Search

This paper presents methods of decomposition of musculoskeletal structures from radiographs into multiple individual muscle and bone structures. While existing solutions require dual-energy scan for the training dataset and are mainly applied to structures with high-intensity contrast, such as bones, we focused on multiple superimposed muscles with subtle contrast in addition to bones. The decomposition problem is formulated as an image translation problem between (1) a real X-ray image and (2) multiple digitally reconstructed radiographs, each of which contains a single muscle or bone structure, and solved using unpaired training based on the CycleGAN framework. The training dataset was created via automatic computed tomography (CT) segmentation of muscle/bone regions and virtually projecting them with geometric parameters similar to the real X-ray images. Two additional features were incorporated into the CycleGAN framework to achieve a high-resolution and accurate decomposition: hierarchical learning and reconstruction loss with the gradient correlation similarity metric. Furthermore, we introduced a new diagnostic metric for muscle asymmetry directly measured from a plain X-ray image to validate the proposed method. Our simulation and real-image experiments using real X-ray and CT images of 475 patients with hip diseases suggested that each additional feature significantly enhanced the decomposition accuracy. The experiments also evaluated the accuracy of muscle volume ratio measurement, which suggested a potential application to muscle asymmetry assessment from an X-ray image for diagnostic and therapeutic assistance. The improved CycleGAN framework can be applied for investigating the decomposition of musculoskeletal structures from single radiographs.

Background Cerebral amyloid angiopathy is a common cause of subcortical hemorrhage in older adults. Although open hematoma removal may be performed for severe subcortical hemorrhage, its safety in patients with cerebral amyloid angiopathy has not been established, and postoperative rebleeding may occur. Therefore, this study aimed to investigate factors associated with postoperative rebleeding. Methods Out of 145 consecutive patients who had undergone craniotomy for surgical removal of subcortical intracerebral hemorrhage between April 2010 and August 2019 at a single institution in Japan, we examined 109 patients with subcortical hemorrhage who met the inclusion criteria. After excluding 30 patients whose tissue samples were unsuitable for the study, the final study cohort comprised 79 patients. Results Of the 79 patients, 50 (63%) were diagnosed with cerebral amyloid angiopathy (cerebral amyloid angiopathy group) and 29 (37%) were not diagnosed with noncerebral amyloid angiopathy (noncerebral amyloid angiopathy group). Postoperative rebleeding occurred in 12 patients (24%) in the cerebral amyloid angiopathy group and in 2 patients (7%) in the noncerebral amyloid angiopathy group. Preoperative prothrombin time–international normalized ratio and intraoperative bleeding volume were significantly associated with postoperative rebleeding in the cerebral amyloid angiopathy group (odds ratio = 42.4, 95% confidence interval = 1.14–1578; p = 0.042 and odds ratio = 1.005, 95% confidence interval = 1.001–1.008; p = 0.007, respectively). Conclusions Patients with cerebral amyloid angiopathy-related cerebral hemorrhage who are receiving antithrombotic therapy, particularly warfarin therapy, are at a high risk of postoperative rebleeding. Trial registration Registry and Registration Number of the study: 19–220, 2019/12/23, retrospectively registered.

