Ich erkläre hiermit an Eides statt, dass ich die vorliegende kumulative Dissertation mit dem Titel The Role of Image-Guided Robotic Radiosurgery in the Management of Glomus Jugulare Tumors selbständig verfasst, mich außer der angegebenen keiner weiteren Hilfsmittel bedient und alle Erkenntnisse, die aus dem Schrifttum ganz oder annähernd übernommen sind, als solche kenntlich gemacht und nach ihrer Herkunft unter Bezeichnung der Fundstelle einzeln nachgewiesen habe. Ich erkläre des Weiteren, dass die hier vorgelegte kumulative Dissertation nicht in gleicher oder in ähnlicher Form bei einer anderen Stelle zur Erlangung eines akademischen Grades eingereicht wurde.
Paradoxically, remote military interventions are, by their very nature, restricted to the closed system of computerised, heavily mediated environments of control. The potential for crisis in the cooperation of human and non-human actors is nowhere more apparent than in the ground control sta- tions (GCS) in remotely-controlled drone warfare. A GCS provides image-guided control over the deployment of so-called unmanned weapon systems. It forms the central operative unit for decision-making and action in remote warfare by linking human perception to the sensor techno- logy of the drone. It is, thus, a crucial component of a control setting, where questions of agency culminate in a distinctly
Technological advances continue to revolutionize medical care by enabling earlier diagnoses, safer treatments and more and more sophisticated surgical interventions. Enhancements of minimally invasive surgery allow for access to difficult-to- reach anatomy, shorter hospital stays, and less pain. Yet, the trade-off for these minimal invasive approaches involves decreased visibility and impaired spatial orientation for the surgeon. This is especially the case if two-dimensional endoscopic images without any depth cues serve as the main source of information for navigating through the patient’s anatomy. As a consequence, the surgeon has to rely on his general anatomical knowledge and also a little bit of guesswork to determine the location of vital structures. Yet, the development of image-guided navigation (IGN) systems bears the potential to facilitate this demanding task essentially. IGN enables the surgeon during ongoing surgery to pinpoint the position of surgical instruments in relation to the patient’s anatomy. Core elements of IGN systems include a registration device to align the patient’s current location with preoperative images (e.g., based on computer tomography) and a tracking device (based on an electromagnetic or optical camera) to allow for an intra-operative tracking of surgical instruments as well as the patient’s position. The actual position of the tip of the instrument is then displayed on a screen in relation to the patient’s anatomy. Whenever there emerges any uncertainty about the current localization, the surgeon just needs to inspect this screen. Hence, from a human factors point of view IGN can be referred to as automation of one of the main tasks of the surgeon: The spatial localization of the surgical instrument within the three-dimensional space of the patient’s anatomy is delegated partially to a computer-based system (Manzey, Strauss, Trantakis, Lueth, Roettger, Bahner-Heyne et al., in press). According to the framework model of Parasuraman, Sheridan, and Wickens (2000), IGN systems can be classified as a low to medium level of information automation: The systems support to some degree the information acquisition and analysis of visual information by providing additional
After radiochemotherapy and image-guided adaptive brachytherapy ~98% of the patients achieve a complete remission. The clinical evidence for IGABT is more and more increasing but still limited to mainly retrospective cohort analyses (Pötter et al. 2007; Potter et al. 2011; Lindegaard et al. 2013; Rijkmans et al. 2014). Prospective randomized controlled studies are currently not available. The retrospective multi-centre study “RETRO-EMBRACE” provides currently the most advanced clinical results in IGABT for cervical cancer. The study comprises 731 patients (FIGO IA/IB/IIA: 22.8%, IIB: 50.4%, IIIA-IVB: 26.8%) from 12 international centres. After a median follow-up time of 43 months the overall actuarial 3-year local tumour control rate was 91%, the cancer-specific survival was 79% and the overall survival 74%. The actuarial 3-year local tumour control rates for FIGO stage IB, IIB and IIIB were 98%, 93% and 79%, respectively. G3-G5 morbidity after 5 years for urinary bladder, the gastrointestinal tract and the vagina was observed in 5%, 7% and 5%, respectively (Sturdza et al. 2016). A sub-analysis of the RETRO-EMBRACE study revealed that patients with a CTV HR ≥ 30cm 3 treated with combined intracavitary/interstitial techniques experienced
cal significance other than DCIS and invasive carcinoma, such as atypical ductal or lobular hyperplasia.
In conclusion, we found an upgrade rate to carcinoma at exci- sion for IDPs diagnosed with image-guided biopsy of 16.1%. Le- sions with an upgrade to malignancy after excision were signifi- cantly associated with a greater distance to the nipple. Our data suggest that observation only might not be adequate for the man- agement of all patients with a diagnosis of IDP on biopsy. For pa- tients with peripheral IDP, surgical excision should be considered.
