Takashi Ohya

Intravascular catheter navigation using path planning and virtual visual feedback for oral cancer treatment.

It is tedious and difficult to position a flexible catheter in a target vessel branch within complicated‐shaped vessels owing to the lack of real‐time visual feedback. Digital subtraction angiography and fluoroscopic guidance are currently used for catheter placement.

Vessel bifurcation localization based on intraoperative three-dimensional ultrasound and catheter path for image-guided catheter intervention of oral cancers

We present a method to localize intraoperative target vessel bifurcations under bones for ultrasound (US) image-guided catheter interventions. A catheter path is recorded to acquire skeletons for the target vessel bifurcations that cannot be imaged by intraoperative US. The catheter path is combined with the centerlines of the three-dimensional (3D) US image to construct a preliminary skeleton. Based on the preliminary skeleton, the orientations of target vessels are determined by registration with the preoperative image and the bifurcations were localized by computing the vessel length. An accurate intraoperative vessel skeleton is obtained for correcting the preoperative image to compensate for vessel deformation. A reality check of the proposed method was performed in a phantom experiment. Reasonable results were obtained. The in vivo experiment verified the clinical workflow of the proposed method in an in vivo environment. The accuracy of the centerline length of the vessel for localizing the target artery bifurcation was 2.4mm. These results suggest that the proposed method can allow the catheter tip to stop at the target artery bifurcations and enter into the target arteries. This method can be applied for virtual reality-enhanced image-guided catheter intervention of oral cancers.

High-Quality Intra-operative Ultrasound Reconstruction Based on Catheter Path

This paper presents a new method for improving image quality of intra-operative ultrasound for catheter intervention. A preliminary volume was constructed by extracting all colored pixels of 2D Doppler images. After center line was computed, cross sections of vessel were sampled and segmented by a modified region growing combining intensity distribution, curvature and area threshold. An ellipse was estimated based on the segmented contour. Finally, the estimated ellipses were interpolated onto the new center line which was registered to catheter path. The method was evaluated by a vessel phantom with bifurcations. The results show that the modified region growing can estimate an accurate contour and a high quality reconstruction can be obtained by interpolating estimated cross sections.

Analysis of carotid artery deformation in different head and neck positions for maxillofacial catheter navigation in advanced oral cancer treatment

BackgroundTo improve the accuracy of catheter navigation, it is important to develop a method to predict shifts of carotid artery (CA) bifurcations caused by intraoperative deformation. An important factor affecting the accuracy of electromagnetic maxillofacial catheter navigation systems is CA deformations. We aimed to assess CA deformation in different head and neck positions.MethodsUsing two sets of computed tomography angiography (CTA) images of six patients, displacements of the skull (maxillofacial segments), C1–C4 cervical vertebrae, mandible (mandibular segment), and CA along with its branches were analyzed. Segmented rigid bones around CA were considered the main causes of CA deformation. After superimposition of maxillofacial segments, C1–C4 and mandible segments were superimposed separately for displacement measurements. Five bifurcation points (vA–vE) were assessed after extracting the CA centerline. A new standardized coordinate system, regardless of patient-specific scanning positions, was employed. It was created using the principal axes of inertia of the maxillofacial bone segments of patients. Position and orientation parameters were transferred to this coordinate system. CA deformation in different head and neck positions was assessed.ResultsAbsolute shifts in the center of gravity in the bone models for different segments were C1, 1.02 ± 0.9; C2, 2.18 ± 1.81; C3, 4.25 ± 3.85; C4, 5.90 ± 5.14; and mandible, 1.75 ± 2.76 mm. Shifts of CA bifurcations were vA, 5.52 ± 4.12; vB, 4.02 ± 3.27; vC, 4.39 ± 2.42; vD, 4.48 ± 1.88; and vE, 2.47 ± 1.32. Displacements, position changes, and orientation changes of C1–C4 segments as well as the displacements of all CA bifurcation points were similar in individual patients.ConclusionsCA deformation was objectively proven as an important factor contributing to errors in maxillofacial navigation. Our study results suggest that small movements of the bones around CA can result in small CA deformations. Although patients’ faces were not fixed properly during CT scanning, C1–C4 and vA–vE displacements were similar in individual patients. We proposed a novel method for accumulation of the displacement data, and this study indicated the importance of surrounding bone displacements in predicting CA bifurcation.

