Combining Multiple Dynamic Models and Deep Learning Architectures for Tracking the Left Ventricle Endocardium in Ultrasound Data

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 35; no. 11; pp. 2592 - 2607
• Main Authors: Carneiro, G., Nascimento, J. C.
• Format: Journal Article
• Language: English
• Published: Los Alamitos, CA IEEE 01.11.2013; IEEE Computer Society
• Subjects: Algorithms; Applied sciences; Artificial Intelligence; Biological and medical sciences; Cardiovascular system; Computational modeling; Computer science; control theory; systems; Computer Simulation; Connectionism. Neural networks; deep belief networks; discriminative classifiers; dynamical model; Echocardiography - methods; Endocardium - diagnostic imaging; Exact sciences and technology; Heart Ventricles - diagnostic imaging; Humans; Image Interpretation, Computer-Assisted - methods; Image segmentation; Imaging; Investigative techniques, diagnostic techniques (general aspects); Learning and adaptive systems; Left ventricle segmentation; Medical sciences; Models, Cardiovascular; Motion segmentation; particle filters; Pattern Recognition, Automated - methods; Pattern recognition. Digital image processing. Computational geometry; Shape; Tracking; Training; Ultrasonic investigative techniques; Ventricular Dysfunction, Left - diagnostic imaging; Great depth; Left ventricle segmentation; Tracking; Image processing; Ultrasound imaging; Endocardium; discriminative classifiers; Modeling; Belief networks; Sequential method; Classification; Cardiology; Dynamic model; Ultrasound; Particle filter; Importance sampling; Multimodel; Statistical analysis; Probabilistic approach; Healthy subject; Pattern recognition; Neural network; Experimental study; dynamical model; Left ventricle; Image segmentation; deep belief networks; Remolded sample; particle filters; Tracking(movable target); Time complexity; Edge detection; Artificial intelligence
• ISSN: 0162-8828; 1939-3539; 2160-9292; 1939-3539
• DOI: 10.1109/TPAMI.2013.96

We present a new statistical pattern recognition approach for the problem of left ventricle endocardium tracking in ultrasound data. The problem is formulated as a sequential importance resampling algorithm such that the expected segmentation of the current time step is estimated based on the appearance, shape, and motion models that take into account all previous and current images and previous segmentation contours produced by the method. The new appearance and shape models decouple the affine and nonrigid segmentations of the left ventricle to reduce the running time complexity. The proposed motion model combines the systole and diastole motion patterns and an observation distribution built by a deep neural network. The functionality of our approach is evaluated using a dataset of diseased cases containing 16 sequences and another dataset of normal cases comprised of four sequences, where both sets present long axis views of the left ventricle. Using a training set comprised of diseased and healthy cases, we show that our approach produces more accurate results than current state-of-the-art endocardium tracking methods in two test sequences from healthy subjects. Using three test sequences containing different types of cardiopathies, we show that our method correlates well with interuser statistics produced by four cardiologists.