Jimmy Qiu

Improving superficial target delineation in radiation therapy with endoscopic tracking and registration

PURPOSE Target delineation within volumetric imaging is a critical step in the planning process of intensity modulated radiation therapy. In endoluminal cancers, endoscopy often reveals superficial areas of visible disease beyond what is seen on volumetric imaging. Quantitatively relating these findings to the volumetric imaging is prone to human error during the recall and contouring of the target. We have developed a method to improve target delineation in the radiation therapy planning process by quantitatively registering endoscopic findings contours traced on endoscopic images to volumetric imaging. METHODS Using electromagnetic sensors embedded in an endoscope, 2D endoscopic images were registered to computed tomography (CT) volumetric images by tracking the position and orientation of the endoscope relative to a CT image set. Regions-of-interest (ROI) in the 2D endoscopic view were delineated. A mesh created within the boundary of the ROI was projected onto the 3D image data, registering the ROI with the volumetric image. This 3D ROI was exported to clinical radiation treatment planning software. The precision and accuracy of the procedure was tested on two solid phantoms with superficial markings visible on both endoscopy and CT images. The first phantom was T-shaped tube with X-marks etched on the interior. The second phantom was an anatomically correct skull phantom with a phantom superficial lesion placed on the pharyngeal surface. Markings were contoured on the endoscope images and compared with contours delineated in the treatment planning system based on the CT images. Clinical feasibility was tested on three patients with early stage glottic cancer. Image-based rendering using manually identified landmarks was used to improve the registration. RESULTS Using the T-shaped phantom with X-markings, the 2D to 3D registration accuracy was 1.5-3.5 mm, depending on the endoscope position relative to the markings. Intraobserver standard variation was 0.5 mm. Rotational accuracy was within 2°. Using the skull phantom, registration accuracy was assessed by calculating the average surface minimum distance between the endoscopy and treatment planning contours. The average surface distance was 0.92 mm with 93% of all points in the 2D-endoscopy ROI within 1.5 mm of any point within the ROI contoured in the treatment planning software. This accuracy is limited by the CT imaging resolution and the electromagnetic (EM) sensor accuracy. The clinical testing demonstrated that endoscopic contouring is feasible. With registration based on em tracking only, accuracy was 5.6-8.4 mm. Image-based registration reduced this error to less than 3.5 mm and enabled endoscopic contouring in all cases. CONCLUSIONS Registration of contours generated on 2D endoscopic images to 3D planning space is feasible, with accuracy smaller than typical set-up margins. Used in addition to standard 3D contouring methods in radiation planning, the technology may improve gross tumour volume (GTV) delineation for superficial tumors in luminal sites that are only visible in endoscopy.

Localization of Pulmonary Nodules Using Navigation Bronchoscope and a Near-Infrared Fluorescence Thoracoscope

BACKGROUND Video-assisted thoracoscopic wedge resection of multiple small, non-visible, and nonpalpable pulmonary nodules is a clinical challenge. We propose an ultra-minimally invasive technique for localization of pulmonary nodules using the electromagnetic navigation bronchoscope (ENB)-guided transbronchial indocyanine green (ICG) injection and intraoperative fluorescence detection with a near-infrared (NIR) fluorescence thoracoscope. METHODS Fluorescence properties of ICG topically injected into the lung parenchyma were determined using a resected porcine lung. The combination of ENB-guided ICG injection and NIR fluorescence detection was tested using a live porcine model. An electromagnetic sensor integrated flexible bronchoscope was geometrically registered to the three-dimensional chest computed tomographic image data by way of a real-time electromagnetic tracking system. The ICG mixed with iopamidol was injected into the pulmonary nodules by ENB guidance; ICG fluorescence was visualized by a near-infrared (NIR) thoracoscope. RESULTS The ICG existing under 24-mm depth of inflated lung was detectable by the NIR fluorescence thoracoscope. The size of the fluorescence spot made by 0.1 mL of ICG was 10.4 ± 2.2 mm. An ICG or iopamidol spot remained at the injected point of the lung for more than 6 hours in vivo. The ICG fluorescence spot injected into the pulmonary nodule with ENB guidance was identified at the pulmonary nodule with the NIR thoracoscope. CONCLUSIONS The ENB-guided transbronchial ICG injection and intraoperative NIR thoracoscopic detection is a feasible method to localize multiple pulmonary nodules.

