Segmentation Reconstruction Research Articles

Deep-Separation Guided Progressive Reconstruction Network for Semantic Segmentation of Remote Sensing Images

The success of deep learning and the segmentation of remote sensing images (RSIs) has improved semantic segmentation in recent years. However, existing RSI segmentation methods have two inherent problems: (1) detecting objects of various scales in RSIs of complex scenes is challenging, and (2) feature reconstruction for accurate segmentation is difficult. To solve these problems, we propose a deep-separation-guided progressive reconstruction network that achieves accurate RSI segmentation. First, we design a decoder comprising progressive reconstruction blocks capturing detailed features at various resolutions through multi-scale features obtained from various receptive fields to preserve accuracy during reconstruction. Subsequently, we propose a deep separation module that distinguishes various classes based on semantic features to use deep features to detect objects of different scales. Moreover, adjacent middle features are complemented during decoding to improve the segmentation performance. Extensive experimental results on two optical RSI datasets show that the proposed network outperforms 11 state-of-the-art methods.

Quantitative analysis of mesoporous structures by electron tomography: A phantom study

Electron tomography (ET) has attracted significant attention for a quantitative analysis of mesoporous materials, especially for complex disordered pore structures, as no priori assumption on the pore shape is needed, which is normally inevitable when using traditional bulk characterization techniques. However, a reliable quantification of such pore structures from ET critically depends on the fidelity of the segmented reconstruction, which can be significantly affected, e.g. by the raw data quality, the limited tilting range, artifacts introduced during alignment and further depends on the reconstruction algorithm. Therefore, we systematically investigated the reconstruction reliability of three main-stream algorithms including simultaneous iterative reconstruction technique (SIRT), total variation minimization (TVM) and discrete algebraic reconstruction technique (DART) for mesoporous materials using different imperfect (realistic) tilt-series based on a set of phantom simulations. We found that DART outperforms the other two methods in reliably revealing small pores and narrow channels, especially when the number of projections is strongly constrained. The accurately segmented reconstruction from DART makes it possible to achieve reliable quantification of the pores structure, which in turn leads to reliable evaluation of effective diffusion coefficients. We discuss the influence of different acquisition and reconstruction parameters on the reconstructed 3D volume and the quantitative analysis of pore features. We aim to provide a practical guideline for optimizing acquisition and reconstruction parameters and how to evaluate the accuracy when describing the mesoporous structure.

Reconstruction of Dentomaxillofacial Deformity Secondary to Mandibular Defect Using Concomitant Orthognathic Surgery and Fibula Free Flap.

The reconstruction of mandibular defects may be delayed or compromised for many reasons, especially in pediatric patients. With the growth of the remaining mandible and the maxilla in the malocclusion status, secondary dentomaxillofacial deformity is plausible. To treat the concomitant mandibular defect and secondary dentomaxillofacial deformity, a hierarchical algorithm using orthodontics, orthognathic surgery, and fibula free flap was developed. This retrospective case series included six patients with long-term mandibular defects caused by tumor resection without repair or with compromised costochondral reconstruction. All patients were treated using the same staged protocol, but with minor changes: (1) presurgical orthodontics, (2) virtual surgical planning, (3) fabrication of the guides and splints, (4) sequenced operations, and (5) postoperative care. The sequence of surgery included the Le Fort I osteotomy, mandibular sagittal split ramus osteotomy of the remaining ramus, final occlusion registration, repositioning of the distal segment of the mandible, segmented fibula reconstruction, and finally, the fixation of mandibular sagittal split ramus osteotomy. The operations and wound healing were uneventful in all patients, and no flap failure or severe complications were detected. Also, the patients exhibited no signs of temporomandibular joint ankylosis during the follow-up. The subspinale-nasion-supramental angle was significantly reduced after surgery. A significant improvement was detected in the facial contour symmetry measurements postoperatively. This proposed workflow of concomitant orthognathic surgery and the fibula free flap is effective and reliable for the reconstruction of dentomaxillofacial deformity secondary to the long-term mandibular defect.

