Radiomics of NSCLC: Quantitative CT Image Feature Characterization and Tumor Shrinkage Prediction

Radiomics is the high-throughput extraction and analysis of quantitative image features. For non-small cell lung cancer (NSCLC) patients, radiomics can be applied to standard of care computed tomography (CT) images to improve tumor diagnosis, staging, and response assessment. The first objective of this work was to show that CT image features extracted from pre-treatment NSCLC tumors could be used to predict tumor shrinkage in response to therapy. This is important since tumor shrinkage is an important cancer treatment endpoint that is correlated with probability of disease progression and overall survival. Accurate prediction of tumor shrinkage could also lead to individually customized treatment plans. To accomplish this objective, 64 stage NSCLC patients with similar treatments were all imaged using the same CT scanner and protocol. Quantitative image features were extracted and principal component regression with simulated annealing subset selection was used to predict shrinkage. Cross validation and permutation tests were used to validate the results. The optimal model gave a strong correlation between the observed and predicted shrinkages with . The second objective of this work was to identify sets of NSCLC CT image features that are reproducible, non-redundant, and informative across multiple machines. Feature sets with these qualities are needed for NSCLC radiomics models to be robust to machine variation and spurious correlation. To accomplish this objective, test-retest CT image pairs were obtained from 56 NSCLC patients imaged on three CT machines from two institutions. For each machine, quantitative image features with concordance correlation coefficient values greater than 0.90 were considered reproducible. Multi-machine reproducible feature sets were created by taking the intersection of individual machine reproducible feature sets. Redundant features were removed through hierarchical clustering. The findings showed that image feature reproducibility and redundancy depended on both the CT machine and the CT image type (average cine 4D-CT imaging vs. end-exhale cine 4D-CT imaging vs. helical inspiratory breath-hold 3D CT). For each image type, a set of cross-machine reproducible, non-redundant, and informative image features was identified. Compared to end-exhale 4D-CT and breath-hold 3D-CT, average 4D-CT derived image features showed superior multi-machine reproducibility and are the best candidates for clinical correlation.

Video-based, real-time multi-view stereo

We investigate the problem of obtaining a dense reconstruction in real-time, from a live video stream. In recent years, multi-view stereo (MVS) has received considerable attention and a number of methods have been proposed. However, most methods operate under the assumption of a relatively sparse set of still images as input and unlimited computation time. Video based MVS has received less attention despite the fact that video sequences offer significant benefits in terms of usability of MVS systems. In this paper we propose a novel video based MVS algorithm that is suitable for real-time, interactive 3d modeling with a hand-held camera. The key idea is a per-pixel, probabilistic depth estimation scheme that updates posterior depth distributions with every new frame. The current implementation is capable of updating 15 million distributions/s. We evaluate the proposed method against the state-of-the-art real-time MVS method and show improvement in terms of accuracy. © 2011 Elsevier B.V. All rights reserved.

Leveraging Multiple Features for Image Retrieval and Matching

The goal of image retrieval and matching is to find and locate object instances in images from a large-scale image database. While visual features are abundant, how to combine them to improve performance by individual features remains a challenging task. In this work, we focus on leveraging multiple features for accurate and efficient image retrieval and matching. We first propose two graph-based approaches to rerank initially retrieved images for generic image retrieval. In the graph, vertices are images while edges are similarities between image pairs. Our first approach employs a mixture Markov model based on a random walk model on multiple graphs to fuse graphs. We introduce a probabilistic model to compute the importance of each feature for graph fusion under a naive Bayesian formulation, which requires statistics of similarities from a manually labeled dataset containing irrelevant images. To reduce human labeling, we further propose a fully unsupervised reranking algorithm based on a submodular objective function that can be efficiently optimized by greedy algorithm. By maximizing an information gain term over the graph, our submodular function favors a subset of database images that are similar to query images and resemble each other. The function also exploits the rank relationships of images from multiple ranked lists obtained by different features. We then study a more well-defined application, person re-identification, where the database contains labeled images of human bodies captured by multiple cameras. Re-identifications from multiple cameras are regarded as related tasks to exploit shared information. We apply a novel multi-task learning algorithm using both low level features and attributes. A low rank attribute embedding is joint learned within the multi-task learning formulation to embed original binary attributes to a continuous attribute space, where incorrect and incomplete attributes are rectified and recovered. To locate objects in images, we design an object detector based on object proposals and deep convolutional neural networks (CNN) in view of the emergence of deep networks. We improve a Fast RCNN framework and investigate two new strategies to detect objects accurately and efficiently: scale-dependent pooling (SDP) and cascaded rejection classifiers (CRC). The SDP improves detection accuracy by exploiting appropriate convolutional features depending on the scale of input object proposals. The CRC effectively utilizes convolutional features and greatly eliminates negative proposals in a cascaded manner, while maintaining a high recall for true objects. The two strategies together improve the detection accuracy and reduce the computational cost.