Childhood-onset inflammatory bowel diseases associated with mutation of Wiskott-Aldrich syndrome protein gene

AIM To screen primary immunodeficiency, Wiskott-Aldrich syndrome (WAS), and chronic granulomatous disease (CGD) among children with inflammatory bowel disease (IBD). METHODS This was a single-center retrospective study. Eighteen children with IBD were investigated. We analyzed their expression of Wiskott-Aldrich syndrome protein (WASP) in lymphocytes and superoxide generation in phagocytes using flow cytometry. When the expression of WASP or superoxide generation was low or absent, we performed genetic analysis to determine the cause of this. RESULTS Eighteen patients were classified as having ulcerative colitis (n = 10), Crohn’s disease (n = 5), or IBD-unclassified (n = 3). In total, three patients revealed low expression of WASP associated with a WAS gene c.1378 C>T p.Pro460Ser mutation, which has previously been reported as a pathogenic mutation in WAS and X-linked thrombocytopenia. However, with respect to the major symptoms of WAS, none of these three patients showed either thrombocytopenia or increased susceptibility to infection, but one patient showed generalized eczema. No CGD patients were discovered in this study. CONCLUSION Despite the lack of typical clinical manifestations of WAS, low expression of WASP could be associated with the pathogenesis of a subtype of IBD patients.

Real-time Visualized Vessel and Catheter Path Reconstruction for Intravascular Intervention

Minimally invasive intravascular intervention is performed with the aid of catheters for therapy. We apply intra-operative freehand ultrasound and tracked catheter for guiding intravascular intervention. Current freehand 3D ultrasound techniques separate the scanning or acquisition step from the visualization step. This leads to no feedback during image acquisition. The reconstruction quality is highly dependent on scanning experience. A low-quality imaging even causes re-scanning and prolongs the time of image acquisition. In addition, the catheter path is important information implying the skeleton of the target vessel. It also helps doctor understand the catheter position relative to vessels for reducing the risk of vessel injuries. We developed a catheter navigation system that will allow real-time visualizing the scanned vessel volume as well as catheter path. The system consists of a freehand US for vessel reconstruction, an electromagnetic tracking system for locating the position and orientation of the ultrasound probe and catheter tip, a video frame grabber for capturing ultrasound frames, and a typical PC for performing real-time reconstruction and scene rendering. The system incorporates novel methods for visualizing data acquisition of image guided intravascular intervention in real time. This paper reports on current work in progress, and focuses on methods to achieving real-time visualizing reconstructions of freehand 3D ultrasound and catheter path. We demonstrate the system on a vessel phantom. The results suggest the system could visualize the reconstruction in real-time.

Rapid Calibration of 3D Freehand Ultrasound for Vessel Intervention

To create a freehand three-dimensional (3D) ultrasound (US) system for image-guided vessel interventions, a US probe calibration must be performed to build 3D volumes from 2D vessel images. In this paper, a novel calibration method of 3D freehand US is described. It simply images the tip of a tracked needle that could freely move around within the coupling medium while the probe is fixed. The physical coordinates of the tip and the probe location are simultaneously recorded. Image coordinates of the tip are mapped to physical points. A RANSAC algorithm is applied to minimize the error matching image points and physical points. An accuracy evaluation experiment is performed to verify the proposed method. Results show that the needle tip can be accurately located and the RMS calibration error is 1.56mm. A phantom experiment and an in vivo experiment verify the availability of the technique in clinic. That means the present method is applicable to the calibration of a 3D freehand US in vessel interventions.