The accuracy of minimally invasive stabilization techniques has been significantly improved by the implementation of imageguided technologies [ 5 , 8 , 9 , 11 , 12 ]. In addition, optical tracking systems have been established [ 10 , 11 ]. However, these are surface based and the required usage of flexible instruments may not be applicable in case of navigation over long distances. Another major disadvantage of this method is its susceptibility to masking or shielding of the cameras by instruments or the surgeon himself (line of sight effect) [ 13 – 17 ]. “Electromagnetic Track- ing” (EMT) offers a method that allows real time navigation also in subjacent parts of the body. Furthermore, it avoids any radiation during the surgical procedure [ 13 , 18 ]. However, potential interference factors on the electromagnetic field avoided the widespread clinical use of this technique for a long time [ 13 , 15 , 16 , 18 , 19 ]. New technologies caused an improvement of these drawbacks. Also, automatic error detection and the necessity of smaller electromagnetic fields offer new options for the application [ 15 , 16 ].
The outcome of radiation therapy is, amongst other reasons, limited by the clinical uncertainties such as the precise definition of geometric target boundaries and problems in precise delivery of radiation dose to the clinical target volume. In practice, large uncertainties exist in tumor delineation and in target localization due to intra- and inter-organ motions, especially since the dose in radiotherapy is delivered in the course of several fractions. These uncertainties affect the safety margin to ensure coverage of the target tumor. Recently, the advances in imaging technology and integration of these imaging technologies in the treatment machine have resulted in an increase of the accuracy and precision of treatment. Furthermore, the normal tissue complication (63) should be minimized. Several developments of radiation therapy technology contributed to this development in two last decades. Mainly, the development of 3D conformal radiation therapy (3D-CRT) and intensity-modulated radiation therapy (IMRT) has resulted in an increased accuracy of beam targeting. It has become possible to create highly conformal distributions for dose escalation and normal tissue avoidance (27; 63). Furthermore, target localization in the treatment room is performed regularly using two-dimensional (2D) megavoltage (MV) portal films and electronic portal imaging de- vices (EPID) (33; 67; 104). Unfortunately, the benefits of IMRT are limited when large margins have to be used to accommodate uncertainties in target and normal tissue po- sition at the time of treatment. However, in locations where day-to-day inter fractional organ motion is present, imageguided therapy offers the possibility for reduced margin and dose escalation.
Objective To prospectively compare the interobserver variability of combined transrectal ultrasound (TRUS)/computed tomography (CT)- vs. CT only- vs. magnetic resonance imaging (MRI) only-based contouring of the high-risk clinical target volume (CTV HR ) in image-guided adaptive brachytherapy (IGABT) for locally advanced cervical cancer (LACC). Methods Five patients with LACC (FIGO stages IIb–IVa) treated with radiochemotherapy and IGABT were included. CT, TRUS, and T2-weighted MRI images were performed after brachytherapy applicator insertion. 3D-TRUS image acquisition was performed with a customized ultrasound stepper device and software. Automatic applicator reconstruction using optical tracking was performed in the TRUS dataset and TRUS and CT images were fused with rigid image registration with the applicator as reference structure. The CTV HR (based on the GEC-ESTRO recommendations) was
For patients with locally advanced cervical cancer (LACC), concurrent radiochemotherapy including brachytherapy (BT) is the standard therapy, leading to good oncologic results [ 1 – 3 ]. The reported absolute risk of developing grade 3–4 (G3–4) genitourinary (GU) or gastrointestinal (GI) fistulae through definitive radiation therapy including standard BT with dose prescription to point A is around 9% (overall G3–4 toxicity 30%) [ 4 ]. Within the last few years, modern radiotherapy techniques have been introduced into the treatment of LACC. These include magnetic resonance (MR) image-guided adaptive brachytherapy (IGABT) [ 5 ] with dose prescription individualised to the target at the time of BT. In a recently published cohort of 731 patients, the overall actuarial rate of G3–4 GI and GU toxicity af- ter definitive radiochemotherapy and IGABT was 11% at 5 years [ 6 , 18 ]. A previous publication from our centre on 156 patients in whom systematic IGABT was performed reported only 2/156 fistula events [ 5 ]. Despite excellent overall local tumour control results, several series [ 5 , 7 , 8 ] still observed relatively high rates of distant recur- rences after IGABT, especially in patients with advanced stage (stage III/IV) and/or nodal disease [ 6 ]. These distant metastases require cytotoxic treatment.