A User Interaction Bug Analyzer Based on Image

Context: Mobile applications support a set of user-interaction features that are independent of the application logic. Rotating the device, scrolling, or zooming are examples of such features. Some bugs in mobile applications can be attributed to user-interaction features. Objective: This paper proposes and evaluates a bug analyzer based on user-interaction features that uses digital image processing to find bugs. Method: Our bug analyzer detects bugs by comparing the similarity between images taken before and after a user-interaction. SURF, an interest point detector and descriptor, is used to compare the images. To evaluate the bug analyzer, we conducted a case study with 15 randomly selected mobile applications. First, we identified user-interaction bugs by manually testing the applications. Images were captured before and after applying each user-interaction feature. Then, image pairs were processed with SURF to obtain interest points, from which a similarity percentage was computed, to finally decide whether there was a bug. Results: We performed a total of 49 user-interaction feature tests. When manually testing the applications, 17 bugs were found, whereas when using image processing, 15 bugs were detected. Conclusions: 8 out of 15 mobile applications tested had bugs associated to user-interaction features. Our bug analyzer based on image processing was able to detect 88% (15 out of 17) of the user-interaction bugs found with manual testing.

Interobserver variability of radiation therapists aligning to fiducial markers for prostate radiation therapy

As the use of fiducial markers (FMs) for the localisation of the prostate during external beam radiation therapy (EBRT) has become part of routine practice, radiation therapists (RTs) have become increasingly responsible for online image interpretation. The aim of this investigation was to quantify the limits of agreement (LoA) between RTs when localising to FMs with orthogonal kilovoltage (kV) imaging. Methods Six patients receiving prostate EBRT utilising FMs were included in this study. Treatment localisation was performed using kV imaging prior to each fraction. Online stereoscopic assessment of FMs, performed by the treating RTs, was compared with the offline assessment by three RTs. Observer agreement was determined by pairwise Bland-Altman analysis. Results Stereoscopic analysis of 225 image pairs was performed online at the time of treatment, and offline by three RT observers. Eighteen pairwise Bland-Altman analyses were completed to assess the level of agreement between observers. Localisation by RTs was found to be within clinically acceptable 95% LoAs. Conclusions Small differences between RTs, in both the online and offline setting, were found to be within clinically acceptable limits. RTs were able to make consistent and reliable judgements when matching FMs on planar kV imaging.

Autonomous helicopter hover using an artificial neural network

Details the developments to date of an unmanned air vehicle (UAV) based on a standard size 60 model helicopter. The design goal is to have the helicopter achieve stable hover with the aid of an INS and stereo vision. The focus of the paper is on the development of an artificial neural network (ANN) that makes use of only the INS data to generate hover commands, which are used to directly manipulate the flight servos. Current results show that networks incorporating some form of recurrency (state history) offer little advantage over those without. At this stage, the ANN has partially maintained periods of hover even with misaligned sensors.

FeatureMatch: A General ANNF Estimation Technique and its Applications

In this paper, we propose FeatureMatch, a generalised approximate nearest-neighbour field (ANNF) computation framework, between a source and target image. The proposed algorithm can estimate ANNF maps between any image pairs, not necessarily related. This generalisation is achieved through appropriate spatial-range transforms. To compute ANNF maps, global colour adaptation is applied as a range transform on the source image. Image patches from the pair of images are approximated using low-dimensional features, which are used along with KD-tree to estimate the ANNF map. This ANNF map is further improved based on image coherency and spatial transforms. The proposed generalisation, enables us to handle a wider range of vision applications, which have not been tackled using the ANNF framework. We illustrate two such applications namely: 1) optic disk detection and 2) super resolution. The first application deals with medical imaging, where we locate optic disks in retinal images using a healthy optic disk image as common target image. The second application deals with super resolution of synthetic images using a common source image as dictionary. We make use of ANNF mappings in both these applications and show experimentally that our proposed approaches are faster and accurate, compared with the state-of-the-art techniques.

Efficient block segmentation algorithm for true motion estimation

A block-based motion estimation technique is proposed which permits a less general segmentation performed using an efficient deterministic algorithm. Applied to image pairs from the Flower Garden and Table Tennis sequences, the algorithm successfully localizes motion discontinuities and detects uncovered regions. The algorithm is implemented in C on a Sun Sparcstation 20. The gradient-based motion estimation required 28.8 s CPU time, and 500 iterations of the segmentation algorithm required 32.6 s.

Automated Computation of the Fundamental Matrix for Vision-Based Construction Site Applications

Estimating the fundamental matrix (F), to determine the epipolar geometry between a pair of images or video frames, is a basic step for a wide variety of vision-based functions used in construction operations, such as camera-pair calibration, automatic progress monitoring, and 3D reconstruction. Currently, robust methods (e.g., SIFT + normalized eight-point algorithm + RANSAC) are widely used in the construction community for this purpose. Although they can provide acceptable accuracy, the significant amount of required computational time impedes their adoption in real-time applications, especially video data analysis with many frames per second. Aiming to overcome this limitation, this paper presents and evaluates the accuracy of a solution to find F by combining the use of two speedy and consistent methods: SURF for the selection of a robust set of point correspondences and the normalized eight-point algorithm. This solution is tested extensively on construction site image pairs including changes in viewpoint, scale, illumination, rotation, and moving objects. The results demonstrate that this method can be used for real-time applications (5 image pairs per second with the resolution of 640 × 480) involving scenes of the built environment.

