Search

Respiratory motion and the associated deformations of abdominal organs and tumors are essential information in clinical applications. However, inter- and intra-patient multi-organ deformations are complex and have not been statistically formulated, whereas single organ deformations have been widely studied. In this paper, we introduce a multi-organ deformation library and its application to deformation reconstruction based on the shape features of multiple abdominal organs. Statistical multi-organ motion/deformation models of the stomach, liver, left and right kidneys, and duodenum were generated by shape matching their region labels defined on four-dimensional computed tomography images. A total of 250 volumes were measured from 25 pancreatic cancer patients. This paper also proposes a per-region-based deformation learning using the non-linear kernel model to predict the displacement of pancreatic cancer for adaptive radiotherapy. The experimental results show that the proposed concept estimates deformations better than general per-patient-based learning models and achieves a clinically acceptable estimation error with a mean distance of 1.2 ± 0.7 mm and a Hausdorff distance of 4.2 ± 2.3 mm throughout the respiratory motion.

Because implicit medical knowledge and experience are used to perform medical treatment, such decisions must be clarified when systematizing surgical procedures. We propose an algorithm that extracts low-dimensional features that are important for determining the number of fibular segments in mandibular reconstruction using the enumeration of Lasso solutions (eLasso). To perform the multi-class classification, we extend the eLasso using an importance evaluation criterion that quantifies the contribution of the extracted features. Experiment results show that the extracted 7-dimensional feature set has the same estimation performance as the set using all 49-dimensional features.

In endoscopic surgery, it is necessary to understand the three-dimensional structure of the target region to improve safety. For organs that do not deform much during surgery, preoperative computed tomography (CT) images can be used to understand their three-dimensional structure, however, deformation estimation is necessary for organs that deform substantially. Even though the intraoperative deformation estimation of organs has been widely studied, two-dimensional organ region segmentations from camera images are necessary to perform this estimation. In this paper, we propose a region segmentation method using U-net for the lung, which is an organ that deforms substantially during surgery. Because the accuracy of the results for smoker lungs is lower than that for non-smoker lungs, we improved the accuracy by translating the texture of the lung surface using a CycleGAN.

Computed tomography (CT) and magnetic resonance imaging (MRI) scanners measure three-dimensional (3D) images of patients. However, only low-dimensional local two-dimensional (2D) images may be obtained during surgery or radiotherapy. Although computer vision techniques have shown that 3D shapes can be estimated from multiple 2D images, shape reconstruction from a single 2D image such as an endoscopic image or an X-ray image remains a challenge. In this study, we propose X-ray2Shape, which permits a deep learning-based 3D organ mesh to be reconstructed from a single 2D projection image. The method learns the mesh deformation from a mean template and deep features computed from the individual projection images. Experiments with organ meshes and digitally reconstructed radiograph (DRR) images of abdominal regions were performed to confirm the estimation performance of the methods.

Because the lung deforms during surgery because of pneumothorax, it is important to be able to track the location of a tumor. Deformation of the whole lung can be estimated using intraoperative cone-beam CT (CBCT) images. In this study, we used deformable mesh registration methods for paired CBCT images in the inflated and deflated states, and analyzed their deformation. We proposed a deformable mesh registration framework for deformations of partial organ shapes involving large deformation and rotation. Experimental results showed that the proposed methods reduced errors in point-to-point correspondence. As a result of registration using surgical clips placed on the lung surface during imaging, it was confirmed that an average error of 3.9 mm occurred in eight cases. The result of analysis showed that both tissue rotation and contraction had large effects on displacement.