Navigated Pelvic Osteotomy and Tumor Resection A Study Assessing the Accuracy and Reproducibility of Resection Planes in Sawbones and Cadavers

BACKGROUND This Sawbones and cadaver study was performed to assess the accuracy and reproducibility of pelvic bone cuts made with use of a novel navigation system with a navigated osteotome and oscillating saw. METHODS Using a novel navigation system and a three-dimensional planning tool, we navigated pelvic bone cuts that were representative of typical cuts made in pelvic tumor resections. The system includes a prototype mobile C-arm for intraoperative cone-beam computed tomography, real-time optical tracking (Polaris), and three-dimensional visualization software. Three-dimensional virtual radiographs were utilized in addition to triplanar (axial, sagittal, and coronal) navigation. In part one of the study, we navigated twenty-four sacral bone cuts in Sawbones models and validated our results in sixteen similar cuts in cadavers. In part two, we developed three Sawbones models of pelvic tumors based on actual patient scenarios and compared three navigated resections with three non-navigated resections for each tumor model. Part three assessed the accuracy of the system with multiple users. RESULTS There were ninety navigated cuts in Sawbones that were compared with fifty-four non-navigated cuts. In the navigated Sawbones cuts, the mean entry and exit cuts were 1.4 ± 1 mm and 1.9 ± 1.2 mm from the planned cuts, respectively. In comparison, the entry and exit cuts in Sawbones that were not navigated were 2.8 ± 4.9 mm and 3.5 ± 4.6 mm away from the planned osteotomy site. The navigated cuts were significantly more accurate (p ≤ 0.01). In the cadaver study, navigated entry and exit cuts were 1.5 ± 0.9 mm and 2.1 ± 1.5 mm from the planned cuts. The variation among three different users was 1 mm on both the entry and exit cuts. CONCLUSIONS Navigation to guide pelvic bone cuts is accurate and feasible. Three-dimensional radiographs should be used for improved accuracy. Navigated cuts were significantly more accurate than non-navigated cuts were. A margin of 5 mm between the target tumor volume and the planned cut plane would result in a negative margin resection in more than 95% of the cuts. CLINICAL RELEVANCE The accuracy of pelvic bone tumor resections and pelvic osteotomies can be improved with navigation to within 5 mm of the planned cut.

Displaying 3D radiation dose on endoscopic video for therapeutic assessment and surgical guidance

We have developed a method to register and display 3D parametric data, in particular radiation dose, on two-dimensional endoscopic images. This registration of radiation dose to endoscopic or optical imaging may be valuable in assessment of normal tissue response to radiation, and visualization of radiated tissues in patients receiving post-radiation surgery. Electromagnetic sensors embedded in a flexible endoscope were used to track the position and orientation of the endoscope allowing registration of 2D endoscopic images to CT volumetric images and radiation doses planned with respect to these images. A surface was rendered from the CT image based on the air/tissue threshold, creating a virtual endoscopic view analogous to the real endoscopic view. Radiation dose at the surface or at known depth below the surface was assigned to each segment of the virtual surface. Dose could be displayed as either a colorwash on this surface or surface isodose lines. By assigning transparency levels to each surface segment based on dose or isoline location, the virtual dose display was overlaid onto the real endoscope image. Spatial accuracy of the dose display was tested using a cylindrical phantom with a treatment plan created for the phantom that matched dose levels with grid lines on the phantom surface. The accuracy of the dose display in these phantoms was 0.8-0.99 mm. To demonstrate clinical feasibility of this approach, the dose display was also tested on clinical data of a patient with laryngeal cancer treated with radiation therapy, with estimated display accuracy of ∼2-3 mm. The utility of the dose display for registration of radiation dose information to the surgical field was further demonstrated in a mock sarcoma case using a leg phantom. With direct overlay of radiation dose on endoscopic imaging, tissue toxicities and tumor response in endoluminal organs can be directly correlated with the actual tissue dose, offering a more nuanced assessment of normal tissue toxicities following radiation therapy and accurate registration of radiation dose to the surgical field.