Groundtruth system for underwater benchmarking

Oceans - San Diego, 2013

Calibration Method for Underwater Visual Ground-Truth System

This work presents an automatic calibration method for a vision based external underwater ground-truth positioning system. These systems are a relevant tool in benchmarking and assessing the quality of research in underwater robotics applications. A stereo vision system can in suitable environments such as test tanks or in clear water conditions provide accurate position with low cost and flexible operation. In this work we present a two step extrinsic camera parameter calibration procedure in order to reduce the setup time and provide accurate results. The proposed method uses a planar homography decomposition in order to determine the relative camera poses and the determination of vanishing points of detected lines in the image to obtain the global pose of the stereo rig in the reference frame. This method was applied to our external vision based ground-truth at the INESC TEC/Robotics test tank. Results are presented in comparison with an precise calibration performed using points obtained from an accurate 3D LIDAR modelling of the environment.

Efficient topology estimation for large scale optical mapping

Large scale image mosaicing methods are in great demand among scientists who study different aspects of the seabed, and have been fostered by impressive advances in the capabilities of underwater robots in gathering optical data from the seafloor. Cost and weight constraints mean that lowcost Remotely operated vehicles (ROVs) usually have a very limited number of sensors. When a low-cost robot carries out a seafloor survey using a down-looking camera, it usually follows a predetermined trajectory that provides several non time-consecutive overlapping image pairs. Finding these pairs (a process known as topology estimation) is indispensable to obtaining globally consistent mosaics and accurate trajectory estimates, which are necessary for a global view of the surveyed area, especially when optical sensors are the only data source. This thesis presents a set of consistent methods aimed at creating large area image mosaics from optical data obtained during surveys with low-cost underwater vehicles. First, a global alignment method developed within a Feature-based image mosaicing (FIM) framework, where nonlinear minimisation is substituted by two linear steps, is discussed. Then, a simple four-point mosaic rectifying method is proposed to reduce distortions that might occur due to lens distortions, error accumulation and the difficulties of optical imaging in an underwater medium. The topology estimation problem is addressed by means of an augmented state and extended Kalman filter combined framework, aimed at minimising the total number of matching attempts and simultaneously obtaining the best possible trajectory. Potential image pairs are predicted by taking into account the uncertainty in the trajectory. The contribution of matching an image pair is investigated using information theory principles. Lastly, a different solution to the topology estimation problem is proposed in a bundle adjustment framework. Innovative aspects include the use of fast image similarity criterion combined with a Minimum spanning tree (MST) solution, to obtain a tentative topology. This topology is improved by attempting image matching with the pairs for which there is the most overlap evidence. Unlike previous approaches for large-area mosaicing, our framework is able to deal naturally with cases where time-consecutive images cannot be matched successfully, such as completely unordered sets. Finally, the efficiency of the proposed methods is discussed and a comparison made with other state-of-the-art approaches, using a series of challenging datasets in underwater scenarios

What's in a name? Lexical knowledge drives infants' visual preferences in the absence of referential input

Seventeen-month-old infants were presented with pairs of images, in silence or with the non-directive auditory stimulus 'look!'. The images had been chosen so that one image depicted an item whose name was known to the infant, and the other image depicted an image whose name was not known to the infant. Infants looked longer at images for which they had names than at images for which they did not have names, despite the absence of any referential input. The experiment controlled for the familiarity of the objects depicted: in each trial, image pairs presented to infants had previously been judged by caregivers to be of roughly equal familiarity. From a theoretical perspective, the results indicate that objects with names are of intrinsic interest to the infant. The possible causal direction for this linkage is discussed and it is concluded that the results are consistent with Whorfian linguistic determinism, although other construals are possible. From a methodological perspective, the results have implications for the use of preferential looking as an index of early word comprehension.

Extração semi-automática de rodovias em um par de imagens aéreas digitais utilizando geometria epipolar e programação dinâmica

The purpose of this paper is to introduce a methodology for semi-automatic road extraction from aerial digital image pairs by using dynamic programming and epipolar geometry. The method uses both images from where each road feature pair is extracted. The operator identifies the corresponding road featuresand s/he selects sparse seed points along them. After all road pairs have been extracted, epipolar geometry is applied to determine the automatic point-to-point correspondence between each correspondent feature. Finally, each correspondent road pair is georeferenced by photogrammetric intersection. Experiments were made with rural aerial images. The results led to the conclusion that the methodology is robust and efficient, even in the presence of shadows of trees and buildings or other irregularities.

Extração semi-automática da malha viária em imagens aéreas digitais de áreas rurais utilizando otimização por programação dinâmica no espaço objeto

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)