Reattachable fiducial skin marker for automatic multimodality registration.

PURPOSE: Fusing image information has become increasingly important for optimal diagnosis and treatment of the patient. Despite intensive research towards markerless registration approaches, fiducial marker-based methods remain the default choice for a wide range of applications in clinical practice. However, as especially non-invasive markers cannot be positioned reproducibly in the same pose on the patient, pre-interventional imaging has to be performed immediately before the intervention for fiducial marker-based registrations. METHODS: We propose a new non-invasive, reattachable fiducial skin marker concept for multi-modal registration approaches including the use of electromagnetic or optical tracking technologies. We furthermore describe a robust, automatic fiducial marker localization algorithm for computed tomography (CT) and magnetic resonance imaging (MRI) images. Localization of the new fiducial marker has been assessed for different marker configurations using both CT and MRI. Furthermore, we applied the marker in an abdominal phantom study. For this, we attached the marker at three poses to the phantom, registered ten segmented targets of the phantom's CT image to live ultrasound images and determined the target registration error (TRE) for each target and each marker pose. RESULTS: Reattachment of the marker was possible with a mean precision of 0.02 mm ± 0.01 mm. Our algorithm successfully localized the marker automatically in all ([Formula: see text]) evaluated CT/MRI images. Depending on the marker pose, the mean ([Formula: see text]) TRE of the abdominal phantom study ranged from 1.51 ± 0.75 mm to 4.65 ± 1.22 mm. CONCLUSIONS: The non-invasive, reattachable skin marker concept allows reproducible positioning of the marker and automatic localization in different imaging modalities. The low TREs indicate the potential applicability of the marker concept for clinical interventions, such as the puncture of abdominal lesions, where current image-based registration approaches still lack robustness and existing marker-based methods are often impractical.

Deep learning-based classification of DSA image sequences of patients with acute ischemic stroke.

PURPOSE: Recently, a large number of patients with acute ischemic stroke benefited from the use of thrombectomy, a minimally invasive intervention technique for mechanically removing thrombi from the cerebrovasculature. During thrombectomy, 2D digital subtraction angiography (DSA) image sequences are acquired simultaneously from the posterior-anterior and the lateral view to control whether thrombus removal was successful, and to possibly detect newly occluded areas caused by thrombus fragments split from the main thrombus. However, such new occlusions, which would be treatable by thrombectomy, may be overlooked during the intervention. To prevent this, we developed a deep learning-based approach to automatic classification of DSA sequences into thrombus-free and non-thrombus-free sequences. METHODS: We performed a retrospective study based on the single-center DSA data of thrombectomy patients. For classifying the DSA sequences, we applied Long Short-Term Memory or Gated Recurrent Unit networks and combined them with different Convolutional Neural Networks used as feature extractor. These network variants were trained on the DSA data by using five-fold cross-validation. The classification performance was determined on a test data set with respect to the Matthews correlation coefficient (MCC) and the area under the curve (AUC). Finally, we evaluated our models on patient cases, in which overlooking thrombi during thrombectomy had happened. RESULTS: Depending on the specific model configuration used, we obtained a performance of up to 0.77[Formula: see text]0.94 for the MCC[Formula: see text]AUC, respectively. Additionally, overlooking thrombi could have been prevented in the reported patient cases, as our models would have classified the corresponding DSA sequences correctly. CONCLUSION: Our deep learning-based approach to thrombus identification in DSA sequences yielded high accuracy on our single-center test data set. External validation is now required to investigate the generalizability of our method. As demonstrated, using this new approach may help reduce the incident risk of overlooking thrombi during thrombectomy in the future.

Polhemus EM tracked Micro Sensor for CT-guided interventions.

