arXiv subjects

Ulas Bagci

Publications and source records attributed to Ulas Bagci.

At least 199 records · Page 11Linked to original sources

arXiv2018

Deep Geodesic Learning for Segmentation and Anatomical Landmarking

In this paper, we propose a novel deep learning framework for anatomy segmentation and automatic landmark- ing. Specifically, we focus on the challenging problem of mandible segmentation from cone-beam computed tomography (CBCT) scans and identification of 9 anatomical landmarks of the mandible on the geodesic space. The overall approach employs three inter-related steps. In step 1, we propose a deep neu- ral network architecture with carefully designed regularization, and network hyper-parameters to perform image segmentation without the need for data augmentation and complex post- processing refinement. In step 2, we formulate the landmark localization problem directly on the geodesic space for sparsely- spaced anatomical landmarks. In step 3, we propose to use a long short-term memory (LSTM) network to identify closely- spaced landmarks, which is rather difficult to obtain using other standard detection networks. The proposed fully automated method showed superior efficacy compared to the state-of-the- art mandible segmentation and landmarking approaches in craniofacial anomalies and diseased states. We used a very challenging CBCT dataset of 50 patients with a high-degree of craniomaxillofacial (CMF) variability that is realistic in clinical practice. Complementary to the quantitative analysis, the qualitative visual inspection was conducted for distinct CBCT scans from 250 patients with high anatomical variability. We have also shown feasibility of the proposed work in an independent dataset from MICCAI Head-Neck Challenge (2015) achieving the state-of-the-art performance. Lastly, we present an in-depth analysis of the proposed deep networks with respect to the choice of hyper-parameters such as pooling and activation functions.

cs.CV↗

arXiv2018

Automatically Designing CNN Architectures for Medical Image Segmentation

Deep neural network architectures have traditionally been designed and explored with human expertise in a long-lasting trial-and-error process. This process requires huge amount of time, expertise, and resources. To address this tedious problem, we propose a novel algorithm to optimally find hyperparameters of a deep network architecture automatically. We specifically focus on designing neural architectures for medical image segmentation task. Our proposed method is based on a policy gradient reinforcement learning for which the reward function is assigned a segmentation evaluation utility (i.e., dice index). We show the efficacy of the proposed method with its low computational cost in comparison with the state-of-the-art medical image segmentation networks. We also present a new architecture design, a densely connected encoder-decoder CNN, as a strong baseline architecture to apply the proposed hyperparameter search algorithm. We apply the proposed algorithm to each layer of the baseline architectures. As an application, we train the proposed system on cine cardiac MR images from Automated Cardiac Diagnosis Challenge (ACDC) MICCAI 2017. Starting from a baseline segmentation architecture, the resulting network architecture obtains the state-of-the-art results in accuracy without performing any trial-and-error based architecture design approaches or close supervision of the hyperparameters changes.

stat.ML↗

arXiv2018

End to End Brain Fiber Orientation Estimation using Deep Learning

In this work, we explore the various Brain Neuron tracking techniques, which is one of the most significant applications of Diffusion Tensor Imaging. Tractography provides us with a non-invasive method to analyze underlying tissue micro-structure. Understanding the structure and organization of the tissues facilitates us with a diagnosis method to identify any aberrations and provide acute information on the occurrences of brain ischemia or stroke, the mutation of neurological diseases such as Alzheimer, multiple sclerosis and so on. Time if of essence and accurate localization of the aberrations can help save or change a diseased life. Following up with the limitations introduced by the current Tractography techniques such as computational complexity, reconstruction errors during tensor estimation and standardization, we aim to elucidate these limitations through our research findings. We introduce an end to end Deep Learning framework which can accurately estimate the most probable likelihood orientation at each voxel along a neuronal pathway. We use Probabilistic Tractography as our baseline model to obtain the training data and which also serve as a Tractography Gold Standard for our evaluations. Through experiments we show that our Deep Network can do a significant improvement over current Tractography implementations by reducing the run-time complexity to a significant new level. Our architecture also allows for variable sized input DWI signals eliminating the need to worry about memory issues as seen with the traditional techniques. The advantage of this architecture is that it is perfectly desirable to be processed on a cloud setup and utilize the existing multi GPU frameworks to perform whole brain Tractography in minutes rather than hours. We evaluate our network with Gold Standard and benchmark its performance across several parameters.