In the last years, proton beam therapy has become one of the most advanced forms of cancer treatment. 1,2 The ability to reduce radiation exposure to adjacent healthy tissue and to spare the tissue behind the tumor almost entirely makes the treatment optimal for tumors where conventional photon- based radiotherapy results are limited. 3–6 Today’s state-of- the-art medical accelerators for proton beams are equipped with X-ray-based devices for image-guided radiation therapy (IGRT), aiming to reduce geometric uncertainties during radiation delivery and to enhance the target conformity of the delivered dose. 7–9 Limitations on the soft-tissue contrast of
The purpose of the thesis was to perform research assisting in implementation of ImageGuided Adaptive Radiotherapy (IGART) in the routine clinical RT treatment process at the Depart- ment of Radiotherapy, Medical University of Vienna. The aims of the thesis address specifically physical and clinical aspects of IGART for several patient groups. Following clinical sites, where combination of sophisticated treatment techniques with novel imaging modalities can bring sig- nificant improvement, were proposed: head-and-neck, prostate and lung tumors. Physical aspects described here refer mainly to the accuracy of commercial dose calculation algorithms available in modern TPS and their applicability to dose re-calculation on kV-CBCT images. The clinical aspects under investigation refer to the interobserver variability for different tumors (prostate, lung) and its consequences for ART implementation. The motivation for the thesis (sources of errors in RT) and general background of CBCT-based IGART are given in the Chapter 1. Chapter 2 is dedicated to the comparison of Monte Carlo (MC) based dose calculation vs. ad- vanced kernel algorithms, performed in various phantoms. As the ART process requires multiple treatment planning sessions per patient case, the error introduced by the dose calculation should be minimal, especially for inhomogeneous treatment regions. On the other hand, practical rea- sons imply tight time gaps available for re-planning. The dose calculation accuracy provided by commercial implementation of MC techniques was evaluated in the following way. Dose calcula- tion was initially compared with relative and absolute dosimetry measurements obtained during basic beam data acquisition and TPS commissioning process. Secondly, several square and rect- angular fields were evaluated in homogeneous and heterogeneous phantom settings based on 1D gamma evaluation of depth-dose curves and profiles from EBT film dosimetry. Finally, confor- mal and IMRT plans were compared with film measurements in heterogeneous phantoms by 2D gamma analysis. The presented data showed that commercially available MC algorithms (XVMC implementations in Monaco/Elekta-CMS and iPlan/BrainLab TPS) are slightly superior to the advanced kernel techniques (collapsed cone algorithm in Oncentra MasterPlan/Nucletron and AAA in Eclipse/Varian) with regard to heterogeneous media. However, advanced kernel tech- niques are more advantageous in terms of calculation speed.
An aim for the future should be the adaption not only according to anatomical changes, but also to biological changes in cancers. It is well known, that most tumors are heterogeneous entities. For example, hypoxic subregions of a tumor may be more radioresistant than other parts. Repeated functional imaging with MR and/or PET during therapy may enable response assessment and subsequent treatment adaption to the dynamic tumor biology through dose painting [196,197]. However, several obstacles have hindered the clinical adoption of this concept up until now. Multiple radiotracers have been developed and tested, but the quantitative relationships between PET image intensities and biological parameters like radiosensitivity and hypoxia are not fully established yet. Additional uncertainties come from image registration errors between MR, PET and CT images from different time points, which may be reduced through hybrid PET/MR-scanners and DIR. It has been demonstrated that spatio-temporal changes of tumor hypoxia occur during fractionated radiotherapy, which may pose a challenge for dose painting [198,199]. The role of functional MRI sequences like diffusion-weighted imaging and dynamic contrast-enhanced imaging in biologically guided RT is increasingly investigated, but improvements in terms of quantification and reproducibility are still needed .
MRI guidance in external beam therapy is currently being investigated to improve the current clinical outcomes. Better delineation and more precise description of anatomical variations may have a huge impact on particle therapy, where steep dose gradients are commonly observed around the tumor volume. However, the implementation of such a system, combining an MR core and a particle beam line, proved to be extremely challenging. From a technical point of view, there are still many open issues on how to integrate both systems fulfilling clinical requirements. In addition, computer-assisted workflows will be required for on-line adaptative MRgPT, combining fast and accurate algorithms for treatment planning, monitoring and adaptation. As electromagnetic interactions are present between the MR scanner and the particle beam delivery system, a wide range of MC simulations are foreseen to describe their global effect on beam delivery, image quality and dose calculation. Some of the expected pitfalls were investigated and discussed through this work, focusing on the dosimetric impact of magnetic fields on current treatment planning strategies for proton beam therapy.