PURPOSE: Electromagnetic (EM) tracking is a key technology in image-guided therapy. A new EM Micro Sensor was presented by Polhemus Inc.; it is the first to enable localization of medical instruments through their trackers. Different field generators (FGs) are available by Polhemus, one being almost as small as a sugar cube. As accuracy and robustness of tracking are known challenges to using EM trackers in clinical environments, the goal of this study was a standardized assessment of the Micro Sensor in both a laboratory (lab) and a computed tomography (CT) environment. METHODS: The Micro Sensor was assessed by means of Hummel et al.'s standardized protocol; it was assessed in conjunction with a Polhemus Liberty tracker and three FGs - with edge lengths of 1 (TX1), 2 (TX2), and 4 (TX4) inches. Precision as well as positional and rotational accuracy were determined in a lab and a CT suite. Distortions by four different metallic cylinders and tracking of two typical medical instruments - a hypodermic needle and a flexible endoscope - were also tested. RESULTS: A jitter of 0.02 mm or less was found for all FGs in the different environments, except for the TX2 FG for which no valid data could be obtained in the CT. Errors of 5 cm distance measurements were 0.6 mm or less for all FGs in the lab. While the distance errors of the TX1 FG were only slightly increased up to 1.6 mm in the CT, those of the TX4 FG were found to be up to around 10% of the measured distance (5.4 mm on average). The mean orientation error was found to be 0.9° /0.5° /0.1° for the TX4/TX2/TX1 FG in the lab. In the CT environment, rotation errors were in the same range: less than 1.2° /0.1° for the TX4/TX1 FG. Deviation under the presence of metallic cylinders stayed below 1 mm in most cases. Precision and orientational accuracy do not seem to be affected by instrument tracking and stayed in the same range as for the other measurements whereas distance errors were slightly increased up to 1.7 mm. CONCLUSION: This study shows that accurate tracking of medical instruments is possible with the new Micro Sensor; it demonstrated a jitter of 0.01 mm or less, position errors below 2 mm, and rotation errors of less than 0.3° . As with other EM trackers, errors increase when large tracking volumes with ranges of up to 50 cm are required in clinical environments. For smaller tracking volumes with ranges of up to 15 cm, a high accuracy and robustness was found. This is interesting especially for the TX1 FG which can easily be placed in close vicinity to the region of interest.

MITK-OpenIGTLink for combining open-source toolkits in real-time computer-assisted interventions.

PURPOSE: Due to rapid developments in the research areas of medical imaging, medical image processing and robotics, computer-assisted interventions (CAI) are becoming an integral part of modern patient care. From a software engineering point of view, these systems are highly complex and research can benefit greatly from reusing software components. This is supported by a number of open-source toolkits for medical imaging and CAI such as the medical imaging interaction toolkit (MITK), the public software library for ultrasound imaging research (PLUS) and 3D Slicer. An independent inter-toolkit communication such as the open image-guided therapy link (OpenIGTLink) can be used to combine the advantages of these toolkits and enable an easier realization of a clinical CAI workflow. METHODS: MITK-OpenIGTLink is presented as a network interface within MITK that allows easy to use, asynchronous two-way messaging between MITK and clinical devices or other toolkits. Performance and interoperability tests with MITK-OpenIGTLink were carried out considering the whole CAI workflow from data acquisition over processing to visualization. RESULTS: We present how MITK-OpenIGTLink can be applied in different usage scenarios. In performance tests, tracking data were transmitted with a frame rate of up to 1000 Hz and a latency of 2.81 ms. Transmission of images with typical ultrasound (US) and greyscale high-definition (HD) resolutions of [Formula: see text] and [Formula: see text] is possible at up to 512 and 128 Hz, respectively. CONCLUSION: With the integration of OpenIGTLink into MITK, this protocol is now supported by all established open-source toolkits in the field. This eases interoperability between MITK and toolkits such as PLUS or 3D Slicer and facilitates cross-toolkit research collaborations. MITK and its submodule MITK-OpenIGTLink are provided open source under a BSD-style licence ( http://mitk.org ).

Towards markerless navigation for percutaneous needle insertions.

