Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

A general non-parametric technique is proposed for the analysis of a complex multimodal feature space and to delineate arbitrarily shaped clusters in it. The basic computational module of the technique is an old pattern recognition procedure: the mean shift. For discrete data, we prove the convergence of a recursive mean shift procedure to the nearest stationary point of the underlying density function and, thus, its utility in detecting the modes of the density. The relation of the mean shift procedure to the Nadaraya-Watson estimator from kernel regression and the robust M-estimators; of location is also established. Algorithms for two low-level vision tasks discontinuity-preserving smoothing and image segmentation - are described as applications. In these algorithms, the only user-set parameter is the resolution of the analysis, and either gray-level or color images are accepted as input. Extensive experimental results illustrate their excellent performance.

Mean shift: a robust approach toward feature space analysis

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

A survey on deep learning in medical image analysis

Computer vision applications have come to rely increasingly on superpixels in recent years, but it is not always clear what constitutes a good superpixel algorithm. In an effort to understand the benefits and drawbacks of existing methods, we empirically compare five state-of-the-art superpixel algorithms for their ability to adhere to image boundaries, speed, memory efficiency, and their impact on segmentation performance. We then introduce a new superpixel algorithm, simple linear iterative clustering (SLIC), which adapts a k-means clustering approach to efficiently generate superpixels. Despite its simplicity, SLIC adheres to boundaries as well as or better than previous methods. At the same time, it is faster and more memory efficient, improves segmentation performance, and is straightforward to extend to supervoxel generation.

/pdf/slic-superpixels-compared-to-state-of-the-art-superpixel-5d8sq04am9.pdf

SLIC Superpixels Compared to State-of-the-Art Superpixel Methods

Personalized cardiac models have become a crucial component of the clinical workflow, especially in the context of complex cardiovascular disorders, such as valvular heart disease. In this chapter we present a comprehensive framework for the patient-specific modeling of the valvular apparatus and heart chambers from multi-modal cardiac images. An integrated model of the four heart valves and chambers is introduced, which captures a large spectrum of morphologic, dynamic and pathologic variations. The patient-specific model parameters are estimated from four-dimensional cardiac images using robust learning-based techniques. These include object localization, rigid and non-rigid motion estimation, and surface boundary estimation from dense 4D data (TEE, CT) as well as regression-based techniques for surface reconstruction from sparse 4D data (MRI). Clinical applications based on the patient-specific modeling approach are proposed for decision support in Transcatheter Aortic Valve Implantation and Percutaneous Pulmonary Valve Implantation while performance evaluation is conducted on a population of 476 patients.

/pdf/morphologica-l-and-functional-modeling-of-the-heart-valves-4r1wy54iet.pdf

Morphologica l and Functional Modeling of the Heart Valves and Chambers

Systems and methods are provided for generating a synthesized medical image patch of a nodule. An initial medical image patch and a class label associated with a nodule to be synthesized are received. The initial medical image patch has a masked portion and an unmasked portion. A synthesized medical image patch is generated using a trained generative adversarial network. The synthesized medical image patch includes the unmasked portion of the initial medical image patch and a synthesized nodule replacing the masked portion of the initial medical image patch. The synthesized nodule is synthesized according to the class label. The synthesized medical image patch is output.

Class-Aware Adversarial Pulmonary Nodule Synthesis

Understanding and predicting atrial electrophysiology, for diagnosis and therapy planning purposes, calls for methods able to accurately represent the complex patterns of atrial electrical activity, and to produce very fast predictions to be suitable for use in the clinical practice. We apply a data-driven approach for the model reduction of an atrial cellular model. The reduced model predicts cellular action potentials (AP) in a simple form but is effective in capturing the physiological complexity of the original model. The model construction starts from an AP manifold learning which reduces the AP manifold dimension to 15, and continues with a regression model learning to predict the 15 components in the reduced AP manifold. The regression model has the potential to drastically improve the performance of atrial tissue-level electrophysiology (EP) modeling, enabling a 75 % reduction of the computational cost with the same time step and up to two order of magnitudes smaller computational time with larger time steps. The model is also capable of describing the restitution properties of the AP, as demonstrated in tests with varying diastolic intervals. This model has great potential use for real-time personalized atrial EP modeling, and the same modeling technique can be extended to the study of other excitable myocardial tissues.

Data-Driven Model Reduction for Fast, High Fidelity Atrial Electrophysiology Computations

An intelligent medical imaging scanner system includes an image scanner, an operator interface, a database, processors, and a storage medium. The database includes a learning model for relating configurations of the image scanner to operator input requirements. The storage medium contains programming instructions that, when executed, cause the processors to determine whether the learning model may be used to generate a configuration of the image scanner corresponding to the new input requirements. If the configuration can be generated, the processors use that configuration to acquire images of a patient using the image scanner. If the configuration of the image scanner cannot be generated, the processors perform an accommodation process comprising (a) modifying the learning model to generate a new configuration of the image scanner corresponding to the new input requirements, and (b) using the new configuration to acquire the images of the patient using the image scanner.

Biologically inspired intelligent body scanner

PROBLEM TO BE SOLVED: To detect a boundary of vessel correctly and certainly. SOLUTION: The method and system for detecting vessel boundary and a medium capable of reading out from a computer are characterized by the fact that a plurality of edges in the image are detected based on the change in intensity between data points over some distance, and then a set of edges is selected from the plurality of edges, and an initial vessel boundary is determined based on the selected set of edges, and a shape descriptor is applied to the initial vessel boundary to determine a final vessel boundary. COPYRIGHT: (C)2007,JPO&INPIT

Dorin Comaniciu

Papers

Morphologica l and Functional Modeling of the Heart Valves and Chambers

Class-Aware Adversarial Pulmonary Nodule Synthesis

Data-Driven Model Reduction for Fast, High Fidelity Atrial Electrophysiology Computations

Biologically inspired intelligent body scanner

Method and system for detecting vessel boundary