Roger Schaer

Using MapReduce for Large-Scale Medical Image Analysis

In this paper, MapReduce is used to speed up and make possible three large-scale medical image processing use-cases: (i) parameter optimization for lung texture classification using support vector machines (SVM), (ii) content-based medical image indexing, and (iii) three-dimensional directional wavelet analysis for solid texture classification.

Cloud-Based Evaluation of Anatomical Structure Segmentation and Landmark Detection Algorithms: VISCERAL Anatomy Benchmarks

Variations in the shape and appearance of anatomical structures in medical images are often relevant radiological signs of disease. Automatic tools can help automate parts of this manual process. A cloud-based evaluation framework is presented in this paper including results of benchmarking current state-of-the-art medical imaging algorithms for anatomical structure segmentation and landmark detection: the VISCERAL Anatomy benchmarks. The algorithms are implemented in virtual machines in the cloud where participants can only access the training data and can be run privately by the benchmark administrators to objectively compare their performance in an unseen common test set. Overall, 120 computed tomography and magnetic resonance patient volumes were manually annotated to create a standard Gold Corpus containing a total of 1295 structures and 1760 landmarks. Ten participants contributed with automatic algorithms for the organ segmentation task, and three for the landmark localization task. Different algorithms obtained the best scores in the four available imaging modalities and for subsets of anatomical structures. The annotation framework, resulting data set, evaluation setup, results and performance analysis from the three VISCERAL Anatomy benchmarks are presented in this article. Both the VISCERAL data set and Silver Corpus generated with the fusion of the participant algorithms on a larger set of non-manually-annotated medical images are available to the research community.

Gesture Interaction for Content--based Medical Image Retrieval

Large amounts of medical images are being produced to help physicians in diagnosis and treatment planning. These images are then archived in PACS (Picture Archival and Communication Systems) and usually they are only reused in the context of the same patient during further visits. Medical image retrieval systems allow medical professionals to search for images in institutional archives, the Internet or in the scientific literature. The goal of the search can be in diagnosis but often as well for teaching and research. A large body of research has investigated efficient and effective algorithms to retrieve a set of images to fulfil a specific information need. However, much less research has been done on studying simple and engaging interaction for users of medical image retrieval systems. In this paper we propose an intuitive and engaging web--based interface targeted to be used by a large range of users with gesture control. This interface allows users to retrieve medical images by accessing a system called Parallel Distributed Image Search Engine (ParaDISE), a text-- and content--based image retrieval system. Accepting search with keywords and example images, this interface uses simple gestures to get random example images and mark examples as positive and negative relevance feedback with results being updated after each interaction.

Facilitating medical information search using Google Glass connected to a content-based medical image retrieval system

Wearable computing devices are starting to change the way users interact with computers and the Internet. Among them, Google Glass includes a small screen located in front of the right eye, a camera filming in front of the user and a small computing unit. Google Glass has the advantage to provide online services while allowing the user to perform tasks with his/her hands. These augmented glasses uncover many useful applications, also in the medical domain. For example, Google Glass can easily provide video conference between medical doctors to discuss a live case. Using these glasses can also facilitate medical information search by allowing the access of a large amount of annotated medical cases during a consultation in a non-disruptive fashion for medical staff. In this paper, we developed a Google Glass application able to take a photo and send it to a medical image retrieval system along with keywords in order to retrieve similar cases. As a preliminary assessment of the usability of the application, we tested the application under three conditions (images of the skin; printed CT scans and MRI images; and CT and MRI images acquired directly from an LCD screen) to explore whether using Google Glass affects the accuracy of the results returned by the medical image retrieval system. The preliminary results show that despite minor problems due to the relative stability of the Google Glass, images can be sent to and processed by the medical image retrieval system and similar images are returned to the user, potentially helping in the decision making process.

Comparing fusion techniques for the ImageCLEF 2013 medical case retrieval task

Retrieval systems can supply similar cases with a proven diagnosis to a new example case under observation to help clinicians during their work. The ImageCLEFmed evaluation campaign proposes a framework where research groups can compare case-based retrieval approaches. This paper focuses on the case-based task and adds results of the compound figure separation and modality classification tasks. Several fusion approaches are compared to identify the approaches best adapted to the heterogeneous data of the task. Fusion of visual and textual features is analyzed, demonstrating that the selection of the fusion strategy can improve the best performance on the case-based retrieval task.