The present study aimed to better characterize the morphogenesis of the femur from the embryonic to the early fetal periods. Sixty-two human fetal specimens (crown–rump length [CRL] range: 11.4-185 mm) from the Kyoto Collection were used for this study. The morphogenesis and internal differentiation process of the femur were analyzed in 3D using phase-contrast X-ray computed tomography and magnetic resonance imaging. The cartilaginous femur was first observed at Carnegie stage 18. Major anatomical landmarks were formed prior to the initiation of ossification at the center of the diaphysis (CRL, 40 mm), as described by Bardeen. The region with very high signal intensity (phase 5 according to Streeter’s classification; i.e., area described as cartilage disintegration) emerged at the center of the diaphysis, which split the region with slightly low signal intensity (phase 4; i.e., cartilage cells of maximum size) in fetuses with a CRL of 40.0 mm. The phase 4 and phase 5 regions became confined to the metaphysis, which might become the epiphyseal cartilage plate. Femur length and ossified shaft length (OSL) showed a strong positive correlation with CRL. The OSL-to-femur length ratio rapidly increased in fetuses with CRL between 40 and 75 mm, which became moderately increased in fetuses with a CRL of ≥75 mm. Cartilage canal invasion occurred earlier at the proximal epiphysis (CRL, 62 mm) than at the distal epiphysis (CRL, 75 mm). Morphometry and Procrustes analysis indicated that changes in the femur shape after ossification were limited, which were mainly detected at the time of initial ossification and shortly after that. In contrast, femoral neck anteversion and torsion of the femoral head continuously changed during the fetal period. Our data could aid in understanding the morphogenesis of the femur and in differentiating normal and abnormal development during the early fetal period.

Objective: This paper introduces methods to automate preoperative planning of fibular segmentation and placement for mandibular reconstruction with fibular flaps. Methods: Preoperative virtual planning for this type of surgery has been performed by manual adjustment of many parameters, or based upon a single feature of the reconstruction. We propose a novel planning procedure formulated as a non-convex minimization problem of an objective function using the multilateral shape descriptors. Results: A retrospective study was designed and 120 reconstruction plans were reproduced using computed tomography images with oral surgeons. The proposed automated planning model was quantitatively compared with both the existing model and the surgeons’ plans. Conclusion: The results show that the developed framework attains stable automated planning that agrees with the surgeons’ decisions. Significance: This method addresses trade-off problems between symmetric reconstruction and restoration of the native contour of the mandible.

The pelvic skeleton is formed via endochondral ossification. However, it is not known how the normal cartilage is formed before ossification occurs. Furthermore, the overall timeline of cartilage formation and the morphology of the cartilage in the pelvis are unclear. In this study, cartilage formation in the pelvic skeletons of 25 human fetuses (crown-rump length [CRL] = 11.9-75.0 mm) was observed using phase-contrast computed tomography and 7T magnetic resonance imaging. The chondrification center of the ilium, ischium, and pubis first appeared simultaneously at Carnegie stage (CS) 18, was located around the acetabulum, and grew radially in the later stage. The iliac crest formed at CS20 while the iliac body's central part remained chondrified. The iliac body formed a discoid at CS22. The growth rate was greater in the ilium than in the sacrum-coccyx, pubis, and ischium. Connection and articulation formed in a limited period, while the sacroiliac joint formed at CS21. The articulation of the pubic symphysis, connection of the articular column in the sacrum, and Y-shape connection of the three parts of the hip bones to the acetabulum were observed at CS23; the connection of the ischium and pubic ramus was observed at the early-fetal stage. Furthermore, the degree of connection at the center of the sacrum varied among samples. Most of the pelvimetry data showed a high correlation with CRL. The transverse and antero-posterior lengths of the pelvic inlet of the lesser pelvis varied among samples (R2 = 0.11). The subpubic angle also varied (65-90°) and was not correlated with CRL (R2 = 0.22). Moreover, cartilaginous structure formation appeared to influence bone structure. This study provides valuable information regarding the morphogenesis of the pelvic structure.

