External-beam radiotherapy followed by high dose rate (HDR) brachytherapy is the standard-of-care for treating gynecologic cancers. The enhanced soft-tissue contrast provided magnetic resonance imaging (MRI) makes it a valuable modality diagnosing and these However, in to computed tomography (CT) imaging, appearance of catheters, through which radiation sources are inserted reach cancerous tissue later on, often variable across images. This paper reports, first time, new deep-learning-based...

This study presents a new visuo-haptic virtual reality (VR) training and planning system for percutaneous transhepatic cholangio-drainage (PTCD) based on partially segmented patient models. We only use image data instead of full segmentation circumvent the necessity surface or volume mesh Haptic interaction with during palpation, ultrasound probing needle insertion is provided. Furthermore, VR simulator includes X-ray simulation image-guided training. The visualization techniques are...

This article presents methods for direct visuo-haptic 4D volume rendering of virtual patient models under respiratory motion. Breathing are computed based on patient-specific CT image data sequences. Virtual visualized in real-time by ray casting a reference warped time-variant displacement field, which is using the motion at run-time. Furthermore, haptic interaction with animated provided displacements high rates to translate position device into space image. concept applied palpation and...

This work presents an evaluation study using a force feedback framework for novel direct needle volume rendering concept in the context of liver puncture simulation. PTC/PTCD interventions targeting bile ducts have been selected to illustrate this concept. The haptic algorithms simulator system are based on (1) partially segmented patient image data and (2) non-linear spring model effective at organ borders. primary aim is quantitatively evaluate errors caused by our modeling approach,...

A two-step concept for 3D segmentation on 5 abdominal organs inside volumetric CT images is presented. Firsteach relevant organ’s volume of interest extracted as bounding box. The acts input asecond stage, wherein two compared U-Nets with different architectural dimensions re-construct an organ segmen-tation label mask. In this work, we focus comparing 2D vs. U-Net counterparts. Our initial resultsindicate Dice improvements about 6% at maximum. study to our surprise, liver and kidneys...

For patient-specific voxel-based visuo-haptic rendering of CT scans the liver area, fully automatic segmentation large volume structures such as skin, soft tissue, lungs and intestine (risk structures) is important. Using a machine learning based approach, several existing segmentations from 10 segmented gold-standard patients are learned by random decision forests individually collectively. The core this paper feature selection application classifiers to new patient data set. In...

The preparation phase for surgical simulations often comprises the segmentation of patient data, which is needed realistic visual and haptic rendering. Expert 3D data sets can last from several hours to days. In this paper we introduce a direct volume rendering approach lumbar punctures. Preparation time spent much shorter compared our reference system nearly identical force output at needle tip be observed. number structures completely segmented by an expert reduced 11 3 tissues in...

Virtual reality (VR) training simulators of liver needle insertion in the hepatic area breathing virtual patients often need 4D image data acquisitions as a prerequisite. Here, first population-based patient atlas is built and second requirement dose-relevant or expensive acquisition CT MRI set for new can be mitigated by warping mean motion. The breakthrough contribution this work construction reuse population-based, learned motion models.

Virtual reality techniques can be used for the training of needle insertion interventions. Visuo-haptic environments consist a visual display simulated scene and force-feedback device haptic interaction. Here, we address visualization deformations soft tissue in real-time. A finite differences method is to calculate inverse displacement fields deform volumetric image data liver puncturing scenario. Real patient CT-image was this. diffusive linear-elastic formulation propagation field...

A system for the fully automatic segmentation of liver and spleen is presented. In a multi-atlas based framework, several existing segmentations are deformed in parallel to image intensity registrations targeting unseen patient. new locally adaptive label fusion method presented as core this paper. patch comparison approach, transformed compared weak target organ The roughly estimates hidden truth. Traditional approaches just rely on expert only. result confidence weight neighboring...

Palpation is the first step for many medical interventions. To provide an immersive virtual training and planning environment, palpation has to be successfully modeled simulated. Here, we present a multiproxy approach that calculates friction surface resistance forces multiple contact points on finger tips or tools like ultrasound probes displays resulting force torque 6DOF haptic device. No manual time intensive segmentation of patient image data needed create simulation based CT thus our...

A two-step concept for 3D segmentation on 5 abdominal organs inside volumetric CT images is presented.First each relevant organ's volume of interest extracted as bounding box.The acts input a second stage, wherein two compared U-Nets with different architectural dimensions re-construct an organ label mask.In this work, we focus comparing 2D vs. U-Net counterparts.Our initial results indicate Dice improvements about 6% at maximum.In study to our surprise, liver and kidneys instance were...

Accurate and reliable segmentation of catheters in MR-guided interventions remains a challenge, step critical importance clinical workflows. In this work, under reasonable assumptions, mechanical model based heuristics guide the process allows correct catheter identification rates greater than 98% (error 2.88 mm), reduction outliers to one-fourth compared state art. Given distal tips, searching towards proximal ends is guided by models that are estimated on per-catheter basis. Their bending...

A two-step concept for 3D segmentation on 5 abdominal organs inside volumetric CT images is presented. First each relevant organ's volume of interest extracted as bounding box. The acts input a second stage, wherein two compared U-Nets with different architectural dimensions re-construct an organ label mask. In this work, we focus comparing 2D vs. U-Net counterparts. Our initial results indicate Dice improvements about 6\% at maximum. study to our surprise, liver and kidneys instance were...

Current virtual reality (VR) training simulators of liver punctures often rely on static 3D patient data and use an unrealistic (sinusoidal) periodic animation the respiratory movement. Existing methods for breathing motion support simple mathematical or patient-specific, estimated models. However with personalized models each new patient, a heavily dose relevant expensive 4D acquisition is mandatory keyframe-based modeling. Given reference data, first model building stage using linear...

Real-time surgical simulation relies on the fast computation of soft tissue deformations. In this paper, we present image-based algorithms for computing deformations a volumetric image during needle insertion in real-time. The are based diffusive and linear elastic finite difference methods as utilized registration. For an evaluation, compared to element pre-puncture phase insertion. Furthermore, improved tested; contrary our assumption, diffusion approach outperforms approach. used perform...