An improved algorithm for three-dimensional reconstruction of nonwoven microstructures by using multi-focus images

The filtration performance of nonwovens, such as virus protection and air purification, depends largely on their microstructure, that is, their internal structure. With the development of manufacturing techniques, nonwovens can be produced with complex microstructures to deliver low air resistance and high filtration efficiency. Therefore, it seems important to reconstruct an accurate three-dimensional (3D) model to describe the various microstructures for performance evaluation and product development. Multi-focus image fusion is one of the most effective ways used for the 3D nonwoven reconstruction. However, the overlapping and crossing of fibers in multi-focus images taken on an optional microscope make 3D fiber reconstruction incomplete and discontinuous. In this study, we developed a novel fiber segmentation and 3D nonwoven reconstruction algorithm to restore fiber continuity. The fiber web image was firstly thinned into the fiber mid-axis or skeleton, and fiber fragments were detected by searching the intersections. The features, such as the angle and depth, of each fiber fragment were obtained using Harris corner detection. A cluster technique was used to group the head-points of the fiber fragments that have similar features into one cluster. All fiber fragments in the same cluster are fitted with a polynomial curve to form a complete fiber in the 3D space. The experimental results show that this new method is suitable for the reconstruction of nonwovens with complex structures, such as fiber bonding or crossing, and can automatically detect and connect fiber fragments in 3D space. It provides an effective way for the subsequent simulation/evaluation of filtration characteristics.

Visible Lymph Affluents in the D3 Volume: An MDCTA Pictorial Essay.

Background: There seems to be a gap in knowledge of the anatomy of mesenteric lymphatics between the superior mesenteric nodes and the intestinal trunk. To our knowledge, these central lymph vessels were not hitherto systematically searched for, described, or morphometrically analyzed. Our aim was to identify those vessels on the routine multidetector computerized tomography angiography (MDCTA), performed prior to right colectomy for cancer, with extended mesenterectomy, central vascular ligation, and D3 lymphadenectomy. Methods: A total of 420 MDCTA datasets were analyzed utilizing manual segmentation and 3D reconstruction, with the aid of image processing software Osirix, Mimics, and 3-matic. The 3D models and masks underwent a detailed topographic and morphometric analysis. Results: Significant vascular-like structures, having neither origin nor termination on the blood vessels, were noted in 18 cases (4.3%) in the D3 volume. The dimensions of visible lymph vessels varied, their mean diameter was 1.81 ± 0.61 mm, and the mean length was 38.07 ± 22.19 mm. In the vast majority of cases, the lymph vessels were situated in front of the superior mesenteric artery (SMA), coursing either longitudinally cranially (13 cases) or transversely/obliquely to the left (5 cases). In all cases but one, the lymph vessel passed at the left-hand side of the middle colic artery. As for the course shape, in seven cases, the lymph vessel appeared highly serpiginous. Conclusions: The regular MDCTA can provide valuable information on mesenteric lymphatics and aid in surgical planning.

Hydrogel: A Potential Material for Bone Tissue Engineering Repairing the Segmental Mandibular Defect.

Free flap surgery is currently the only successful method used by surgeons to reconstruct critical-sized defects of the jaw, and is commonly used in patients who have had bony lesions excised due to oral cancer, trauma, infection or necrosis. However, donor site morbidity remains a significant flaw of this strategy. Various biomaterials have been under investigation in search of a suitable alternative for segmental mandibular defect reconstruction. Hydrogels are group of biomaterials that have shown their potential in various tissue engineering applications, including bone regeneration, both through in vitro and in vivo pre-clinical animal trials. This review discusses different types of hydrogels, their fabrication techniques, 3D printing, their potential for bone regeneration, outcomes, and the limitations of various hydrogels in preclinical models for bone tissue engineering. This review also proposes a modified technique utilizing the potential of hydrogels combined with scaffolds and cells for efficient reconstruction of mandibular segmental defects.