Background Cerebral amyloid angiopathy (CAA) occurs in 80% of patients with Alzheimer’s disease (AD) and is mainly caused by the abnormal deposition of Aβ in the walls of cerebral blood vessels. Cerebrovascular molecular mechanisms in CAA were investigated by using comprehensive and accurate quantitative proteomics. Methods Concerning the molecular mechanisms specific to CAA, formalin-fixed paraffin-embedded (FFPE) sections were prepared from patients having AD neuropathologic change (ADNC) with severe cortical Aβ vascular deposition (ADNC +/CAA +), and from patients having ADNC without vascular deposition of Aβ (ADNC +/CAA −; so called, AD). Cerebral cortical vessels were isolated from FFPE sections using laser microdissection (LMD), processed by pressure cycling technology (PCT), and applied to SWATH (sequential window acquisition of all theoretical fragment ion spectra) proteomics. Results The protein expression levels of 17 proteins in ADNC +/CAA +/H donors (ADNC +/CAA + donors with highly abundant Aβ in capillaries) were significantly different from those in ADNC +/CAA − and ADNC −/CAA − donors. Furthermore, we identified 56 proteins showing more than a 1.5-fold difference in average expression levels between ADNC +/CAA + and ADNC −/CAA − donors, and were significantly correlated with the levels of Aβ or Collagen alpha-2(VI) chain (COL6A2) (CAA markers) in 11 donors (6 ADNC +/CAA + and 5 ADNC −/CAA −). Over 70% of the 56 proteins showed ADNC +/CAA + specific changes in protein expression. The comparative analysis with brain parenchyma showed that more than 90% of the 56 proteins were vascular-specific pathological changes. A literature-based pathway analysis showed that 42 proteins are associated with fibrosis, oxidative stress and apoptosis. This included the increased expression of Heat shock protein HSP 90-alpha, CD44 antigen and Carbonic anhydrase 1 which are inhibited by potential drugs against CAA. Conclusions The combination of LMD-based isolation of vessels from FFPE sections, PCT-assisted sample processing and SWATH analysis (FFPE-LMD-PCT-SWATH method) revealed for the first time the changes in the expression of many proteins that are involved in fibrosis, ROS production and cell death in ADNC +/CAA + (CAA patients) vessels. The findings reported herein would be useful for developing a better understanding of the pathology of CAA and for promoting the discovery and development of drugs and biomarkers for CAA.

We investigated the accuracy of the acetabular cup position and orientation in robotic-assisted total hip arthroplasty (rTHA) compared to navigated THA (nTHA) using computed tomography (CT) for patients with osteoarthritis secondary to developmental dysplasia of the hip (DDH). We studied 31 hips of 28 patients who underwent rTHA and 119 hips of 112 patients who underwent nTHA with the same target cup orientation. After propensity score matching, each group comprised 29 hips. Post-operative cup position and orientation were measured from the postoperative CT data. Errors from the target cup position and orientation were compared between the two groups. Post-operatively, the absolute error of the anteroposterior and superoinferior cup positions from the target position was significantly lower in the rTHA group than in the nTHA group. The change in the post-operative radiographic inclination from the target orientation was lower in the rTHA group than in the nTHA group. Screw fixation for cup implantation was required for three hips in the nTHA group but not in the rTHA group. rTHA achieved more precise cup implantation with reduced variation from the target orientation compared to nTHA in patients with osteoarthritis secondary to DDH.

In quantitative computed tomography (CT), manual selection of the intensity calibration phantom’s region of interest is necessary for calculating density (mg/cm3) from the radiodensity values (Hounsfield units: HU). However, as this manual process requires effort and time, the purposes of this study were to develop a system that applies a convolutional neural network (CNN) to automatically segment intensity calibration phantom regions in CT images and to test the system in a large cohort to evaluate its robustness. This cross-sectional, retrospective study included 1040 cases (520 each from two institutions) in which an intensity calibration phantom (B-MAS200, Kyoto Kagaku, Kyoto, Japan) was used. A training dataset was created by manually segmenting the phantom regions for 40 cases (20 cases for each institution). The CNN model’s segmentation accuracy was assessed with the Dice coefficient, and the average symmetric surface distance was assessed through fourfold cross-validation. Further, absolute difference of HU was compared between manually and automatically segmented regions. The system was tested on the remaining 1000 cases. For each institution, linear regression was applied to calculate the correlation coefficients between HU and phantom density. The source code and the model used for phantom segmentation can be accessed at https://github.com/keisuke-uemura/CT-Intensity-Calibration-Phantom-Segmentation . The median Dice coefficient was 0.977, and the median average symmetric surface distance was 0.116 mm. The median absolute difference of the segmented regions between manual and automated segmentation was 0.114 HU. For the test cases, the median correlation coefficients were 0.9998 and 0.999 for the two institutions, with a minimum value of 0.9863. The proposed CNN model successfully segmented the calibration phantom regions in CT images with excellent accuracy.