Minimally Invasive Electro-Magnetic Navigational Bronchoscopy-Integrated Near-Infrared-Guided Sentinel Lymph Node Mapping in the Porcine Lung

Background The use of near-infrared (NIR) fluorescence imaging with indocyanine green (ICG) for sentinel lymph node (SN) mapping has been investigated in lung cancer; however, this has not been fully adapted for minimally invasive surgery (MIS). The aim of our study was to develop a minimally invasive SN mapping integrating pre-operative electro-magnetic navigational bronchoscopy (ENB)-guided transbronchial ICG injection and intraoperative NIR thoracoscopic imaging. Methods A NIR thoracoscope was used to visualize ICG fluorescence. ICG solutions in a 96-well plate and ex vivo porcine lungs were examined to optimize ICG concentrations and injection volumes. Transbronchial ICG injection (n=4) was assessed in comparison to a traditional transpleural approach (n=3), where after thoracotomy an ICG solution (100μL at 100μg/mL) was injected into the porcine right upper lobe for SN identification. For further translation into clinical use, transbronchial ICG injection prior to thoracotomy followed by NIR thoracoscopic imaging was validated (n=3). ENB was used for accurate targeting in two pigs with a pseudo-tumor. Results The ICG fluorescence at 10 μg/mL was the brightest among various concentrations, unchanged by the distance between the thoracoscope and ICG solutions. Injected ICG of no more than 500μL showed a localized fluorescence area. All 7 pigs showed a bright paratracheal lymph node within 15 minutes post-injection, with persistent fluorescence for 60 minutes. The antecedent transbronchial ICG injection succeeded in SN identification in all 3 cases at the first thoracoscopic inspection within 20 minutes post-injection. The ENB system allowed accurate ICG injection surrounding the pseudo-tumors. Conclusions ENB-guided ICG injection followed by NIR thoracoscopy was technically feasible for SN mapping in the porcine lung. This promising platform may be translated into human clinical trials and is suited for MIS.

Three-dimensional virtual navigation versus conventional image guidance: A randomized controlled trial

Providing image guidance in a 3‐dimensional (3D) format, visually more in keeping with the operative field, could potentially reduce workload and lead to faster and more accurate navigation. We wished to assess a 3D virtual‐view surgical navigation prototype in comparison to a traditional 2D system.

Quantitative comparison of cranial approaches in the anatomy laboratory: A neuronavigation based research method

AIM To describe the development and validation of a novel neuronavigation-based method, which allows the quantification of the anatomical features that define an approach, as well as real-time visualization of the surgical pyramid. METHODS The method was initially developed with commercially-available hardware for coordinate collection (a digitizer and a frameless navigation system) and software for volume rendering; dedicated neuronavigation software (ApproachViewer, part of GTx-UHN) was then developed. The accuracy of measurements and the possibility of volumetric rendering of surgical approaches simulated in a phantom were compared among three different methods and commercially-available radiological software. In the anatomy laboratory, ApproachViewer was applied to the comparative quantitative analysis of multiple neurosurgical approaches and was used by many surgeons who were untrained for the research method. RESULTS The accuracy of ApproachViewer is comparable to commercially-available radiological software. In the anatomy laboratory, the method appears versatile. The system can be easily used after brief training. ApproachViewer allows for real-time evaluation and comparison of surgical approaches, as well as post-dissection analyses of collected data. The accuracy of the method depends on the navigation registration: with a 1-2 mm registration error, it is adequate for evaluation and comparison of most neurosurgical approaches. CONCLUSION This new research method and software allows semi-automated visualization, quantification, and comparison of neurosurgical approaches in the anatomy laboratory.

