cortex and the lateral occipital cortex also showed a significantly positive correlation with SEP amplitudes. These results suggest that we can evaluate neurological and electrophysiological severity in patients with DCM by analyzing FC strengths between certain brain regions.

Impact of fat on the apparent T1 value of the liver: assessment by water-only derived T1 mapping.

OBJECTIVES: To determine the impact of fat on the apparent T1 value of the liver using water-only derived T1 mapping. METHODS: 3-T MRI included 2D Look-Locker T1 mapping and proton density fat fraction (PDFF) mapping. T1 values of the liver were compared among T1 maps obtained by in-phase (IP), opposed-phase (OP), and Dixon water sequences using paired t-test. The correlation between T1 values of the liver on each T1 map and PDFF was assessed using Spearman correlation coefficient. The absolute differences between T1 value of the liver on Dixon water images and that on IP or OP images were also correlated with PDFF. RESULTS: One hundred sixty-two patients (median age, 70 [range, 24-91] years, 90 men) were retrospectively evaluated. The T1 values of the liver on each T1 map were significantly different (p < 0.001). The T1 value of the liver on IP images was significantly negatively correlated with PDFF (r = - 0.438), while the T1 value of the liver on OP images was slightly positively correlated with PDFF (r = 0.164). The T1 value of the liver on Dixon water images was slightly negatively correlated with PDFF (r = - 0.171). The absolute differences between T1 value of the liver on Dixon water images and that on IP or OP images were significantly correlated with PDFF (r = 0.606, 0.722; p < 0.001). CONCLUSION: Fat correction for the apparent T1 value by water-only derived T1 maps will be helpful for accurately evaluating the T1 value of the liver. CLINICAL RELEVANCE STATEMENT: Fat-corrected T1 mapping of the liver with the water component only obtained from the 2D Dixon Look-Locker sequence could be useful for accurately evaluating the T1 value of the liver without the impact of fat in daily clinical practice. KEY POINTS: • The T1 values of the liver on the conventional T1 maps are significantly affected by the presence of fat. • The apparent T1 value of the liver on water-only derived T1 maps would be slightly impacted by the presence of fat. • Fat correction for the apparent T1 values is necessary for the accurate assessment of the T1 values of the liver.

The pancreatic exocrine function in patients with pancreatic endocrine insufficiency: the evaluation with cine-dynamic magnetic resonance cholangiopancreatography using a spatially selective inversion-recovery pulse and T1 mapping.

PURPOSE: To evaluate the association of the pancreatic exocrine function estimated by cine-dynamic magnetic resonance cholangiopancreatography (MRCP) using a spatially selective inversion-recovery (IR) pulse with the pancreatic endocrine function estimated by the T1 relaxation time of the pancreatic parenchyma and HbA1c values. MATERIALS AND METHODS: Forty-three patients with suspected hepatobiliary or pancreatic diseases were included. Patients were classified into three groups: HbA1c < 5.7% (normal group), 5.7% ≤ HbA1c < 6.5% (prediabetes group), and HbA1c ≥ 6.5% (diabetes group). The frequency of the secretory flow of the pancreatic juice was observed within the area of the IR pulse, and the moving distance (mean secretion grade) of the pancreatic juice inflow within the area of the IR pulse on cine-dynamic MRCP, and the T1 relaxation time of the pancreatic parenchyma on the T1 map images were assessed. The MR imaging measurements were compared using Spearman's rank correlation coefficient analysis and the Kruskal-Wallis and Mann-Whitney U tests. RESULTS: Both the mean secretion grade and frequency of the pancreatic secretory inflow had a significant negative correlation with the T1 relaxation time of the pancreatic parenchyma (r = - 0.335, p = 0.028 and r = - 0.305, p = 0.047, respectively) and HbA1c values (r = - 0.308, p = 0.044 and r = - 0.313, p = 0.041, respectively). Both the mean secretion grade and frequency of the pancreatic secretory inflow in the elevated HbA1c (prediabetes and diabetes) group were significantly lower than those in the normal group (p = 0.030 and p = 0.029, respectively). CONCLUSION: The pancreatic exocrine function estimated by cine-dynamic MRCP was significantly lower in patients with prediabetes and diabetes than in controls. Cine-dynamic MRCP with a spatially selective IR pulse may be useful for the early diagnosis of pancreatic exocrine insufficiency in patients with pancreatic endocrine insufficiency.

Feasibility of high-resolution magnetic resonance imaging of the liver using deep learning reconstruction based on the deep learning denoising technique.