Age-related regeneration failure in the central nervous system can occur as a result of a decline in remyelination efficacy. The responsiveness of myelin-forming cells to signals for remyelination is affected by aging-related epigenetic modification; however, the molecular mechanism is not fully clarified. In the present study, we report that the apelin receptor (APJ) mediates remyelination efficiency with age. APJ expression in myelin-forming cells is correlated with age-associated changes in remyelination efficiency, and the activation of APJ promotes remyelination through the translocation of myelin regulatory factor. APJ signaling activation promoted remyelination in both aged mice with toxin-induced demyelination and mice with experimental autoimmune encephalomyelitis. In human cells, APJ activation enhanced the expression of remyelination markers. Impaired oligodendrocyte function in aged animals can be reversibly reactivated; thus, the results demonstrate that dysfunction of the apelin–APJ system mediates remyelination failure in aged animals, and that their myelinating function can be reactivated by APJ activation. Ito et al. show that regeneration in the aging brain is impaired due to reduced expression of the apelin receptor APJ. Circulating apelin signals oligodendrocytes via APJ to support remyelination, and this pathway can be restored in older mice with an APJ agonist.

This study aimed to investigate whether use of a computed tomography (CT)-based navigation system reduce the risk of dislocation after total hip arthroplasty (THA) in patients with osteonecrosis of the femoral head (ONFH). A total of 271 hips from 192 consecutive patients that underwent primary THA for ONFH were included. There were 110 hips in non-navigation group, and 161 hips in navigation group. After applying exclusion criteria, 209 hips from 149 patients were selected for analysis. Clinical outcomes and complication rates were evaluated, and implant alignments were also calculated. To identify whether the navigation system was useful to prevent dislocation, the inverse probability of treatment-weighted Cox regression analysis using a propensity score in relationship to sex, age at surgery, body mass index, and femoral head size was performed. No significant difference was observed in clinical scores between both groups. Dislocation was significantly lower in the navigation group (3 hips, 2.7%) than in the non-navigation group (11 hips, 11.2%; p = 0.012), whereas periprosthetic joint infection and aseptic loosening did not differ between the groups. Variance of cup inclination and anteversion angles was smaller in the navigation group than in the non-navigation group (p

In total hip arthroplasty, analysis of postoperative medical images is important to evaluate surgical outcome. Since Computed Tomography (CT) is most prevalent modality in orthopedic surgery, we aimed at the analysis of CT image. In this work, we focus on the metal artifact in postoperative CT caused by the metallic implant, which reduces the accuracy of segmentation especially in the vicinity of the implant. Our goal was to develop an automated segmentation method of the bones and muscles in the postoperative CT images. We propose a method that combines Normalized Metal Artifact Reduction (NMAR), which is one of the state-of-the-art metal artifact reduction methods, and a Convolutional Neural Network-based segmentation using two U-net architectures. The first U-net refines the result of NMAR and the Bayesian muscle segmentation is performed by the second U-net. We conducted experiments using simulated images of 20 patients and real images of three patients to evaluate the segmentation accuracy of 19 muscles. In simulation study, the proposed method showed statistically significant improvement (p

