Alborz Amir-Khalili

Automatic segmentation of occluded vasculature via pulsatile motion analysis in endoscopic robot-assisted partial nephrectomy video

Hilar dissection is an important and delicate stage in partial nephrectomy, during which surgeons remove connective tissue surrounding renal vasculature. Serious complications arise when the occluded blood vessels, concealed by fat, are missed in the endoscopic view and as a result are not appropriately clamped. Such complications may include catastrophic blood loss from internal bleeding and associated occlusion of the surgical view during the excision of the cancerous mass (due to heavy bleeding), both of which may compromise the visibility of surgical margins or even result in a conversion from a minimally invasive to an open intervention. To aid in vessel discovery, we propose a novel automatic method to segment occluded vasculature from labeling minute pulsatile motion that is otherwise imperceptible with the naked eye. Our segmentation technique extracts subtle tissue motions using a technique adapted from phase-based video magnification, in which we measure motion from periodic changes in local phase information albeit for labeling rather than magnification. Based on measuring local phase through spatial decomposition of each frame of the endoscopic video using complex wavelet pairs, our approach assigns segmentation labels by detecting regions exhibiting temporal local phase changes matching the heart rate. We demonstrate how our technique is a practical solution for time-critical surgical applications by presenting quantitative and qualitative performance evaluations of our vessel detection algorithms with a retrospective study of fifteen clinical robot-assisted partial nephrectomies.

Auto Localization and Segmentation of Occluded Vessels in Robot-Assisted Partial Nephrectomy

Hilar dissection is an important and delicate stage in partial nephrectomy during which surgeons remove connective tissue surrounding renal vasculature. Potentially serious complications arise when vessels occluded by fat are missed in the endoscopic view and are not appropriately clamped. To aid in vessel discovery, we propose an automatic method to localize and label occluded vasculature. Our segmentation technique is adapted from phase-based video magnification, in which we measure subtle motion from periodic changes in local phase information albeit for labeling rather than magnification. We measure local phase through spatial decomposition of each frame of the endoscopic video using complex wavelet pairs. We then assign segmentation labels based on identifying responses of regions exhibiting temporal local phase changes matching the heart rate frequency. Our method is evaluated with a retrospective study of eight real robot-assisted partial nephrectomies demonstrating utility for surgical guidance that could potentially reduce operation times and complication rates.

Uncertainty-Encoded Augmented Reality for Robot-Assisted Partial Nephrectomy: A Phantom Study

In most robot-assisted surgical interventions, multimodal fusion of pre- and intra-operative data is highly valuable, affording the surgeon a more comprehensive understanding of the surgical scene observed through the stereo endoscopic camera. More specifically, in the case of partial nephrectomy, fusing pre-operative segmentations of kidney and tumor with the stereo endoscopic view can guide tumor localization and the identification of resection margins. However, the surgeons are often unable to reliably assess the levels of trust they can bestow on what is overlaid on the screen. In this paper, we present the proof-of-concept of an uncertainty-encoded augmented reality framework and novel visualizations of the uncertainties derived from the pre-operative CT segmentation onto the surgeon’s stereo endoscopic view. To verify its clinical potential, the proposed method is applied to an ex vivo lamb kidney. The results are contrasted to different visualization solutions based on crisp segmentation demonstrating that our method provides valuable additional information that can help the surgeon during the resection planning.

Towards multi-modal image-guided tumour identification in robot-assisted partial nephrectomy

Tumour identification is a critical step in robot-assisted partial nephrectomy (RAPN) during which the surgeon determines the tumour localization and resection margins. To help the surgeon in achieving this step, our research work aims at leveraging both pre- and intra-operative imaging modalities (CT, MRI, laparoscopic US, stereo endoscopic video) to provide an augmented reality view of kidney-tumour boundaries with uncertainty-encoded information. We present herein the progress of this research work including segmentation of preoperative scans, biomechanical simulation of deformations, stereo surface reconstruction from stereo endoscopic camera, pre-operative to intra-operative data registration, and augmented reality visualization.

Real-time extraction of local phase features from volumetric medical image data

We present a novel real-time implementation of local phase feature extraction from volumetric image data based on 3D directional (log-Gabor) filters. We achieve drastic performance gains without compromising the signal-to-noise ratio by pre-computing the filters and adaptive noise estimation parameters, and streamlining the remainder of the computations to efficiently run on a multi-processor graphic processing unit (GPU). We validate our method on clinical ultrasound data and demonstrate a 15-fold speedup in computation time over state-of-the art methods, which could potentially facilitate a wide range of practical applications for real-time image-guided procedures.

Automatic Vessel Segmentation from Pulsatile Radial Distension

Identification of vascular structures from medical images is integral to many clinical procedures. Most vessel segmentation techniques ignore the characteristic pulsatile motion of vessels in that formulation. In a recent effort to automatically segment vessels that are hidden under fat, we motivated the use of the magnitude of local pulsatile motion extracted from surgical endoscopic video. In this paper we propose a new approach that leverages the local orientation, in addition to magnitude of motion, and demonstrate that the extended computation of motion vectors can improve the segmentation of vascular structures. We implement our approach using two alternatives to magnitude-only motion estimation by using traditional optical flow and by exploiting the monogenic signal for fast flow estimation. Our evaluations are conducted on both synthetic phantoms as well as real ultrasound data showing improved segmentation results (0.36 increase in DSC and 0.11 increase in AUC) with negligible change in computational performance.