As recently described by Sanhai et al., one of the top priorities in nanomedicine is determination of nanoparticle biodistribution. Concerning kinetics of biodistributional changes, visualization over time or at certain timepoints in the same animal is crucial, and quantification of changes on a mass-balance basis is so far only provided by nuclear imaging (160). PET and SPECT are routinely used for diagnosis of cancer, inflammation, cardiac disease or neurological disorders (161). While the first small animal PET scans were of poor spatial resolution, today’s small animal PET scanners with a resolution of 1 mm (162) even outrange pinhole-based small animal SPECT imaging. SPECT or gamma camera imaging detects gamma photons or X-rays of radionuclides that decay by electron capture or isomeric transition (163). A sodium iodide crystal in SPECT scanners detects the distribution of γ-ray emission, converts it into an electrical signal and eventually into an image. By (pinhole-)collimation between radioactive signal and detector, scattered radiation can be decreased. The γ-radiation energy of the radionuclide used needs to be high enough (80-250 keV) to penetrate the body in order to reach the detector, while energy can be decreased in small-animal imaging. The great advantage of SPECT is that various radionuclides with different photon energies can be acquired simultaneously in dual-isotope SPECT images. Sensitivity of SPECT is partly decreased due to the absorption of lead collimators, allowing only about 0.1 % of the γ-radiation to be detected. Therefore, high radiation doses and long acquisition times are needed for SPECT imaging (163). PET, on the other hand, uses radionuclides that emit positrons by β + decay which annihilate ubiquitous electrons, leading to conversion into two gamma photons of 511 keV. The detection of photons can either be instrumented with photon sensitive crystals or with detectors based on multi-wire gas chambers, where the photons are converted into electrons, which are multiplied in the gas and detected at the anode (164). The list of suitable positron emitters is not as broad as the one of gamma emitters. Additionally, PET imaging is more expensive, especially due to the high energy radionuclides, and can not be performed over several days because of their short half-lives.
Keywords: Image-based localization, Mobile navigation, Structure from motion, 3D point clouds, Outlier removal
Due to the rapid growth of technologies, pedestrian nav- igation has become widely accessible in the recent years. In the developed countries, smartphones are no longer considered as luxury items and are owned by the major- ity of population. The sensors installed in modern mobile devices, such as global positioning system (GPS) receiver, accelerometer, compass, gyroscope, and camera, provide a broad field of methods that can be applied for mobile navigation.
Abbreviations. Comparison: non, non-IGS; PB, pointer-based; IN, instrument navigation; PV, process visualization; UV, uncertainty visualization; DV, distance visualization; PW, proximity warning; ARS, augmented risk structures; ARS+, augmented rick structures and conventional endoscopic video display; ATV, augmented target volume; AT, augmented target; 3DV, three-dimensional virtual image guidance; ID, instrument disablement; MR, movement restriction; STS, semiautomatic trepanation system. App (Application): PS/ASB, paranasal sinuses and anterior skull base; TB/LSB, temporal bone and lateral skull base; OS, orbital surgery; NS, neurosurgery; SS, spinal surgery. SE (study environment): p, patients; c, cadavers; s, phantoms (simulation); a, animal. n (sample size/number of participating surgeons): n.i., not indicated. Experience: R, residents; F, fellows; E, experienced; U, unexperienced; MS, medical students; n.i., not indicated. Comments: a non-surgery – participants were evaluated on non-surgical day (control day); b process visualization through
At the beginning of the image interaction task, a dimmed user interface indicated the inactive state of the system. After activation with the given input modality, the grid of four equally sized viewports was displayed and the primary task started for group A. At first, a head-controlled cursor had to be used to select one of the viewports using a selection input method (voice or head gestures). The viewport’s border turned blue upon selection and panning and zooming mode was entered immediately. Each viewport accommodated geometric patterns in light gray, a small filled black shape and a corresponding, more prominent outline at the center. The small shape had to be translated and zoomed in such a way that it matches the outline. This had to be done using an image manipulation technique (facial expression or leaning). During panning and zooming, the head cursor was not displayed anymore as the head movement was used for panning the image at this point. When the filled shape matched the contour, it turned blue, indicating completion of this subtask. The viewport had to be exited using the current selection method. After exiting a viewport, the head-controlled cursor was displayed again to select the next viewport. After panning and zooming all of the viewport contents correctly, the system had to be deactivated. The user interface during these states can be seen in Figure 6.15.
S3 Text. Depth Dose Rate Interpolation. (PDF)
S1 Fig. Micro-CT and equipment used in this study. (a) Industrial X-ray system YXLON Y. Fox (YXLON International GmbH, Hamburg, Germany). (b) The X-ray tube (A) is mounted at a fixed position on the top of the system. The beam directs towards the manipulator (B), which can be moved in three axes (x,y,z) and rotated by 360°. The beam is detected by a 12-bit direct digital flat panel detector (C). (c) This image shows the X-ray tube outlet with a ring (D) holding the target (E) in place when the system is evacuated. The lower picture shows the ani- mal couch with anesthesia support (F) and ear bar fixations (G). (d) Mounted base plate (left side) and collimator (right side).