PURPOSE: Percutaneous needle insertions are increasingly used for diagnosis and treatment of abdominal lesions. The challenging part of computed tomography (CT)-guided punctures is the transfer of the insertion trajectory planned in the CT image to the patient. Conventionally, this often results in several needle repositionings and control CT scans. To address this issue, several navigation systems for percutaneous needle insertions have been presented; however, none of them has thus far become widely accepted in clinical routine. Their benefit for the patient could not exceed the additional higher costs and the increased complexity in terms of bulky tracking systems and specialized markers for registration and tracking. METHODS: We present the first markerless and trackerless navigation concept for real-time patient localization and instrument guidance. It has specifically been designed to be integrated smoothly into the clinical workflow and does not require markers or an external tracking system. The main idea is the utilization of a range imaging device that allows for contactless and radiation-free acquisition of both range and color information used for patient localization and instrument guidance. RESULTS: A first feasibility study in phantom and porcine models yielded a median targeting accuracy of 6.9 and 19.4 mm, respectively. CONCLUSIONS: Although system performance remains to be improved for clinical use, expected advances in camera technology as well as consideration of respiratory motion and automation of the individual steps will make this approach an interesting alternative for guiding percutaneous needle insertions.

Electromagnetic tracking in medicine--a review of technology, validation, and applications.

Object tracking is a key enabling technology in the context of computer-assisted medical interventions. Allowing the continuous localization of medical instruments and patient anatomy, it is a prerequisite for providing instrument guidance to subsurface anatomical structures. The only widely used technique that enables real-time tracking of small objects without line-of-sight restrictions is electromagnetic (EM) tracking. While EM tracking has been the subject of many research efforts, clinical applications have been slow to emerge. The aim of this review paper is therefore to provide insight into the future potential and limitations of EM tracking for medical use. We describe the basic working principles of EM tracking systems, list the main sources of error, and summarize the published studies on tracking accuracy, precision and robustness along with the corresponding validation protocols proposed. State-of-the-art approaches to error compensation are also reviewed in depth. Finally, an overview of the clinical applications addressed with EM tracking is given. Throughout the paper, we report not only on scientific progress, but also provide a review on commercial systems. Given the continuous debate on the applicability of EM tracking in medicine, this paper provides a timely overview of the state-of-the-art in the field.

The Medical Imaging Interaction Toolkit: challenges and advances : 10 years of open-source development.

PURPOSE: The Medical Imaging Interaction Toolkit (MITK) has been available as open-source software for almost 10 years now. In this period the requirements of software systems in the medical image processing domain have become increasingly complex. The aim of this paper is to show how MITK evolved into a software system that is able to cover all steps of a clinical workflow including data retrieval, image analysis, diagnosis, treatment planning, intervention support, and treatment control. METHODS: MITK provides modularization and extensibility on different levels. In addition to the original toolkit, a module system, micro services for small, system-wide features, a service-oriented architecture based on the Open Services Gateway initiative (OSGi) standard, and an extensible and configurable application framework allow MITK to be used, extended and deployed as needed. A refined software process was implemented to deliver high-quality software, ease the fulfillment of regulatory requirements, and enable teamwork in mixed-competence teams. RESULTS: MITK has been applied by a worldwide community and integrated into a variety of solutions, either at the toolkit level or as an application framework with custom extensions. The MITK Workbench has been released as a highly extensible and customizable end-user application. Optional support for tool tracking, image-guided therapy, diffusion imaging as well as various external packages (e.g. CTK, DCMTK, OpenCV, SOFA, Python) is available. MITK has also been used in several FDA/CE-certified applications, which demonstrates the high-quality software and rigorous development process. CONCLUSIONS: MITK provides a versatile platform with a high degree of modularization and interoperability and is well suited to meet the challenging tasks of today's and tomorrow's clinically motivated research.

MITK-ToF--range data within MITK.