cs.CV↗

arXiv2018

S4ND: Single-Shot Single-Scale Lung Nodule Detection

The state of the art lung nodule detection studies rely on computationally expensive multi-stage frameworks to detect nodules from CT scans. To address this computational challenge and provide better performance, in this paper we propose S4ND, a new deep learning based method for lung nodule detection. Our approach uses a single feed forward pass of a single network for detection and provides better performance when compared to the current literature. The whole detection pipeline is designed as a single $3D$ Convolutional Neural Network (CNN) with dense connections, trained in an end-to-end manner. S4ND does not require any further post-processing or user guidance to refine detection results. Experimentally, we compared our network with the current state-of-the-art object detection network (SSD) in computer vision as well as the state-of-the-art published method for lung nodule detection (3D DCNN). We used publically available $888$ CT scans from LUNA challenge dataset and showed that the proposed method outperforms the current literature both in terms of efficiency and accuracy by achieving an average FROC-score of $0.897$. We also provide an in-depth analysis of our proposed network to shed light on the unclear paradigms of tiny object detection.

cs.CV↗

arXiv2018

Semi-supervised multi-task learning for lung cancer diagnosis

Early detection of lung nodules is of great importance in lung cancer screening. Existing research recognizes the critical role played by CAD systems in early detection and diagnosis of lung nodules. However, many CAD systems, which are used as cancer detection tools, produce a lot of false positives (FP) and require a further FP reduction step. Furthermore, guidelines for early diagnosis and treatment of lung cancer are consist of different shape and volume measurements of abnormalities. Segmentation is at the heart of our understanding of nodules morphology making it a major area of interest within the field of computer aided diagnosis systems. This study set out to test the hypothesis that joint learning of false positive (FP) nodule reduction and nodule segmentation can improve the computer aided diagnosis (CAD) systems' performance on both tasks. To support this hypothesis we propose a 3D deep multi-task CNN to tackle these two problems jointly. We tested our system on LUNA16 dataset and achieved an average dice similarity coefficient (DSC) of 91% as segmentation accuracy and a score of nearly 92% for FP reduction. As a proof of our hypothesis, we showed improvements of segmentation and FP reduction tasks over two baselines. Our results support that joint training of these two tasks through a multi-task learning approach improves system performance on both. We also showed that a semi-supervised approach can be used to overcome the limitation of lack of labeled data for the 3D segmentation task.

cs.CV↗

arXiv2018

A Collaborative Computer Aided Diagnosis (C-CAD) System with Eye-Tracking, Sparse Attentional Model, and Deep Learning

There are at least two categories of errors in radiology screening that can lead to suboptimal diagnostic decisions and interventions:(i)human fallibility and (ii)complexity of visual search. Computer aided diagnostic (CAD) tools are developed to help radiologists to compensate for some of these errors. However, despite their significant improvements over conventional screening strategies, most CAD systems do not go beyond their use as second opinion tools due to producing a high number of false positives, which human interpreters need to correct. In parallel with efforts in computerized analysis of radiology scans, several researchers have examined behaviors of radiologists while screening medical images to better understand how and why they miss tumors, how they interact with the information in an image, and how they search for unknown pathology in the images. Eye-tracking tools have been instrumental in exploring answers to these fundamental questions. In this paper, we aim to develop a paradigm shift CAD system, called collaborative CAD (C-CAD), that unifies both of the above mentioned research lines: CAD and eye-tracking. We design an eye-tracking interface providing radiologists with a real radiology reading room experience. Then, we propose a novel algorithm that unifies eye-tracking data and a CAD system. Specifically, we present a new graph based clustering and sparsification algorithm to transform eye-tracking data (gaze) into a signal model to interpret gaze patterns quantitatively and qualitatively. The proposed C-CAD collaborates with radiologists via eye-tracking technology and helps them to improve diagnostic decisions. The C-CAD learns radiologists' search efficiency by processing their gaze patterns. To do this, the C-CAD uses a deep learning algorithm in a newly designed multi-task learning platform to segment and diagnose cancers simultaneously.

cs.CV↗

arXiv2018

Deep Multi-Modal Classification of Intraductal Papillary Mucinous Neoplasms (IPMN) with Canonical Correlation Analysis