Проблемы и успехи комбинированного применения технологий Илизарова и Masquelet

Introduction Based on the assessment of the problems and effectiveness in the use of the induced membrane and bone transport techniques, a new technological solution was proposed. It combines the methods of Masquelet and Ilizarov for restorative treatment of patients with bone defects and nonunion. Materials and methods The combination of the technologies was successfully applied for filling bone defects in the conditions of active purulent infection and its remission in 24 patients. Patients of the first group (n = 17) had bone defects in the conditions of the osteomyelitic process remission. In the second group of patients (n = 7), the osteomyelitic process was active. The combined technology of bone grafting included segment reconstruction in two stages. At the first stage of treatment, a sanitizing treatment of soft and bone tissues in the area of the defect and nonunion was performed followed by spacer implantation and transosseous fixation of bone fragments with the Ilizarov apparatus. During the second operating session, the spacer was removed, and after osteotomy (corticotomy), the fragment(s) were transported according to Ilizarov. Empirical antibiotic therapy against a wide range of pathogens was started after the verification of the microbial tests of the biomaterial and the determination of sensitivity to antibiotics, the correction of antibiotic therapy was carried out. Results The postoperative wounds in the area of the implanted spacers healed by primary intention in the first group. In two patients of the second group (29 %), purulent fistulas were formed by the time the spacers were removed, and the wounds healed by secondary intention. The duration of distraction in the first group was 45.4 ± 9.8 days. Bone transport in the patients of the second group continued 52.8 ± 5.3 days. The duration of fixation of the segments with the device was 195.1 ± 9.9 days in the first group and 181.8 ± 11.4 days in the second group. Discussion At the initial stage of the combination of the Masquelet technique and non-free Ilizarov bone grafting, the risks of the activity of a purulent process remain if the debridement of the infection nidus is not radical, implanted spacers with a prophylactic dose of antibiotics are massive, and the antibiotic therapy is empirical. The formation of an induced membrane with bactericidal activity at the second stage of surgical treatment, the creation of favorable conditions for bone transport, adequate sanitation of the purulent focus, and targeted antibiotic therapy ensure a stable suppression of the activity of the purulent process. Conclusions Surgical rehabilitation of patients with segmental infected defects results in bone defect filling with distraction regenerates undergoing complete organotypic restructuring, which eliminates the likelihood of deformities or fractures at the level of newly formed bone areas and reduces the risk of recurrence of the osteomyelitis process.

Intraoral versus transcervical approaches in mandibular reconstruction with free flaps: A retrospective study

This study aims to investigate the clinical and functional differences between intraoral and transcervical approaches for segmental mandible resection and reconstruction with free flaps. Patients diagnosed as benign and low-grade mandibular malignant tumors without neck dissections were retrospectively reviewed and divided into intraoral and transcervical groups. Patients of intraoral group underwent intraoral mandibulectomy and vascular anastomosis was performed through a 2-cm submandibular incision, while traditional submandibular approach was used in transcervical group. Clinical characteristics of two groups were assessed including body mass index (BMI), defect types and number of fibular segments, as well as perioperative variables such as operation time, blood loss, drainage volume. The score of appearance, swallowing and speech using the University of Washington Quality of Life Questionnaire (UW-QOL) was recorded and analyzed 6-month postoperatively.A total of 14 patients in intraoral group and 21 patients in transcervical group was collected, respectively. In intraoral group, intraoperative blood loss and postoperative drainage volume were significantly reduced in comparison with transcervical group (p = 0.0146, p = 0.0017; respectively). The score of appearance was 87.50 ± 12.97 in intraoral group, which was significantly higher than 64.29 ± 12.68 in transcervical group (p < 0.0001). Similar results were found in patients of subtype Class II mandibular defect between two groups. However, patients of intraoral group had a significant increase in operative time and a comparable amount of intraoperative blood loss (p = 0.0472, p = 0.1434; respectively).Within the limitations of the study it seems that an intraoral approach combined with a 2-cm submandibular incision should be preferred over a transcervical approach for segmental mandibulectomy and free flap reconstruction whenever appropriate.