PURPOSE: The time-of-flight (ToF) technique is an emerging technique for rapidly acquiring distance information and is becoming increasingly popular for intra-operative surface acquisition. Using the ToF technique as an intra-operative imaging modality requires seamless integration into the clinical workflow. We thus aim to integrate ToF support in an existing framework for medical image processing. METHODS: MITK-ToF was implemented as an extension of the open-source C++ Medical Imaging Interaction Toolkit (MITK) and provides the basic functionality needed for rapid prototyping and development of image-guided therapy (IGT) applications that utilize range data for intra-operative surface acquisition. This framework was designed with a module-based architecture separating the hardware-dependent image acquisition task from the processing of the range data. RESULTS: The first version of MITK-ToF has been released as an open-source toolkit and supports several ToF cameras and basic processing algorithms. The toolkit, a sample application, and a tutorial are available from http://mitk.org. CONCLUSIONS: With the increased popularity of time-of-flight cameras for intra-operative surface acquisition, integration of range data supports into medical image processing toolkits such as MITK is a necessary step. Handling acquisition of range data from different cameras and processing of the data requires the establishment and use of software design principles that emphasize flexibility, extendibility, robustness, performance, and portability. The open-source toolkit MITK-ToF satisfies these requirements for the image-guided therapy community and was already used in several research projects.

Convergent iterative closest-point algorithm to accomodate anisotropic and inhomogenous localization error.

Since its introduction in the early 1990s, the Iterative Closest Point (ICP) algorithm has become one of the most well-known methods for geometric alignment of 3D models. Given two roughly aligned shapes represented by two point sets, the algorithm iteratively establishes point correspondences given the current alignment of the data and computes a rigid transformation accordingly. From a statistical point of view, however, it implicitly assumes that the points are observed with isotropic Gaussian noise. In this paper, we show that this assumption may lead to errors and generalize the ICP such that it can account for anisotropic and inhomogenous localization errors. We 1) provide a formal description of the algorithm, 2) extend it to registration of partially overlapping surfaces, 3) prove its convergence, 4) derive the required covariance matrices for a set of selected applications, and 5) present means for optimizing the runtime. An evaluation on publicly available surface meshes as well as on a set of meshes extracted from medical imaging data shows a dramatic increase in accuracy compared to the original ICP, especially in the case of partial surface registration. As point-based surface registration is a central component in various applications, the potential impact of the proposed method is high.

Navigated liver biopsy using a novel soft tissue navigation system versus CT-guided liver biopsy in a porcine model: a prospective randomized trial.

RATIONALE AND OBJECTIVES: The aim of this prospective, randomized animal study was to compare a new computer guided needle-based navigation system for liver biopsy with conventional computed tomography (CT)-guided liver biopsy. Computer-navigated interventions provide continuous needle tracking during motion and deformation from patient respiration and movement. MATERIALS AND METHODS: Twenty artificial tumors of about 5 mm in diameter were injected into the livers of five pigs, each at a different site. Each tumor was targeted by conventional CT-guided and computer navigated intervention. Intervention was considered complete after successful tumor biopsy. Data on procedure time, number of CT scans performed, accuracy, and success rate were recorded. RESULTS: All tumors (100%) were biopsied successfully. Mean procedural time was comparable between the two techniques (20 ± 9 minutes conventional versus 20 ± 8 minutes navigation). Mean number of CT scans were 1.2 ± 0.4 with navigation and 6.1 ± 3.8 with the conventional technique (P < .01). The dose-length product in the conventional group was significantly higher (212 ± 116 mGy × cm) than in the navigated group (78 ± 22 mGy × cm; P < .001). Mean number of capsule penetrations was 4 ± 1 with navigation versus 2 ± 1 with the conventional technique (P < .001). CONCLUSION: Computer-navigated liver biopsy may provide a promising and innovative device for easy, rapid, and successful liver biopsies with low morbidity. Further technical improvements and clinical studies in humans are required.

RepliExplore: coupling physical and virtual anatomy models.