Pancreatic cancer has the poorest prognosis among all cancer types. Intraductal Papillary Mucinous Neoplasms (IPMNs) are radiographically identifiable precursors to pancreatic cancer; hence, early detection and precise risk assessment of IPMN are vital. In this work, we propose a Convolutional Neural Network (CNN) based computer aided diagnosis (CAD) system to perform IPMN diagnosis and risk assessment by utilizing multi-modal MRI. In our proposed approach, we use minimum and maximum intensity projections to ease the annotation variations among different slices and type of MRIs. Then, we present a CNN to obtain deep feature representation corresponding to each MRI modality (T1-weighted and T2-weighted). At the final step, we employ canonical correlation analysis (CCA) to perform a fusion operation at the feature level, leading to discriminative canonical correlation features. Extracted features are used for classification. Our results indicate significant improvements over other potential approaches to solve this important problem. The proposed approach doesn't require explicit sample balancing in cases of imbalance between positive and negative examples. To the best of our knowledge, our study is the first to automatically diagnose IPMN using multi-modal MRI.

cs.CV↗

arXiv2018

Capsules for Object Segmentation

Convolutional neural networks (CNNs) have shown remarkable results over the last several years for a wide range of computer vision tasks. A new architecture recently introduced by Sabour et al., referred to as a capsule networks with dynamic routing, has shown great initial results for digit recognition and small image classification. The success of capsule networks lies in their ability to preserve more information about the input by replacing max-pooling layers with convolutional strides and dynamic routing, allowing for preservation of part-whole relationships in the data. This preservation of the input is demonstrated by reconstructing the input from the output capsule vectors. Our work expands the use of capsule networks to the task of object segmentation for the first time in the literature. We extend the idea of convolutional capsules with locally-connected routing and propose the concept of deconvolutional capsules. Further, we extend the masked reconstruction to reconstruct the positive input class. The proposed convolutional-deconvolutional capsule network, called SegCaps, shows strong results for the task of object segmentation with substantial decrease in parameter space. As an example application, we applied the proposed SegCaps to segment pathological lungs from low dose CT scans and compared its accuracy and efficiency with other U-Net-based architectures. SegCaps is able to handle large image sizes (512 x 512) as opposed to baseline capsules (typically less than 32 x 32). The proposed SegCaps reduced the number of parameters of U-Net architecture by 95.4% while still providing a better segmentation accuracy.

stat.ML↗

arXiv2018

How to Fool Radiologists with Generative Adversarial Networks? A Visual Turing Test for Lung Cancer Diagnosis

Discriminating lung nodules as malignant or benign is still an underlying challenge. To address this challenge, radiologists need computer aided diagnosis (CAD) systems which can assist in learning discriminative imaging features corresponding to malignant and benign nodules. However, learning highly discriminative imaging features is an open problem. In this paper, our aim is to learn the most discriminative features pertaining to lung nodules by using an adversarial learning methodology. Specifically, we propose to use unsupervised learning with Deep Convolutional-Generative Adversarial Networks (DC-GANs) to generate lung nodule samples realistically. We hypothesize that imaging features of lung nodules will be discriminative if it is hard to differentiate them (fake) from real (true) nodules. To test this hypothesis, we present Visual Turing tests to two radiologists in order to evaluate the quality of the generated (fake) nodules. Extensive comparisons are performed in discerning real, generated, benign, and malignant nodules. This experimental set up allows us to validate the overall quality of the generated nodules, which can then be used to (1) improve diagnostic decisions by mining highly discriminative imaging features, (2) train radiologists for educational purposes, and (3) generate realistic samples to train deep networks with big data.

cs.CV↗

arXiv2017

Simultaneous Detection and Quantification of Retinal Fluid with Deep Learning

We propose a new deep learning approach for automatic detection and segmentation of fluid within retinal OCT images. The proposed framework utilizes both ResNet and Encoder-Decoder neural network architectures. When training the network, we apply a novel data augmentation method called myopic warping together with standard rotation-based augmentation to increase the training set size to 45 times the original amount. Finally, the network output is post-processed with an energy minimization algorithm (graph cut) along with a few other knowledge guided morphological operations to finalize the segmentation process. Based on OCT imaging data and its ground truth from the RETOUCH challenge, the proposed system achieves dice indices of 0.522, 0.682, and 0.612, and average absolute volume differences of 0.285, 0.115, and 0.156 mm$^3$ for intaretinal fluid, subretinal fluid, and pigment epithelial detachment respectively.

cs.CV↗

arXiv2017

Automatic Response Assessment in Regions of Language Cortex in Epilepsy Patients Using ECoG-based Functional Mapping and Machine Learning