3D objects reconstruction from frontal images: an example with guitars

This work deals with the automatic 3D reconstruction of objects from frontal RGB images. This aims at a better understanding of the reconstruction of 3D objects from RGB images and their use in immersive virtual environments. We propose a complete workflow that can be easily adapted to almost any other family of rigid objects. To explain and validate our method, we focus on guitars. First, we detect and segment the guitars present in the image using semantic segmentation methods based on convolutional neural networks. In a second step, we perform the final 3D reconstruction of the guitar by warping the rendered depth maps of a fitted 3D template in 2D image space to match the input silhouette. We validated our method by obtaining guitar reconstructions from real input images and renders of all guitar models available in the ShapeNet database. Numerical results for different object families were obtained by computing standard mesh evaluation metrics such as Intersection over Union, Chamfer Distance, and the F-score. The results of this study show that our method can automatically generate high-quality 3D object reconstructions from frontal images using various segmentation and 3D reconstruction techniques.

Global and Local Feature Reconstruction for Medical Image Segmentation.

Learning how to capture long-range dependencies and restore spatial information of down-sampled feature maps are the basis of the encoder-decoder structure networks in medical image segmentation. U-Net based methods use feature fusion to alleviate these two problems, but the global feature extraction ability and spatial information recovery ability of U-Net are still insufficient. In this paper, we propose a Global Feature Reconstruction (GFR) module to efficiently capture global context features and a Local Feature Reconstruction (LFR) module to dynamically up-sample features, respectively. For the GFR module, we first extract the global features with category representation from the feature map, then use the different level global features to reconstruct features at each location. The GFR module establishes a connection for each pair of feature elements in the entire space from a global perspective and transfers semantic information from the deep layers to the shallow layers. For the LFR module, we use low-level feature maps to guide the up-sampling process of high-level feature maps. Specifically, we use local neighborhoods to reconstruct features to achieve the transfer of spatial information. Based on the encoder-decoder architecture, we propose a Global and Local Feature Reconstruction Network (GLFRNet), in which the GFR modules are applied as skip connections and the LFR modules constitute the decoder path. The proposed GLFRNet is applied to four different medical image segmentation tasks and achieves state-of-the-art performance.

244.11: Technical Aspects and Outcomes in Hepatic Venous Reconstruction of Left Lateral Segments With Special Emphasis on Anomalous Hepatic Vein in Pediatric Living Donor Liver Transplantation

244.11: Technical Aspects and Outcomes in Hepatic Venous Reconstruction of Left Lateral Segments With Special Emphasis on Anomalous Hepatic Vein in Pediatric Living Donor Liver Transplantation

Cochlear and Vestibular Volumes in Inner Ear Malformations.

Retrospective cohort. Tertiary academic center. High-resolution CT scans of 115 temporal bones (70 with IEM; cochlear hypoplasia [CH]; n = 19), incomplete partition (IP) Types I and III (n = 16), IP Type II with an enlarged vestibular aqueduct (Mondini malformation; n = 16), enlarged vestibular aqueduct syndrome (n = 19), and 45 controls. Volumetry by software-based, semiautomatic segmentation, and 3D reconstruction. Differences in volumes among IEM and between IEM types and controls; interrater reliability. Compared with controls (mean volume, 78.0 mm3), only CH showed a significantly different cochlear volume (mean volume, 30.2 mm3; p < 0.0001) among all types of IEM. A cutoff value of 60 mm3 separated 100% of CH cases from controls. Compared with controls, significantly larger vestibular system volumes were found in Mondini malformation (mean difference, 22.9 mm3; p = 0.009) and IP (mean difference, 24.1 mm3; p = 0.005). In contrast, CH showed a significantly smaller vestibular system volume (mean difference, 41.1 mm3; p < 0.0001). A good interrater reliability was found for all three-dimensional measurements (ICC = 0.86-0.91). Quantitative reference values for IEM obtained in this study were in line with existing qualitative diagnostic characteristics. A cutoff value less than 60 mm3 may indicate an abnormally small cochlea. Normal reference values for volumes of the cochlea and vestibular system may aid in diagnosing IEM.