Commercial software is generally needed to measure the areal bone mineral density (aBMD) of the proximal femur from clinical computed tomography (CT) images. This study developed and verified an open-source reproducible system to quantify CT-aBMD to screen osteoporosis using clinical CT images.For existing CT images acquired for various reasons other than osteoporosis, it might be beneficial to estimate areal BMD as assessed by dual-energy X-ray absorptiometry (DXA-based BMD) to ascertain the bone status based on DXA. In this study, we aimed to (1) develop an open-source reproducible measurement system to quantify DXA-based BMD from CT images and (2) validate its accuracy.This study analyzed 75 pairs of hip CT and DXA images of women that were acquired for the preoperative assessment of total hip arthroplasty. From the CT images, the femur and a calibration phantom were automatically segmented using pre-trained codes/models available at https://github.com/keisuke-uemura . The proximal femoral region was isolated by manually selected landmarks and was projected onto the coronal plane to measure the areal density (CT-aHU). The calibration phantom was employed to convert the CT-aHU into CT-aBMD. Each parameter was correlated with DXA-based BMD, and the residual errors of CT images to estimate the T-scores in DXA were calculated using the standard error of estimate (SEE).The correlation coefficients of DXA-based BMD with CT-aHU and CT-aBMD were 0.947 and 0.950, respectively (both p 0.001). The SEE for quantifying the T-scores in DXA were 0.51 and 0.50 for CT-aHU and CT-aBMD, respectively.With the method developed herein, CT permits estimation of the DXA-based BMD of the proximal femur within the standard DXA total hip region of interest with an SEE of 0.5 in T-scores. The radiation dose for CT acquisition needs consideration; therefore, our data do not provide a rationale for performing CT for screening osteoporosis. However, on CT images already acquired for clinical indications other than osteoporosis, researchers may use this open-source system to investigate osteoporosis status through the estimated DXA-based BMD of the proximal femur.

We evaluated the course of the femoral blood vessels of patients with acetabular dysplasia. Patients were divided into five groups: those with Crowe type I, II, III, and IV dysplastic hips and those with normal hips. A computed tomography-based hip navigation software was used to measure the distance between the femoral blood vessels and the anterior pelvic wall in four axial planes located 10–40 mm proximal to the pelvic teardrop. In Crowe Groups I through IV, the distance was shortest at a point 20 mm proximal to the pelvic teardrop. Furthermore, the distance decreased as the Crowe classification grade increased. Because the femoral blood vessels pass close to the pelvis in many patients in Crowe III and IV hips, caution is required during surgery in these patients.

The purpose of the present study was to determine which factors affect the positional accuracy of iliosacral screws inserted using 3D fluoroscopic navigation. Specifically, we asked: (1) does the screw insertion angle in the coronal and axial planes affect the positional accuracy of iliosacral screw insertion using 3D fluoroscopic navigation? (2) Is the positional accuracy of iliosacral screw insertion using 3D fluoroscopic navigation affected by the type of screw (transsacral versus standard iliosacral), site of screw insertion (S1 versus S2), patient position (supine versus prone), presence of a dysmorphic sacrum, or AO/OTA classification (type B versus C)? Twenty-seven patients with AO/OTA type B or C pelvic ring fracture were treated by percutaneous iliosacral screw fixation. A total of 55 screws were inserted into S1 or S2 using 3D fluoroscopic navigation combined with preoperative CT-based planning. The positional accuracy of screw placement was assessed by matching postoperative CT images with preoperative CT images. The distance between the central axis of the inserted screw and that of the planned screw placement was measured in the sagittal plane passing through the center of the vertebral body. The mean deviation between the planned and the inserted screw position was 2.9 ± 1.7 mm (range 0–8.5 mm) at the vertebral body center. Multiple regression analysis showed that the screw insertion angle relative to the vertical line of the bone surface in the axial plane (β = 0.354, p = 0.013) and the use of a transsacral screw (β = 0.317, p = 0.017) were correlated with the positional accuracy of screw placement (adjusted R2 = 0.276, p = 0.002). A greater screw insertion angle relative to the vertical line on the bone surface and the use of transsacral screws increases the positional error of iliosacral screws inserted using 3D fluoroscopic navigation. Level IV, therapeutic study.

Patient-specific quantitative assessments of muscle mass and biomechanical musculoskeletal simulations require segmentation of the muscles from medical images. The objective of this work is to automate muscle segmentation from CT data of the hip and thigh. We propose a hierarchical multi-atlas method in which each hierarchy includes spatial normalization using simpler pre-segmented structures in order to reduce the inter-patient variability of more complex target structures. The proposed hierarchical method was evaluated with 19 muscles from 20 CT images of the hip and thigh using the manual segmentation by expert orthopedic surgeons as ground truth. The average symmetric surface distance was significantly reduced in the proposed method (1.53 mm) in comparison with the conventional method (2.65 mm). We demonstrated that the proposed hierarchical multi-atlas method improved the accuracy of muscle segmentation from CT images, in which large inter-patient variability and insufficient contrast were involved.