Accurate localization of brain regions responsible for language and cognitive functions in Epilepsy patients should be carefully determined prior to surgery. Electrocorticography (ECoG)-based Real Time Functional Mapping (RTFM) has been shown to be a safer alternative to the electrical cortical stimulation mapping (ESM), which is currently the clinical/gold standard. Conventional methods for analyzing RTFM signals are based on statistical comparison of signal power at certain frequency bands. Compared to gold standard (ESM), they have limited accuracies when assessing channel responses. In this study, we address the accuracy limitation of the current RTFM signal estimation methods by analyzing the full frequency spectrum of the signal and replacing signal power estimation methods with machine learning algorithms, specifically random forest (RF), as a proof of concept. We train RF with power spectral density of the time-series RTFM signal in supervised learning framework where ground truth labels are obtained from the ESM. Results obtained from RTFM of six adult patients in a strictly controlled experimental setup reveal the state of the art detection accuracy of $\approx 78\%$ for the language comprehension task, an improvement of $23\%$ over the conventional RTFM estimation method. To the best of our knowledge, this is the first study exploring the use of machine learning approaches for determining RTFM signal characteristics, and using the whole-frequency band for better region localization. Our results demonstrate the feasibility of machine learning based RTFM signal analysis method over the full spectrum to be a clinical routine in the near future.

q-bio.NC↗

arXiv2017

Multi-Planar Deep Segmentation Networks for Cardiac Substructures from MRI and CT

Non-invasive detection of cardiovascular disorders from radiology scans requires quantitative image analysis of the heart and its substructures. There are well-established measurements that radiologists use for diseases assessment such as ejection fraction, volume of four chambers, and myocardium mass. These measurements are derived as outcomes of precise segmentation of the heart and its substructures. The aim of this paper is to provide such measurements through an accurate image segmentation algorithm that automatically delineates seven substructures of the heart from MRI and/or CT scans. Our proposed method is based on multi-planar deep convolutional neural networks (CNN) with an adaptive fusion strategy where we automatically utilize complementary information from different planes of the 3D scans for improved delineations. For CT and MRI, we have separately designed three CNNs (the same architectural configuration) for three planes, and have trained the networks from scratch for voxel-wise labeling for the following cardiac structures: myocardium of left ventricle (Myo), left atrium (LA), left ventricle (LV), right atrium (RA), right ventricle (RV), ascending aorta (Ao), and main pulmonary artery (PA). We have evaluated the proposed method with 4-fold-cross validation on the multi-modality whole heart segmentation challenge (MM-WHS 2017) dataset. The precision and dice index of 0.93 and 0.90, and 0.87 and 0.85 were achieved for CT and MR images, respectively. While a CT volume was segmented about 50 seconds, an MRI scan was segmented around 17 seconds with the GPUs/CUDA implementation.

stat.ML↗

arXiv2017

CardiacNET: Segmentation of Left Atrium and Proximal Pulmonary Veins from MRI Using Multi-View CNN

Anatomical and biophysical modeling of left atrium (LA) and proximal pulmonary veins (PPVs) is important for clinical management of several cardiac diseases. Magnetic resonance imaging (MRI) allows qualitative assessment of LA and PPVs through visualization. However, there is a strong need for an advanced image segmentation method to be applied to cardiac MRI for quantitative analysis of LA and PPVs. In this study, we address this unmet clinical need by exploring a new deep learning-based segmentation strategy for quantification of LA and PPVs with high accuracy and heightened efficiency. Our approach is based on a multi-view convolutional neural network (CNN) with an adaptive fusion strategy and a new loss function that allows fast and more accurate convergence of the backpropagation based optimization. After training our network from scratch by using more than 60K 2D MRI images (slices), we have evaluated our segmentation strategy to the STACOM 2013 cardiac segmentation challenge benchmark. Qualitative and quantitative evaluations, obtained from the segmentation challenge, indicate that the proposed method achieved the state-of-the-art sensitivity (90%), specificity (99%), precision (94%), and efficiency levels (10 seconds in GPU, and 7.5 minutes in CPU).

cs.CV↗

arXiv2017

Risk Stratification of Lung Nodules Using 3D CNN-Based Multi-task Learning

Risk stratification of lung nodules is a task of primary importance in lung cancer diagnosis. Any improvement in robust and accurate nodule characterization can assist in identifying cancer stage, prognosis, and improving treatment planning. In this study, we propose a 3D Convolutional Neural Network (CNN) based nodule characterization strategy. With a completely 3D approach, we utilize the volumetric information from a CT scan which would be otherwise lost in the conventional 2D CNN based approaches. In order to address the need for a large amount for training data for CNN, we resort to transfer learning to obtain highly discriminative features. Moreover, we also acquire the task dependent feature representation for six high-level nodule attributes and fuse this complementary information via a Multi-task learning (MTL) framework. Finally, we propose to incorporate potential disagreement among radiologists while scoring different nodule attributes in a graph regularized sparse multi-task learning. We evaluated our proposed approach on one of the largest publicly available lung nodule datasets comprising 1018 scans and obtained state-of-the-art results in regressing the malignancy scores.