Hip Labral Reconstruction With a Polyurethane Scaffold: Restoration of Femoroacetabular Contact Biomechanics.

Background:Many advances have been made in hip labral repair and reconstruction and in the restoration of the suction seal.Purpose/Hypothesis:The purpose of this study was to evaluate the biomechanical effects of segmental labral reconstruction with a synthetic polyurethane scaffold (PS) in comparison with segmental labrectomy. Our hypothesis was that reconstruction with a icroporous polyurethane implant would normalize joint kinetics of the hip and restore the suction seal.Study Design:Controlled laboratory study.Methods:We used 10 hips from 5 fresh-frozen pelvises with an intact acetabular labrum without osteoarthritis. Using an intra-articular pressure measurement system, the contact area, contact pressure, and peak force were assessed for the following conditions: intact labrum, partial anterosuperior labrectomy, and PS reconstruction. For each condition, all specimens were analyzed in 4 positions (90° of flexion, 90° of flexion and internal rotation, 90° of flexion and external rotation, and 20° of extension) and underwent a labral seal test. The relative change from the intact condition was determined for all conditions and positions.Results:Compared with the intact labrum, labrectomy resulted in a significant decrease in the contact area (P < .001) and a significant increase in the peak force (P < .001) and contact pressure (P < .001) across all positions. Compared with labrectomy, PS reconstruction resulted in a significant increase in the contact area (P < .001) and a significant decrease in the contact pressure (P ≤ .02) and peak force (P < .001) across all positions. Compared with the intact labrum, PS reconstruction restored the contact area and peak force to normal values in all positions (P > .05), whereas the contact pressure was significantly decreased compared with labrectomy (P < .05) but did not return to normal values. The labral seal was lost in all specimens after labrectomy but was restored in 80% of the specimens after PS reconstruction.Conclusion:Femoroacetabular contact biomechanics significantly worsened after partial labrectomy; reconstruction using a PS restored the contact area and peak force to the intact state and improved the contact pressure increases seen after partial labrectomy. The contact area and peak force were normalized, and the labral seal was re-established in most cases.Clinical Relevance:This study provides biomechanical evidence for the use of a scaffold for labral reconstruction.

Complete fully automatic detection, segmentation and 3D reconstruction of tumor volume for non-small cell lung cancer using YOLOv4 and region-based active contour model

We aim to develop a fully automatic system that will detect, segment and accurately reconstruct non-small cell lung cancer tumors into space using YOLOv4 and region-based active contour model. The system consists of two main sections which are detection and volumetric rendering. The detection section is composed of image enhancement, augmentation, labeling and localization while the volumetric rendering is mainly image filtering, tumor extraction, region-based active contour and 3D reconstruction. In this method the images are enhanced to eliminate noise before augmentation which is intended to multiply and diversify the image data. Labeling was then carried out in order to create a solid learning foundation for the localization model. Images with localized tumors were passed through smoothing filters and then clustered to extract tumor masks. Lastly contour information was obtained to render the volumetric tumor. The designed system displays a strong detection performance with a precision of 96.57%, sensitivity and F 1 score of 97.02% and 96.79% respectively at a detection speed of 34 fps, prediction time per image of 21.38 ms. The system segmentation validation achieved a dice score coefficient of 92.19 % on tumor extraction. A 99.74 % accuracy was obtained during the verification of the method’s volumetric rendering using a 3D printed image of the rendered tumor. The rendering of the volumetric tumor was obtained at an average time of 11 s. This system shows a strong performance and reliability due to its ability to detect, segment and reconstruct a volumetric tumor into space with high confidence.