In microendoscopic discectomy for spinal disorders, bone cutting procedures are performed in tight spaces while observing a small portion of the target structures. Although optical tracking systems are able to measure the tip of the surgical tool during surgery, the poor shape information available during surgery makes accurate cutting difficult, even if preoperative computed tomography and magnetic resonance images are used for reference. Shape estimation and visualization of the target structures are essential for accurate cutting. However, time-varying shape changes during cutting procedures are still challenging issues for intraoperative navigation. This paper introduces a concept of endoscopic image augmentation that overlays shape changes to support bone cutting procedures. This framework handles the history of the location of the measured drill tip as a volume label and visualizes the remains to be cut overlaid on the endoscopic image in real time. A cutting experiment was performed with volunteers, and the feasibility of this concept was examined using a clinical navigation system. The efficacy of the cutting aid was evaluated with respect to the shape similarity, total moved distance of a cutting tool, and required cutting time. The results of the experiments showed that cutting performance was significantly improved by the proposed framework.

This study was performed to quantitatively analyze medical knowledge of, and experience with, decision-making in preoperative virtual planning of mandibular reconstruction. Three shape descriptors were designed to evaluate local differences between reconstructed mandibles and patients’ original mandibles. We targeted an asymmetrical, wide range of cutting areas including the mandibular sidepiece, and defined a unique three-dimensional coordinate system for each mandibular image. The generalized algorithms for computing the shape descriptors were integrated into interactive planning software, where the user can refine the preoperative plan using the spatial map of the local shape distance as a visual guide. A retrospective study was conducted with two oral surgeons and two dental technicians using the developed software. The obtained 120 reconstruction plans show that the participants preferred a moderate shape distance rather than optimization to the smallest. We observed that a visually plausible shape could be obtained when considering specific anatomical features (e.g., mental foramen. mandibular midline). The proposed descriptors can be used to multilaterally evaluate reconstruction plans and systematically learn surgical procedures.

In preoperative planning for mandibular reconstructive surgery, it is necessary to determine the osteotomy lines for fibular shaping and the proper placement of fibular segments in the mandible. Although virtual surgical planning has been utilized in preoperative decision making, current software designs require manual operation and a trial-and-error process to refine the reconstruction plan. We have developed volumetric fibular transfer simulation software that can quickly design a preoperative plan based on direct volume manipulation and quantitative comparison with the patient's original mandible. We propose three quantitative shape indicators-volume ratio, contour error, and maximum projection-for symmetrical lesions of the mandible, and have implemented a parallel computation algorithm for the semiautomatic placement of fibular segments. Using this virtual planning software, we conducted a retrospective study of the computed tomography data from nine patients. We found that combining direct volume manipulation with real-time local search of placement improved the applicability of the planning system to optimize mandibular reconstruction.

This study examined the external morphology and morphometry of the human embryonic inner ear membranous labyrinth and documented its three-dimensional position in the developing embryo using phase-contrast X-ray computed tomography and magnetic resonance imaging. A total of 27 samples between Carnegie stage (CS) 17 and the postembryonic phase during trimester 1 (approximately 6-10 weeks after fertilization) were included. The otic vesicle elongated along the dorso-ventral axis and differentiated into the end lymphatic appendage and cochlear duct (CD) at CS 17. The spiral course of the CD began at CS18, with anterior and posterior semicircular ducts (SDs) forming prominent circles with a common crus. The spiral course of the CD comprised more than two turns at the postembryonic phase, at which time the height of the CD was evident. A linear increase was observed in the length of anterior, posterior, and lateral SDs, in that order, and the length of the CD increased exponentially over the course of development. Bending in the medial direction was observed between the cochlear and vestibular parts from the latero-caudal view, with the angle decreasing during development. The position of the inner ear was stable throughout the period of observation on the lateral to ventral side of the rhombencephalon, caudal to the pontine flexure, and adjacent to the auditory ganglia. The plane of the lateral semicircular canal was approximately 8.0°-14.6° with respect to the cranial caudal (z-)axis, indicating that the orientation of the inner ear changes during growth to adulthood.

A (13)C-enriched phosphorylcholine polymer ((13)C-PMPC) as a self-traceable MR (magnetic resonance) tag was conjugated with a fragment (scFv) of Herceptin, a clinical antibody against antigen Her2. When injected in model mice bearing Her2(+) (gastric) and Her2(-) (pancreatic) tumors, the antibody-tag conjugate (13)C-PMPC-scFv selectively accumulated in the Her2(+) tumor with a rapid build-up/decay (accumulation/clearance) profile and, with the use of the (1)H-(13)C double-resonance (heteronuclear correlation) technique, the Her2(+) gastric tumor was clearly MR imaged.