Quantification of Working Volumes, Exposure, and Target-Specific Maneuverability of the Pterional Craniotomy and Its Minimally Invasive Variants

OBJECTIVE Quantitative comparison of minimally invasive and standard cranial approaches remains a methodologic challenge. The aim of this study was to apply a new digital, navigation-based method to quantify multiple parameters of the pterional, supraorbital, lateral supraorbital and mini-pterional approaches and to describe a target-specific maneuverability score. METHODS Supraorbital, lateral supraorbital, mini-pterional, and standard pterional craniotomies were performed on 8 sides in 4 cadaver heads. The limits of superficial and deep exposure and surgical target points were registered with image-guidance (Medtronic Stealth). A custom software (Guided Therapeutics Eyes II Software [GTxEyesII]) calculated superficial and deep surface areas, working volumes, and target distances. Volumes were reconstructed digitally and visualized with GTxEyesII. Finally, we defined for each approach a target-specific maneuverability score based on surface exposure, target distance, target position, and anatomical obstacles. RESULTS Surface exposure and volumes were significantly smaller for keyhole approaches compared with standard pterional craniotomy (P < 0.01). Maneuverability index scores showed the greatest values for standard pterional craniotomies, but nearly equally high scores could be achieved by one or several of the less-invasive approaches for most of the targets. CONCLUSIONS The combination of a navigation-based measurement method with custom software (GTxEyesII) provides a powerful tool for visualization and quantification of surgical approaches. Compared with standard pterional craniotomy, alternative keyhole craniotomies offer comparable deep exposure with smaller working volumes and relatively high maneuverability for specific targets.

Accuracy and reproducibility of virtual cutting guides and 3D-navigation for osteotomies of the mandible and maxilla.

Background We set out to determine the accuracy of 3D-navigated mandibular and maxillary osteotomies with the ultimate aim to integrate virtual cutting guides and 3D-navigation into ablative and reconstructive head and neck surgery. Methods Four surgeons (two attending, two clinical fellows) completed 224 unnavigated and 224 3D-navigated osteotomies on anatomical models according to preoperative 3D plans. The osteotomized bones were scanned and analyzed. Results Median distance from the virtual plan was 2.1 mm unnavigated (IQR 2.6 mm, ≥3 mm in 33%) and 1.2 mm 3D-navigated (IQR 1.1 mm, ≥3 mm in 6%) (P<0.0001); median pitch was 4.5° unnavigated (IQR 7.1°) and 3.5° 3D-navigated (IQR 4.0°) (P<0.0001); median roll was 7.4° unnavigated (IQR 8.5°) and 2.6° 3D-navigated (IQR 3.8°) (P<0.0001). Conclusion 3D-rendering enables osteotomy navigation. 3 mm is an appropriate planning distance. The next steps are translating virtual cutting guides to free bone flap reconstruction and clinical use.

Optically-tracked handheld fluorescence imaging platform for monitoring skin response in the management of soft tissue sarcoma

Abstract. Standard clinical management of extremity soft tissue sarcomas includes surgery with radiation therapy. Wound complications (WCs) arising from treatment may occur due to bacterial infection and tissue breakdown. The ability to detect changes in these parameters during treatment may lead to earlier interventions that mitigate WCs. We describe the use of a new system composed of an autofluorescence imaging device and an optical three-dimensional tracking system to detect and coregister the presence of bacteria with radiation doses. The imaging device visualized erythema using white light and detected bacterial autofluorescence using 405-nm excitation light. Its position was tracked relative to the patient using IR reflective spheres and registration to the computed tomography coordinates. Image coregistration software was developed to spatially overlay radiation treatment plans and dose distributions on the white light and autofluorescence images of the surgical site. We describe the technology, its use in the operating room, and standard operating procedures, as well as demonstrate technical feasibility and safety intraoperatively. This new clinical tool may help identify patients at greater risk of developing WCs and investigate correlations between radiation dose, skin response, and changes in bacterial load as biomarkers associated with WCs.