Sewer defect instance segmentation, localization, and 3D reconstruction for sewer floating capsule robots

Most automated sewer inspection tasks are based on closed-circuit television (CCTV) methods and focus on image classification or object detection but fail to obtain information on fine-grained sewer defects. Targeting the sewer defect detection from a video and performing the sewer inspection in a lightweight, low-cost, and practical manner, this study develops a sewer detection framework based on computer vision technology for the sewer floating capsule robot. The proposed framework performs three main sub-tasks, including instance segmentation, real-time sewer inspection device localization, and real three-dimensional (3D) model reconstruction, to realize sewer defect detection. An improved mask regional convolutional neural network (Mask RCNN), which integrates split attention module and balanced L1 loss module, is proposed for robust feature representation. In addition, an improved data augmentation method developed according to the sewer defect instance segmentation tasks is introduced. The 3D reconstruction and real-time localization of a sewer scene are achieved using the structure-from-motion techniques for the sewer floating capsule robot. Extensive experiments are conducted, and the experimental results show that the proposed method is effective and robust. The average precision of instance segmentation at the intersection of the union value of 0.5 is 92.7%, and the maximum 3D model measurement error is 1 m. However, video sequence information and multi-sensor fusion, combining the inertial measurement unit and vision technique, could be studied in the future to achieve better generalization and robust results.

IV segment portal vein reconstruction in split-liver transplantation with extended right grafts

BackgroundLiver transplantation is one of the most effective treatments for end-stage liver disease. Split liver transplantation (SLT) can effectively improve the utilization efficiency of grafts. However, split liver transplantation still faces shortcomings and is not widely used in surgery. How to improve the effective transplantation volume of split liver transplantation and promote the postoperative recovery of patients has important clinical significance.MethodsIn our study, the donor’s liver was split into the extended right graft and left lateral sector, and the IV segment occur ischemia. To guarantee the functional graft size, and avoid complications, we reconstructed the IV segment portal vein and left portal vein. And we analyzed the operation time, intraoperative bleeding, liver function, and postoperative complications.ResultsIn our research, 14 patients underwent IV segment portal vein reconstruction, and 8 patients did not undergo vascular reconstruction. We found that the ischemic area of the IV segment decreased significantly after IV segment portal vein reconstruction. We found that there was no significant difference in operation time and postoperative complications between the patients of the groups. There were significant differences in ALT on the 1st day and albumin on the 6th day after the operation.ConclusionIt indicates that IV segment reconstruction in SLT surgery can alleviate the graft ischemic and promote the recovery of liver function after the operation. And, IV segment reconstruction as a novel operating procedure may be widely used in SLT.

Automated Three-Dimensional Liver Reconstruction with Artificial Intelligence for Virtual Hepatectomy

ObjectiveTo validate the newly developed artificial intelligence (AI)-assisted simulation by evaluating the speed of three-dimensional (3D) reconstruction and accuracy of segmental volumetry among patients with liver tumors. BackgroundAI with a deep learning algorithm based on healthy liver computer tomography images has been developed to assist three-dimensional liver reconstruction in virtual hepatectomy. Methods3D reconstruction using hepatic computed tomography scans of 144 patients with liver tumors was performed using two different versions of Synapse 3D (Fujifilm, Tokyo, Japan): the manual method based on the tracking algorithm and the AI-assisted method. Processing time to 3D reconstruction and volumetry of whole liver, tumor-containing and tumor-free segments were compared. ResultsThe median total liver volume and the volume ratio of a tumor-containing and a tumor-free segment were calculated as 1035 mL, 9.4%, and 9.8% by the AI-assisted reconstruction, whereas 1120 mL, 9.9%, and 9.3% by the manual reconstruction method. The mean absolute deviations were 16.7 mL and 1.0% in the tumor-containing segment and 15.5 mL and 1.0% in the tumor-free segment. The processing time was shorter in the AI-assisted (2.1 vs. 35.0 min; p < 0.001). ConclusionsThe virtual hepatectomy, including functional liver volumetric analysis, using the 3D liver models reconstructed by the AI-assisted methods, was reliable for the practical planning of liver tumor resections.