Novel chiral dendrimer-triamine-coordinated Gd complexes with polyaminoalcohol end groups were synthesized and shown to have longitudinal relaxivity (r1) values 5 times higher than that of clinically used Gd-DTPA. The affinities of Gd-4a and Gd-4b for bovine serum albumin (BSA), respectively, were estimated by a quartz crystal microbalance (QCM) measurement. The amino groups of the dendrimer were then conjugated with PEG. This conjugation with PEG strongly affected its ability to attenuate signals in T1-weighted MRI.

[Background]Recent advances in microscopy enable the acquisition of large numbers of tomographic images from living tissues. Three-dimensional microscope images are often displayed with volume rendering by adjusting the transfer functions. However, because the emissions from fluorescent materials and the optical properties based on point spread functions affect the imaging results, the intensity value can differ locally, even in the same structure. Further, images obtained from brain tissues contain a variety of neural structures such as dendrites and axons with complex crossings and overlapping linear structures. In these cases, the transfer functions previously used fail to optimize image generation, making it difficult to explore the connectivity of these tissues. [Results]This paper proposes an interactive visual exploration method by which the transfer functions are modified locally and interactively based on multidimensional features in the images. A direct editing interface is also provided to specify both the target region and structures with characteristic features, where all manual operations can be performed on the rendered image. This method is demonstrated using two-photon microscope images acquired from living mice, and is shown to be an effective method for interactive visual exploration of overlapping similar structures. [Conclusions]An interactive visualization method was introduced for local improvement of visualization by volume rendering in two-photon microscope images containing regions in which linear nerve structures crisscross in a complex manner. The proposed method is characterized by the localized multidimensional transfer function and interface where the parameters can be determined by the user to suit their particular visualization requirements.

[Background]Dendrites of cortical neurons are widely spread across several layers of the cortex. Recently developed two-photon microscopy systems are capable of visualizing the morphology of neurons within deeper layers of the brain and generate large amounts of volumetric imaging data from living tissue. [Method]For visual exploration of the three-dimensional (3D) structure of dendrites and the connectivity among neurons in the brain, we propose a visualization software and interface for 3D images based on a new transfer function design using volume rendered feature spaces. This software enables the visualization of multidimensional descriptors of shape and texture extracted from imaging data to characterize tissue. It also allows the efficient analysis and visualization of large data sets. [Results]We apply and demonstrate the software to two-photon microscopy images of a living mouse brain. By applying the developed visualization software and algorithms to two-photon microscope images of the mouse brain, we identified a set of feature values that distinguish characteristic structures such as soma, dendrites and apical dendrites in mouse brain. Also, the visualization interface was compared to conventional 1D/2D transfer function system. [Conclusions]We have developed a visualization tool and interface that can represent 3D feature values as textures and shapes. This visualization system allows the analysis and characterization of the higher-dimensional feature values of living tissues at the micron level and will contribute to new discoveries in basic biology and clinical medicine.

Magnetic resonance imaging (MRI) is one of the most promising techniques for the non-invasive visualization of biomarkers and biologically relevant species, both in vivo and ex vivo. Although 1H MRI with paramagnetic contrast agents, such as Gd3+ complexes and iron oxide, is widely used, it often suffers from low contrast because of the large background signals caused by the abundant distribution of protons in biological samples. Here we report the use of supramolecular organic nanoparticles to detect specific proteins by 19F-based MRI in an off/on mode. In NMR spectroscopy these designed probes are silent when aggregated, but in the presence of a target protein they disassemble to produce a sharp signal. This 'turn-on' response allowed us to visualize clearly the protein within live cells by 19F MRI and construct an in-cell inhibitor assay. This recognition-driven disassembly of nanoprobes for a turn-on 19F signal is unprecedented and may extend the use of 19F MRI for specific protein imaging.