The purposes of this study were to document novel Raman spectroscopic findings in femoral heads affected by osteonecrosis and to identify molecular structure differences based on aetiology.We obtained 13 femoral heads with osteonecrosis from 13 different patients who underwent total hip arthroplasty. Comparisons were made between the viable zones of each femoral head examined. The samples were scanned with X-ray micro-CT for structural mapping and a central coronal section slab was prepared for Raman spectroscopy and histological analyses. Raman spectra were collected at different locations, including the viable and necrotic zones of the femoral head, using a highly spectrally resolved Raman microprobe.Significant alterations in the spectral morphology in the high wavenumber region were found, with a pronounced inhibition of peculiar lipid signals in the frequency interval 2851 ~ 2890 cmWe systematically found a decrease in Raman intensity for sphingomyelin and phenylalanine fingerprint bands in the necrotic zones, and these differences may be related to the etiology of osteonecrosis.

For successful joint preservation in osteonecrosis of the femoral head (ONFH), it is important to understand the mechanism of collapse. The purpose of this study was to investigate the initiation of subchondral fracture in ONFH by using micro-CT imaging of the whole femoral head, focusing on the three-dimensional relationship between the subchondral fracture and the bone resorption area. A total of 40 femoral heads from 37 patients retrieved during total hip arthroplasty for stage 3A or 3B ONFH by Japanese Investigation Committee criteria were scanned using micro-CT with a 0.146-mm thickness cuts. We divided the cohort into early and late collapsed stages according to a threshold of 3 mm of collapse as measured by micro-CT. According to the analysis on multiple radial plane views in the whole femoral head, there were two interesting findings. First, the initial fracture cracks ran between separated bone resorption areas at the anterosuperior portions of all 18 femoral heads in the early collapsed stage. Second, fractures of the necrotic bone at the sclerotic boundary and a fibrous, granulation-like, low-density tissue along the necrotic side of the sclerotic boundary were seen in 19 of the 22 in the late collapsed stage. After bone resorption around the retinaculum and teres insertion initiates the subchondral fracture, bone resorption expanding at the anterosuperior portion of the femoral head may result in the spread of fracture and the potential for massive collapse. Three-dimensional micro-CT showed bone resorption around the reparative zone initiates the subchondral fracture in ONFH.

Reduced range of motion (ROM) in flexion and internal rotation are associated with impaired activities of daily living (ADL) after rotational acetabular osteotomy (RAO). This study focused on the morphological variation of the anterior inferior iliac spine in developmental dysplasia of the hip (DDH) and its impact on post-operative bony ROM after RAO. This study aimed to investigate the association between bony ROM after RAO and pre-operative morphological factors of the pelvis and femur, including a positional variation of the anterior inferior iliac spine. Bony ROM in 52 patients with DDH was assessed after virtual RAO using computed tomography (CT). Post-operative acetabular coverage was set at lateral and anterior centre-edge angles of 30° and 55°, respectively. The position of the anterior inferior iliac spine was classified as higher or lower. Multiple regression analysis revealed that the lower anterior inferior iliac spine and higher femoral neck shaft angle were significantly associated with the lower flexion angle after RAO. Lower femoral anteversion, higher femoral neck shaft angle and higher alpha angle at the anterosuperior part were significantly associated with lower internal rotation angle at 90° flexion after RAO. Therefore, morphological variation of the anterior inferior iliac spine affected bony ROM in flexion and that of the femoral neck affected bony ROM in internal rotation at flexion after RAO.

Parkinson's disease (PD) is pathologically characterized by the presence of intraneuronal inclusions, termed Lewy bodies and Lewy neurites, whose main component is alpha-synuclein. Based on the topographic distribution of Lewy bodies and neurites established after autopsy from PD patients, Braak and coworkers hypothesized that PD pathology may start in the gastrointestinal tract then spread through the vagus nerve to the brain. This hypothesis has been reinforced by the discovery that alpha-synuclein may be capable of spreading transcellularly, thereby providing a mechanistic basis for Braak's hypothesis. This 'gut to brain' scenario has ignited heated debates within the movement disorders community and prompted a large number of studies in both humans and animals. Here, we review the arguments for and against the gut as the origin of PD. We conclude that the human autopsy evidence does not support the hypothesis and that it is too early to draw any definitive conclusions. We discuss how this issue might be further addressed in future research.