Live ECG readings using Google Glass in emergency situations

Most sudden cardiac problems require rapid treatment to preserve life. In this regard, electrocardiograms (ECG) shown on vital parameter monitoring systems help medical staff to detect problems. In some situations, such monitoring systems may display information in a less than convenient way for medical staff. For example, vital parameters are displayed on large screens outside the field of view of a surgeon during cardiac surgery. This may lead to losing time and to mistakes when problems occur during cardiac operations. In this paper we present a novel approach to display vital parameters such as the second derivative of the ECG rhythm and heart rate close to the field of view of a surgeon using Google Glass. As a preliminary assessment, we run an experimental study to verify the possibility for medical staff to identify abnormal ECG rhythms from Google Glass. This study compares 6 ECG rhythms readings from a 13.3 inch laptop screen and from the prism of Google Glass. Seven medical residents in internal medicine participated in the study. The preliminary results show that there is no difference between identifying these 6 ECG rhythms from the laptop screen versus Google Glass. Both allow close to perfect identification of the 6 common ECG rhythms. This shows the potential of connected glasses such as Google Glass to be useful in selected medical applications.

QuantImage: An Online Tool for High-Throughput 3D Radiomics Feature Extraction in PET-CT

Abstract The processes of radiomics consist of image-based personalized tumor phenotyping for precision medicine. They complement slow, costly, and invasive molecular analysis of tumoral tissue. Whereas the relevance of a large variety of quantitative imaging biomarkers has been demonstrated for various cancer types, most studies were based on 2D image analysis of relatively small patient cohorts. In this work, we propose an online tool for automatically extracting 3D state-of-the-art quantitative imaging features from large batches of patients. The developed platform is called QuantImage and can be accessed from any web browser. Its use is straightforward and can be further parameterized for refined analyses. It relies on a robust 3D processing pipeline allowing normalization across patients and imaging protocols. The user can simply drag-and-drop a large zip file containing all image data for a batch of patients and the platform returns a spreadsheet with the set of quantitative features extracted for each patient. It is expected to enable high-throughput reproducible research and the validation of radiomics imaging parameters to shape the future of noninvasive personalized medicine.

Semi---supervised Learning for Image ModalityźClassification

Searching for medical image content is a regular task for many physicians, especially in radiology. Retrieval of medical images from the scientific literature can benefit from automatic modality classification to focus the search and filter out non---relevant items. Training datasets are often unevenly distributed regarding the classes resulting sometimes in a less than optimal classification performance. This article proposes a semi---supervised learning approach applied using a k---Nearest Neighbours k---NN classifier to exploit unlabelled data and to expand the training set. The algorithmic implementation is described and the method is evaluated on the ImageCLEFmed modality classification benchmark. Results show that this approach achieves an improved performance over supervised k---NN and Random Forest classifiers. Moreover, medical case---based retrieval also obtains higher performance when using the classified modalities as filter. This shows that image types can be classified well using visual information and they can then be used in a variety of applciations.

Optimized Distributed Hyperparameter Search and Simulation for Lung Texture Classification in CT Using Hadoop

Many medical image analysis tasks require complex learning strategies to reach a quality of image-based decision support that is sufficient in clinical practice. The analysis of medical texture in tomographic images, for example of lung tissue, is no exception. Via a learning framework, very good classification accuracy can be obtained, but several parameters need to be optimized. This article describes a practical framework for efficient distributed parameter optimization. The proposed solutions are applicable for many research groups with heterogeneous computing infrastructures and for various machine learning algorithms. These infrastructures can easily be connected via distributed computation frameworks. We use the Hadoop framework to run and distribute both grid and random search strategies for hyperparameter optimization and cross-validations on a cluster of 21 nodes composed of desktop computers and servers. We show that significant speedups of up to 364× compared to a serial execution can be achieved using our in-house Hadoop cluster by distributing the computation and automatically pruning the search space while still identifying the best-performing parameter combinations. To the best of our knowledge, this is the first article presenting practical results in detail for complex data analysis tasks on such a heterogeneous infrastructure together with a linked simulation framework that allows for computing resource planning. The results are directly applicable in many scenarios and allow implementing an efficient and effective strategy for medical (image) data analysis and related learning approaches.

Using Crowdsourcing for Multi-label Biomedical Compound Figure Annotation

Information analysis or retrieval for images in the biomedical literature needs to deal with a large amount of compound figures (figures containing several subfigures), as they constitute probably more than half of all images in repositories such as PubMed Central, which was the data set used for the task. The ImageCLEFmed benchmark proposed among other tasks in 2015 and 2016 a multi-label classification task, which aims at evaluating the automatic classification of figures into 30 image types. This task was based on compound figures and thus the figures were distributed to participants as compound figures but also in a separated form. Therefore, the generation of a gold standard was required, so that algorithms of participants can be evaluated and compared. This work presents the process carried out to generate the multi-labels of \(\sim \,2650\) compound figures using a crowdsourcing approach. Automatic algorithms to separate compound figures into subfigures were used and the results were then validated or corrected via crowdsourcing. The image types (MR, CT, X–ray, ...) were also annotated by crowdsourcing including detailed quality control. Quality control is necessary to insure quality of the annotated data as much as possible. \(\sim \,625\) h were invested with a cost of \(\sim \,870\