In situ laser profilometry for material segmentation and digital reconstruction of a multicomponent additively manufactured part

In addition to its ability to produce geometrically complex parts, additive manufacturing offers a unique opportunity to collect data about a component while it is being fabricated. However, there has only been limited effort to characterize parts morphologically and compositionally in situ. In this article, we present a layer-by-layer, laser profilometry-based in situ characterization technique as a method to digitally reconstruct a multi-material part. Data collected by the laser profilometer yields height maps and grayscale images which are voxelized using purpose-built software to volumetrically reconstruct the part. The same part was also analyzed using X-ray computed tomography (CT) which was not able to resolve the different compositional regions within the part, but captured the filament morphology. The part was then bisected to compare the digital reconstruction to the actual part morphology and composition. Overall, the digital reconstruction was in good agreement with both the CT and bisected images. Deviations between the digital reconstruction and the CT/bisected images are likely the result of image segmentation settings or material shifts after data was collected. The in situ characterization method demonstrated here sets the stage for real time process monitoring and paves the way for additively manufactured parts that are “born qualified.”

Biochemical and biomechanical characterization of the cervical, thoracic, and lumbar facet joint cartilage in the Yucatan minipig

Facet joint arthrosis causes pain in approximately 7 % of the U.S. population, but current treatments are palliative. The objective of this study was to elucidate structure-function relationships and aid in the development of future treatments for the facet joint. This study characterized the articular surfaces of cervical, thoracic, and lumbar facet cartilage from skeletally mature (18–24 mo) Yucatan minipigs. The minipig was selected as the animal model because it is recognized by the U.S. Food and Drug Administration (FDA) and the American Society for Testing and Materials (ASTM) as a translationally relevant model for spine-related indications. It was found that the thoracic facets had a ∼2 times higher aspect ratio than lumbar and cervical facets. Lumbar facets had 6.9–9.6 times higher % depth than the cervical and thoracic facets. Aggregate modulus values ranged from 135 to 262 kPa, much lower than reported aggregate modulus in the human knee (reported to be 530–701 kPa). The tensile Young’s modulus values ranged from 6.7 to 20.3 MPa, with the lumbar superior facet being 304 % and 286 % higher than the cervical inferior and thoracic superior facets, respectively. Moreover, 3D reconstructions of entire vertebral segments were generated. The results of this study imply that structure-function relationships in the facet cartilage are different from other joint cartilages because biochemical properties are analogous to other articular cartilage sources whereas mechanical properties are not. By providing functional properties and a 3D database of minipig facet geometries, this work may supply design criteria for future facet tissue engineering efforts.

Combatting the effect of image reconstruction settings on lymphoma [18F]FDG PET metabolic tumor volume assessment using various segmentation methods

Background[18F]FDG PET-based metabolic tumor volume (MTV) is a promising prognostic marker for lymphoma patients. The aim of this study is to assess the sensitivity of several MTV segmentation methods to variations in image reconstruction methods and the ability of ComBat to improve MTV reproducibility.MethodsFifty-six lesions were segmented from baseline [18F]FDG PET scans of 19 lymphoma patients. For each scan, EARL1 and EARL2 standards and locally clinically preferred reconstruction protocols were applied. Lesions were delineated using 9 semiautomatic segmentation methods: fixed threshold based on standardized uptake value (SUV), (SUV = 4, SUV = 2.5), relative threshold (41% of SUVmax [41M], 50% of SUVpeak [A50P]), majority vote-based methods that select voxels detected by at least 2 (MV2) and 3 (MV3) out of the latter 4 methods, Nestle thresholding, and methods that identify the optimal method based on SUVmax (L2A, L2B). MTVs from EARL2 and locally clinically preferred reconstructions were compared to those from EARL1. Finally, different versions of ComBat were explored to harmonize the data.ResultsMTVs from the SUV4.0 method were least sensitive to the use of different reconstructions (MTV ratio: median = 1.01, interquartile range = [0.96–1.10]). After ComBat harmonization, an improved agreement of MTVs among different reconstructions was found for most segmentation methods. The regular implementation of ComBat (‘Regular ComBat’) using non-transformed distributions resulted in less accurate and precise MTV alignments than a version using log-transformed datasets (‘Log-transformed ComBat’).ConclusionMTV depends on both segmentation method and reconstruction methods. ComBat reduces reconstruction dependent MTV variability, especially when log-transformation is used to account for the non-normal distribution of MTVs.