PURPOSE: To evaluate the feasibility of High-resolution (HR) magnetic resonance imaging (MRI) of the liver using deep learning reconstruction (DLR) based on a deep learning denoising technique compared with standard-resolution (SR) imaging. MATERIALS AND METHODS: This retrospective study included patients who underwent abdominal MRI including both HR imaging using DLR and SR imaging between April 1 and August 31, 2019. DLR was applied to all HR images using 12 different strength levels of noise reduction to determine the optimal denoised level for HR images. The mean signal-to-noise ratio (SNR) was then compared between the original HR images without DLR and the optimal denoised HR images with DLR and SR images. The mean image noise, sharpness and overall image quality were also compared. Statistical analyses were performed with the Friedman and Dunn-Bonferroni post-hoc test. RESULTS: In total, 49 patients were analyzed (median age, 71 years; 25 women). In quantitative analysis, the mean SNRs on the original HR images without DLR were significantly lower than those on the SR images in all sequences (p < 0.01). Conversely, the mean SNRs on optimal denoised HR images were significantly higher than those on the SR images in all sequences (p < 0.01). In the qualitative analysis, the mean scores for the image noise and overall image quality were significantly higher on optimal denoised HR images than on the SR images in all sequences (p < 0.01) except for the mean image noise score in in-phase (IP) images. CONCLUSIONS: The use of a deep learning-based noise reduction technique substantially and successfully improved the SNR and image quality in HR imaging of the liver. Denoised HR imaging using the DLR technique appears feasible for use in liver MR examinations compared with SR imaging.

Automatic ROI construction for analyzing time-signal intensity curve in dynamic contrast-enhanced MR imaging of the breast.

Our purpose in this study was to construct a 3-dimensional (3D) region of interest (ROI) for analyzing the time-signal intensity curve (TIC) semi-automatically in dynamic contrast-enhanced magnetic resonance (DCE-MR) imaging of the breast. DCE-MR breast imaging datasets were acquired by a 3.0-Tesla MR system with the use of a 3D fast gradient echo sequence. The essential idea in the new method was to analyze each pixel and to construct an ROI made up of pixels with similar TICs. First, an analyst selected a starting point in the contrast media-enhanced tumor. Second, we calculated Pearson's correlation coefficients (CCs) between the TIC in the starting coordinate selected by the analyst and the TIC in the other coordinates. Third, ROI pixels were selected if their CC threshold satisfied a level of coefficient variation of the ROI determined by prior research performed in our institution. We made a retrospective review of patients who underwent breast DCE-MR examination for pre-operative diagnosis. To confirm the feasibility of the resulting 3D-ROI from TIC analysis, we compared Fischer's score obtained from 3D-ROI by applying a new method to a score obtained from a manually selected 2-dimensional (2D) ROI which was used during routine clinical examination. The Fischer's scores obtained from both the automatically selected 3D-ROI and the manually selected 2D-ROI showed almost equivalent results. Thus, we considered that the new method was comparable to the conventional method. Furthermore, the new method has the potential to be used for evaluation of the extent of tumors.

Effectiveness of using low-dose computed tomography to assess patency in gastrointestinal tracts with a patency capsule.

BACKGROUND/AIMS: A patency capsule (PC) is used to safely perform capsule endoscopy. When the PC is not excreted within the defined time frame, radiography often cannot localize the PC. Computed tomography (CT) localizes a PC more definitively than radiography. We evaluated the localization of PCs using low-dose (LD)-CT. METHODOLOGY: Forty-nine patients received a PC and 33 did not excrete the PC within the defined time frame and underwent radiography and LDCT with a 90% exposure reduction. RESULTS: LDCT localized the PC in 31 patients (93.9%), whereas radiography localized it in 7 (21.2%), indicating a significantly higher detection rate with LDCT (P<0.0001). PC retention in the small intestine was confirmed by LDCT in 4 patients. Retention was not observed during capsule endoscopy in the patients with confirmed patency of the gastrointestinal tract. In 21 patients who underwent LDCT with the conventional photographing area, the effective radiographic exposure dose associated with LDCT was 1.43±1.08 mSv. Conversely, in the 12 patients who underwent LDCT with a reduced photographing area, the effective dose was reduced to 0.62±0.27 mSv (P<0.05). CONCLUSION: LDCT with a reduced exposure dose can definitively localize a PC. Therefore, this method may allow capsule endoscopy to be performed for more diseases.

Added value of lung perfused blood volume images using dual-energy CT for assessment of acute pulmonary embolism.