cs.CV↗

arXiv2017

TumorNet: Lung Nodule Characterization Using Multi-View Convolutional Neural Network with Gaussian Process

Characterization of lung nodules as benign or malignant is one of the most important tasks in lung cancer diagnosis, staging and treatment planning. While the variation in the appearance of the nodules remains large, there is a need for a fast and robust computer aided system. In this work, we propose an end-to-end trainable multi-view deep Convolutional Neural Network (CNN) for nodule characterization. First, we use median intensity projection to obtain a 2D patch corresponding to each dimension. The three images are then concatenated to form a tensor, where the images serve as different channels of the input image. In order to increase the number of training samples, we perform data augmentation by scaling, rotating and adding noise to the input image. The trained network is used to extract features from the input image followed by a Gaussian Process (GP) regression to obtain the malignancy score. We also empirically establish the significance of different high level nodule attributes such as calcification, sphericity and others for malignancy determination. These attributes are found to be complementary to the deep multi-view CNN features and a significant improvement over other methods is obtained.

cs.CV↗

arXiv2017

Robust and fully automated segmentation of mandible from CT scans

Mandible bone segmentation from computed tomography (CT) scans is challenging due to mandible's structural irregularities, complex shape patterns, and lack of contrast in joints. Furthermore, connections of teeth to mandible and mandible to remaining parts of the skull make it extremely difficult to identify mandible boundary automatically. This study addresses these challenges by proposing a novel framework where we define the segmentation as two complementary tasks: recognition and delineation. For recognition, we use random forest regression to localize mandible in 3D. For delineation, we propose to use 3D gradient-based fuzzy connectedness (FC) image segmentation algorithm, operating on the recognized mandible sub-volume. Despite heavy CT artifacts and dental fillings, consisting half of the CT image data in our experiments, we have achieved highly accurate detection and delineation results. Specifically, detection accuracy more than 96% (measured by union of intersection (UoI)), the delineation accuracy of 91% (measured by dice similarity coefficient), and less than 1 mm in shape mismatch (Hausdorff Distance) were found.

cs.CV↗

arXiv2016

Characterization of Lung Nodule Malignancy using Hybrid Shape and Appearance Features

Computed tomography imaging is a standard modality for detecting and assessing lung cancer. In order to evaluate the malignancy of lung nodules, clinical practice often involves expert qualitative ratings on several criteria describing a nodule's appearance and shape. Translating these features for computer-aided diagnostics is challenging due to their subjective nature and the difficulties in gaining a complete description. In this paper, we propose a computerized approach to quantitatively evaluate both appearance distinctions and 3D surface variations. Nodule shape was modeled and parameterized using spherical harmonics, and appearance features were extracted using deep convolutional neural networks. Both sets of features were combined to estimate the nodule malignancy using a random forest classifier. The proposed algorithm was tested on the publicly available Lung Image Database Consortium dataset, achieving high accuracy. By providing lung nodule characterization, this method can provide a robust alternative reference opinion for lung cancer diagnosis.

cs.CV↗

arXiv2016

Gaze2Segment: A Pilot Study for Integrating Eye-Tracking Technology into Medical Image Segmentation

This study introduced a novel system, called Gaze2Segment, integrating biological and computer vision techniques to support radiologists' reading experience with an automatic image segmentation task. During diagnostic assessment of lung CT scans, the radiologists' gaze information were used to create a visual attention map. This map was then combined with a computer-derived saliency map, extracted from the gray-scale CT images. The visual attention map was used as an input for indicating roughly the location of a object of interest. With computer-derived saliency information, on the other hand, we aimed at finding foreground and background cues for the object of interest. At the final step, these cues were used to initiate a seed-based delineation process. Segmentation accuracy of the proposed Gaze2Segment was found to be 86% with dice similarity coefficient and 1.45 mm with Hausdorff distance. To the best of our knowledge, Gaze2Segment is the first true integration of eye-tracking technology into a medical image segmentation task without the need for any further user-interaction.

cs.CV↗