3D Surface Reconstruction of Organs Using Patient-Specific Shape Priors in Robot-Assisted Laparoscopic Surgery

With the advent of robot-assisted laparoscopic surgery RALS, intra-operative stereo endoscopy is becoming a ubiquitous imaging modality in abdominal interventions. This high resolution intra-operative imaging modality enables the reconstruction of 3D soft-tissue surface geometry with the help of computer vision techniques. This reconstructed surface is a prerequisite for many clinical applications such as image-guidance with cross-modality registration, telestration, expansion of the surgical scene by stitching/mosaicing, and collision detection. Reconstructing the surface geometry from camera information alone remains a very challenging problem in RALS mainly due to a small baseline between the optical centres of the cameras, presence of blood and smoke, specular highlights, occlusion, and smooth/textureless regions. In this paper, we propose a method for increasing the overall surface reconstruction accuracy by incorporating patient specific shape priors extracted from pre-operative images. Our method is validated on an ini¾?silico phantom and we show that the combination of both pre-operative and intra-operative data significantly improves surface reconstruction as compared to the ground truth. Finally, we verify the clinical potential of the proposed method in the context of abdominal surgery in a phantom study of an exi¾?vivo lamb kidney.

Propagation of registration uncertainty during multi-fraction cervical cancer brachytherapy

Multi-fraction cervical cancer brachytherapy is a form of image-guided radiotherapy that heavily relies on 3D imaging during treatment planning, delivery, and quality control. In this context, deformable image registration can increase the accuracy of dosimetric evaluations, provided that one can account for the uncertainties associated with the registration process. To enable such capability, we propose a mathematical framework that first estimates the registration uncertainty and subsequently propagates the effects of the computed uncertainties from the registration stage through to the visualizations, organ segmentations, and dosimetric evaluations. To ensure the practicality of our proposed framework in real world image-guided radiotherapy contexts, we implemented our technique via a computationally efficient and generalizable algorithm that is compatible with existing deformable image registration software. In our clinical context of fractionated cervical cancer brachytherapy, we perform a retrospective analysis on 37 patients and present evidence that our proposed methodology for computing and propagating registration uncertainties may be beneficial during therapy planning and quality control. Specifically, we quantify and visualize the influence of registration uncertainty on dosimetric analysis during the computation of the total accumulated radiation dose on the bladder wall. We further show how registration uncertainty may be leveraged into enhanced visualizations that depict the quality of the registration and highlight potential deviations from the treatment plan prior to the delivery of radiation treatment. Finally, we show that we can improve the transfer of delineated volumetric organ segmentation labels from one fraction to the next by encoding the computed registration uncertainties into the segmentation labels.

Recovering Missing Connections in Diffusion Weighted MRI Using Matrix Completion

Diffusion weighted magnetic resonance imaging (dwMRI) has become the dominant neuroimaging modality for estimating anatomical connectivity (AC). However, such AC estimation is prone to error due to missing connections resulting from crossing fibers and fiber endpoint uncertainty because of insufficient spatial resolution. Endeavors tackling this problem include improving fiber orientation estimation , applying heuristics to extrapolate fiber endpoints, and increasing spatial resolution. Refining fiber orientation estimation and tractography algorithms can only improve AC estimation to a certain extent, since the attainable improvement is constrained by the current limit on spatial resolution. We thus instead propose using matrix completion (MC) to recover missing connections. The underlying assumption is that the missing connections are intrinsically related to the observed entries of the AC matrix. A critical parameter that governs MC performance is the matrix rank. For this, we present a robust strategy that bypasses selection of a specific rank. Further, standard MC algorithms do not constrain the recovered entries to be non-negative, but this condition is necessary for fiber counts. We thus devise a method to interpolate negative entries based on neighborhood information. On synthetic data, our approach is able to accurately recover deleted AC matrix entries. On real data, AC estimated with our approach achieves higher IQ prediction accuracy than the original AC estimates, fiber endpoint extrapolation, and median filtering.

Segmentation-Free Kidney Localization and Volume Estimation Using Aggregated Orthogonal Decision CNNs

Kidney volume is an important bio-marker in the clinical diagnosis of various renal diseases. For example, it plays an essential role in follow-up evaluation of kidney transplants. Most existing methods for volume estimation rely on kidney segmentation as a prerequisite step, which has various limitations such as initialization-sensitivity and computationally-expensive optimization. In this paper, we propose a hybrid localization-volume estimation deep learning approach capable of (i) localizing kidneys in abdominal CT images, and (ii) estimating renal volume without requiring segmentation. Our approach involves multiple levels of self-learning of image representation using convolutional neural layers, which we show better capture the rich and complex variability in kidney data, demonstrably outperforming hand-crafted feature representations. We validate our method on clinical data of 100 patients with a total of 200 kidney samples (left and right). Our results demonstrate a 55% increase in kidney boundary localization accuracy, and a 30% increase in volume estimation accuracy compared to recent state-of-the-art methods deploying regression-forest-based learning for the same tasks.