PURPOSE: To investigate the added value of lung perfused blood volume (LPBV) using dual-energy CT for the evaluation of intrapulmonary clot (IPC) in patients suspected of having acute pulmonary embolism (PE). MATERIALS AND METHODS: Institutional review board approval was obtained for this retrospective study. Eighty-three patients suspected of having PE who underwent CT pulmonary angiography (CTPA) using a dual-energy technique were enrolled in this study. Two radiologists who were blinded retrospectively and independently reviewed CTPA images alone and the combined images with color-coded LPBV over a 4-week interval, and two separate sessions were performed with a one-month interval. Inter- and intraobserver variability and diagnostic accuracy were evaluated for each reviewer with receiver operating characteristic (ROC) curve analysis. RESULTS: Values for inter- and intraobserver agreement, respectively, were better for CTPA combined with LPBV (ICC=0.847 and 0.937) than CTPA alone (ICC=0.748 and 0.861). For both readers, diagnostic accuracy (area under the ROC curve [Az]) were also superior, when CTPA alone (Az=0.888 [reader 1] and 0.912 [reader 2]) was compared with that after the combination with LPBV images (Az=0.966 [reader 1] and 0.959 [reader 2]) (p<0.001). However, Az values of both images might not have significant difference in statistics, because Az value of CTPA alone was high and 95% confidence intervals overlapped in both images. CONCLUSION: Addition of dual-energy perfusion CT to CTPA improves detection of peripheral IPCs with better interobserver agreement.

The effect of the section thickness used for 2- and 3-dimensional quantification of dual-energy perfusion computed tomography.

PURPOSE: To retrospectively investigate the effect of the section thickness used for quantifying dual-energy perfusion computed tomography (DEpCT) during 2- and 3-dimensional evaluation. METHODS: Sixty-six patients (22 males and 44 females; mean age, 59.3 years) suspected of having an acute pulmonary embolism underwent DEpCT, and 15patients were diagnosed to have intrapulmonary clots (IPCs). Two-dimensional DEpCT images were reconstructed into various section thicknesses from 1 to 10 mm at the main pulmonary artery, and the ratios of the low attenuation area (LAA) ranging from 1 to 5 HU (%LAA5) and 10 HU (%LAA10) on DEpCT were compared with the relative areas of the lung with attenuation coefficients lower than -950 HU (RA-950) using the lung CT images of each section thickness. Three-dimensional values of DEpCT were reconstructed with 3 different section thicknesses (1, 3, and 10 mm) and were analyzed for the presence of IPC burden using the factors suggesting IPC burden, including the right/left ventricular diameter ratio and CT obstruction index. RESULTS: The mean attenuation and image noise were decreased as the section thickness increased. In the 2-dimensional analysis, the %LAA5 and %LAA10 had the smallest value at 1-mm section, and DEpCT with thinner sections had a correlation with the RA-950 (r = 0.22-0.23, P < 0.05). The 3-dimensional values of DEpCT reconstructed with a 1- or 3-mm section thickness had a correlation with the CT obstruction index (r = 0.52-0.59, P < 0.05) and right/left ventricular diameter ratio (r = 0.60-0.68, P < 0.01). CONCLUSIONS: The thinner images should be used for 2- and 3-dimensional quantification of DEpCT.

[Study of high speed rotation SPECT acquisition method with breath-hold -comparison between dual-head system and triple-head system-].

BACKGROUND: Because lung perfusion Tc-99m MAA of SPECT is generally acquired under the non-breath-hold condition, SPECT image quality is degraded for respiratory lung motion. To solve this problem, a new acquisition mode would need to be developed. We suggest that the acquisition mode is the detector rotating at high speed during the deep breath-hold condition and 20 sec. This acquisition method is known as high-speed detector rotation-multiplied projection single photon emission computed tomography (HSRMP-SPECT). Detail is reconstructed from projection data added while projection data are acquired several times under deep breath-hold sets. PURPOSE: The HSRMP-SPECT technique has been developed under the triple-head system. However, the triple-head system is a small field of view. Thus a description of the entire lung area was impossible under the deep breath-hold condition. The HSRMP-SPECT technique seems needed to make the shift to dual-head system, because the generally dual-head system has a large field of view. We were compared triple-head system with dual-head system, and researched some parameter. RESULT: With lung perfusion HSRMP-SPECT using the dual-head system, description of the entire lung area possible. The triple-head system of 5 projection data sets and dual-head system of 6 projection data sets nearly equaled the total count, and also the coefficient of variation is nearly same. Therefore, we were able to make the triple-head system to the dual-head system the shift on HSRMP-SPECT.

Application of breath-holding SPECT with high-speed-rotation technique in hepatic-function scintigraphy.

Recently, a new method for acquiring SPECT data in deep inspiratory breath hold with a high-speed rotation (HSR-SPECT) technique was developed for pulmonary perfusion scintigraphy. Our purpose in this study was to apply this method to hepatic-function scintigraphy. A phantom study was performed for determining data acquisition times. The coefficient of variation (CV) value was used for evaluating nonuniformity of radioactivity in the liver area. Data acquisition times were determined based on the CV value. According to the results of the phantom study, this method needs at least ten projection data sets for acquiring sufficient counts for a proper diagnosis. Under clinical conditions, breath-holding HSR-SPECT can suppress the respiratory motion effects, better than the conventional method can. In addition, the fusion images made from breath-holding HSR-SPECT images match excellently with CT images.