A previous study reported that a massive cerebral infarct in the territory of the middle cerebral artery (MCA) may be associated with development of neurofibrillary tangles (NFTs) in the ipsilateral basal nucleus of Meynert (BNM). We analyzed 19 cases of an MCA territory infarct and 12 with a putaminal hemorrhage (mean age 82.5 years; female/male ratio 8/23; mean time from stroke onset to autopsy 4182 days). In both groups, 74–100% had a significantly higher rate of phosphorylated tau immunoreactive or Gallyas Braak silver stain-positive neurons on the BNM-affected side than on the BNM-unaffected side. These NFTs were immunoreactive for anti-RD3 and anti-RD4 antibodies, and a triple-band pattern was observed by immunoblot analysis with anti-tau antibody. Most NFTs might be formed within the 5–10 years after stroke onset. There were significantly more TAR DNA-binding protein 43 (TDP43) immunoreactive structures on the BNM-affected side than on the BNM-unaffected side. We showed that many NFTs with TDP43-immunoreactive structures were observed in the ipsilateral BNM associated with a massive cerebral infarct in the MCA territory or a putaminal hemorrhage.

The purpose of this study was to clarify bone turnover marker levels in rapidly destructive coxopathy (RDC). Twenty patients with RDC (mean age, 72 ± 11 years; 3 men, 17 postmenopausal women), 111 with osteoarthritis (OA) (age, 60 ± 10 years; 15 men, 13 premenopausal women, 83 postmenopausal women), and 18 with osteonecrosis of femoral head (ON) (55 ± 14 years; 11 men, 3 premenopausal women, 4 postmenopausal women), and 100 patients with femoral neck fracture (FNF) (81 ± 10 years; 27 men, 73 postmenopausal women) were included. Serum tartrate-resistant acid phosphatase 5b (TRACP-5b), bone alkaline phosphatase (BAP), matrix metalloproteinase-3 (MMP-3) levels, and bone mineral density (BMD) of proximal femur and lumbar spine were investigated. TRACP-5b levels were significantly higher in RDC than in OA and ON, whereas BAP levels were higher in RDC than in OA (P

This study describes the use of CT images in atlas-based automated planning methods for acetabular cup implants in total hip arthroplasty (THA). The objective of this study is to develop an automated cup planning method considering the statistical distribution of the residual thickness. From a number of past THA planning datasets, we construct two statistical atlases that represent the surgeon’s expertise. The first atlas is a pelvis-cup merged statistical shape model (PC-SSM), which encodes global spatial relationships between the patient anatomy and implant. The other is a statistical residual thickness map (SRTM) of the implant surface, which encodes local spatial constraints of the anatomy and implant. In addition to PC-SSM and SRTM, we utilized the minimum thickness as a threshold constraint to prevent penetration. The proposed method was applied to the pelvis shapes segmented from CT images of 37 datasets of osteoarthritis patients. Automated planning results with manual segmentation were compared to the plans prepared by an experienced surgeon. There was no significant difference in the average cup size error between the two methods (1.1 and 1.2 mm, respectively). The average positional error obtained by the proposed method, which integrates the two atlases, was significantly smaller (3.2 mm) than the previous method, which uses single atlas (3.9 mm). In the proposed method with automated segmentation, the size error of the proposed method for automated segmentation was comparable (1.1 mm) to that for manual segmentation (1.1 mm). The average positional error was significantly worse (4.2 mm) than that using manual segmentation (3.2 mm). If we only consider mildly diseased cases, however, there was no significance between them (3.2 mm in automated and 2.6 mm in manual segmentation). We infer that integrating PC-SSM and SRTM is a useful approach for modeling experienced surgeon’s preference during cup planning.