Javier Royuela-Del-Val

The emergence of new technologies has changed the way clinicians perform diagnosis. Medical imaging play a crucial role in this process, given the amount of information that they usually provide as non-invasive techniques. Despite the high quality offered by these images and the expertise of clinicians, the diagnostic process is not a straightforward task since different pathologies can have similar signs and symptoms. For this reason, it is extremely useful to assist this process with the inclusion of an automatic tool that reduces the bias when analyzing this kind of images. In this work, we propose an ensemble classifier based on probabilistic Support Vector Machine (SVM) in order to identify relevant patterns while providing information about the reliability of the classification. Specifically, each image is divided into patches and features contained in each one of them are extracted by applying kernel principal component analysis (PCA). The use of base classifiers within an ensemble allows our system to identify the informative patterns regardless of their size or location. Decisions of each individual patch are then combined according to the reliability of each individual classification: the lower the uncertainty, the higher the contribution. Performance is evaluated in a real scenario where distinguishing between pneumonia patients and controls from chest Computed Tomography (CCT) images, yielding an accuracy of 97.86%. The large performance obtained and the simplicity of the system (use of deep learning in CCT images would highly increase the computational cost) evidence the applicability of our proposal in a real-world environment.

Medical image processing is often limited by the computational cost of the involved algorithms. Whereas dedicated computing devices (GPUs in particular) exist and do provide significant efficiency boosts, they have an extra cost of use in terms of housekeeping tasks (device selection and initialization, data streaming, synchronization with the CPU and others), which may hinder developers from using them. This paper describes an OpenCL-based framework that is capable of handling dedicated computing devices seamlessly and that allows the developer to concentrate on image processing tasks. The framework handles automatically device discovery and initialization, data transfers to and from the device and the file system and kernel loading and compiling. Data structures need to be defined only once independently of the computing device; code is unique, consequently, for every device, including the host CPU. Pinned memory/buffer mapping is used to achieve maximum performance in data transfers. Code fragments included in the paper show how the computing device is almost immediately and effortlessly available to the users algorithms, so they can focus on productive work. Code required for device selection and initialization, data loading and streaming and kernel compilation is minimal and systematic. Algorithms can be thought of as mathematical operators (called processes), with input, output and parameters, and they may be chained one after another easily and efficiently. Also for efficiency, processes can have their initialization work split from their core workload, so process chains and loops do not incur in performance penalties. Algorithm code is independent of the device type targeted.