PURPOSE: We present a system which co-registers physical anatomy models with virtual three-dimensional (3D) representations. Interactions performed on the physical model by means of a 3D pointing device are directly reflected on its virtual counterpart. Complex anatomical information integrated into the virtual model thus becomes accessible through the physical interface in a simple and intuitive manner. METHODS: Using an optical tracking system, we implemented and tested a reference application that includes several tools for the exploration and quantification of anatomical models. We theoretically evaluated the accuracy of the landmark-based registration for different landmark configurations. RESULTS: Physicians and computer scientists found the system simple to learn and intuitive to use. By optimizing landmark configurations, the accuracy could be significantly increased, particularly for scenarios in which only selected regions required higher accuracy. CONCLUSIONS: Physical anatomical models can benefit from the combination with a virtual counterpart in several ways. Applications include anatomical education and the study of patient-individual organ models. Optimizing the registration landmark configuration for specific applications can lower the accuracy requirements for the tracking system.

On combining internal and external fiducials for liver motion compensation.

This paper presents an in-vivo accuracy study on combining skin markers (external fiducials) and fiducial needles (internal fiducials) for motion compensation during liver interventions. We compared the target registration error (TRE) for different numbers of skin markers n(s) and fiducial needles n(f), as well as for different transformation types, in two swine using the tip of an additional tracked needle as the target. During continuous breathing, n(f) had the greatest effect on the accuracy, yielding mean root mean square (RMS) errors of 4.8 +/- 1.1 mm (n(f) = 0), 2.0 +/- 0.9 mm (n(f) = 1) and 1.7 +/- 0.8 mm (n(f) = 2) when averaged over multiple tool arrangements (n = 18, 36, 18) with n(s) = 4. These values correspond to error reductions of 11%, 64% and 70%, respectively, compared to the case when no motion compensation is performed, i.e., when the target position is assumed to be constant. At expiration, the mean RMS error ranged from 1.1 mm (n(f) = 0) to 0.8 mm (n(f) = 2), which is of the order of magnitude of the target displacement. Our study further indicates that the fiducial registration error (FRE) of a rigid transformation reflecting tissue motion generally correlates strongly with the TRE. Our findings could be used in practice to (1) decide on a suitable combination of fiducials for a given intervention, considering the trade-off between high accuracy and low invasiveness, and (2) provide an intra-interventional measure of confidence for the accuracy of the system based on the FRE.

In vivo accuracy assessment of a needle-based navigation system for CT-guided radiofrequency ablation of the liver.

Computed tomography (CT)-guided percutaneous radiofrequency ablation (RFA) has become a commonly used procedure in the treatment of liver tumors. One of the main challenges related to the method is the exact placement of the instrument within the lesion. To address this issue, a system was developed for computer-assisted needle placement which uses a set of fiducial needles to compensate for organ motion in real time. The purpose of this study was to assess the accuracy of the system in vivo. Two medical experts with experience in CT-guided interventions and two nonexperts used the navigation system to perform 32 needle insertions into contrasted agar nodules injected into the livers of two ventilated swine. Skin-to-target path planning and real-time needle guidance were based on preinterventional 1 mm CT data slices. The lesions were hit in 97% of all trials with a mean user error of 2.4 +/- 2.1 mm, a mean target registration error (TRE) of 2.1 +/- 1.1 mm, and a mean overall targeting error of 3.7 +/- 2.3 mm. The nonexperts achieved significantly better results than the experts with an overall error of 2.8 +/- 1.4 mm (n=16) compared to 4.5 +/- 2.7 mm (n=16). The mean time for performing four needle insertions based on one preinterventional planning CT was 57 +/- 19 min with a mean setup time of 27 min, which includes the steps fiducial insertion (24 +/- 15 min), planning CT acquisition (1 +/- 0 min), and registration (2 +/- 1 min). The mean time for path planning and targeting was 5 +/- 4 and 2 +/- 1 min, respectively. Apart from the fiducial insertion step, experts and nonexperts performed comparably fast. It is concluded that the system allows for accurate needle placement into hepatic tumors based on one planning CT and could thus enable considerable improvement to the clinical treatment standard for RFA procedures and other CT-guided interventions in the liver. To support clinical application of the method, optimization of individual system modules to reduce intervention time is proposed.