Hrvoje Kalinić

Self-Organizing Maps-based ocean currents forecasting system

An ocean surface currents forecasting system, based on a Self-Organizing Maps (SOM) neural network algorithm, high-frequency (HF) ocean radar measurements and numerical weather prediction (NWP) products, has been developed for a coastal area of the northern Adriatic and compared with operational ROMS-derived surface currents. The two systems differ significantly in architecture and algorithms, being based on either unsupervised learning techniques or ocean physics. To compare performance of the two methods, their forecasting skills were tested on independent datasets. The SOM-based forecasting system has a slightly better forecasting skill, especially during strong wind conditions, with potential for further improvement when data sets of higher quality and longer duration are used for training.

Sensitivity of Self-Organizing Map surface current patterns to the use of radial vs. Cartesian input vectors measured by high-frequency radars

In this paper, the Self-Organizing Map (SOM) method was applied to the surface currents data obtained between February and November 2008 by a network of high-frequency (HF) radars in the northern Adriatic. The sensitivity of the derived SOM solutions was tested in respect to the change of coordinate system of the data introduced to the SOM. In one experiment the original radial data measurements were used, and in the other experiment the Cartesian (total) current vectors derived from original radar data were analyzed. Although the computation of SOM solutions was not a demanding task, comparing both neural lattices yielded the nondeterministic polynomial time (NP) problem for which is difficult to propose a solution that will be globally optimal. Thus, we suggested utilizing the greedy algorithm with underlying assumption of 1-to-1 mapping between lattices. The results suggested that such solution could be local, but not global optimum and that the latter assumption could lower the obtained correlations between the patterns. However, without the assumption of 1-to-1 mapping between lattices, correlation between the derived SOM patterns was quite high, indicating that SOM mapping introduced to the radial current vectors and subsequent transformation into Cartesian coordinate system does not significantly affect obtained patterns in comparison to the SOM mapping done on the derived Cartesian current vectors. The documented similarity corroborates the use of total current vectors in various oceanographic studies, as being representative derivative of original radial measurements. We analyzed high-frequency radar data in the northern Adriatic.Self-Organizing Map method has been applied to radial and Cartesian current vectors.Transformation to Cartesian vectors do not significantly affect the SOM patterns.

The shape of the aortic outflow velocity profile revisited: is there a relation between its asymmetry and ventricular function in coronary artery disease?

AIMS Myocardium contracts in the beginning of ejection causing outflow acceleration, resulting in asymmetric outflow velocity profiles peaking around one-third of ejection and declining when force development declines. This article aimed to demonstrate that decreased contractility in coronary artery disease (CAD) changes outflow timing and profile symmetry. METHODS AND RESULTS Seventy-nine patients undergoing routine full dose dobutamine stress-echo (DSE) were divided into two groups based on resting wall motion and DSE response: DSE negative (DSE(neg)) (35 of 79 patients) and positive (DSE(pos)) (44 of 79 patients) which were compared with 32 healthy volunteers. Aortic CW-Doppler traces at rest were analysed semi-automatically; time-to-peak (T(mod)), ejection-time (ET(mod)), rise-time (t(rise)), and fall-time (t(fall)) were quantified. Asymmetry (asymm) was calculated as the normalized difference of left and right half of the spectrum. Normal curves were triangular, early-peaking, whereas patients showed more rounded shapes and later peaks. T(rise) was longest in DSE(pos). T(fall) was shortest in DSE(pos), followed by controls and DSE(neg). Asymm was lowest in DSE(pos), followed by controls and DSE(neg). Abnormally symmetric profiles (asymm <0.25) were found in none of the controls, 2.9% DSE(neg), and 27.3% DSE(pos). A good correlation was found between assym and ejection fraction (EF) and T(mod)/ET(mod) and EF. Notably, an LV dynamic gradient was induced in 71.4% DSE(neg) and in 18.2% DSE(pos), associated with LV hypertrophy and supernormal (very asymmetric) traces. CONCLUSION Decreased myocardial function results in a more symmetrical outflow, while very asymmetrical traces suggest increased contractility, potentially inducing intra-cavity gradients during DSE. Therefore, including outflow symmetry as a clinical measurement provides additional information on patients with CAD.

Image registration and atlas-based segmentation of cardiac outflow velocity profiles

Cardiovascular disease is the leading cause of death worldwide and for this reason computer-based diagnosis of cardiac diseases is a very important task. In this article, a method for segmentation of aortic outflow velocity profiles from cardiac Doppler ultrasound images is presented. The proposed method is based on the statistical image atlas derived from ultrasound images of healthy volunteers. The ultrasound image segmentation is done by registration of the input image to the atlas, followed by a propagation of the segmentation result from the atlas onto the input image. In the registration process, the normalized mutual information is used as an image similarity measure, while optimization is performed using a multiresolution gradient ascent method. The registration method is evaluated using an in-silico phantom, real data from 30 volunteers, and an inverse consistency test. The segmentation method is evaluated using 59 images from healthy volunteers and 89 images from patients, and using cardiac parameters extracted from the segmented image. Experimental validation is conducted using a set of healthy volunteers and patients and has shown excellent results. Cardiac parameter segmentation evaluation showed that the variability of the automated segmentation relative to the manual is comparable to the intra-observer variability. The proposed method is useful for computer aided diagnosis and extraction of cardiac parameters.

Sensitivity of HF radar-derived surface current self-organizing maps to various processing procedures and mesoscale wind forcing

We performed a number of sensitivity experiments by applying a mapping technique, self-organizing maps (SOM) method, to the surface current data measured by high-frequency (HF) radars in the northern Adriatic and surface winds modelled by two state-of-the-art mesoscale meteorological models, the Aladin (Aire Limitée Adaptation Dynamique Développement InterNational) and the Weather and Research Forecasting models. Surface current data used for the SOM training were collected during a period in which radar coverage was the highest: between February and November 2008. Different pre-processing techniques, such as removal of tides and low-pass filtering, were applied to the data in order to test the sensitivity of characteristic patterns and the connectivity between different SOM solutions. Topographic error did not exceed 15 %, indicating the applicability of the SOM method to the data. The largest difference has been obtained when comparing SOM patterns originating from unprocessed and low-pass filtered data. Introduction of modelled winds in joint SOM analyses stabilized the solutions, while sensitivity to wind forcing coming from the two different meteorological models was found to be small. Such a low sensitivity is considered to be favourable for creation of an operational ocean forecasting system based on neural networks, HF radar measurements and numerical weather prediction mesoscale models.

A method for registration and model-based segmentation of Doppler ultrasound images

Morphological changes of Doppler ultrasound images are an important source of information for diagnosis of cardiovascular diseases. Quantification of these flow profiles requires segmentation of the ultrasound images. In this article, we propose a new model-based method for segmentation of (aortic outflow) velocity profiles. The method is based on a procedure for registration using a geometric transformation specifically designed for matching Doppler ultrasound profiles. After manual segmentation of a model image, the model image is temporarily registered to a new image using two manually defined points in time. Next, a non-rigid registration was carried out in the velocity direction. As a similarfity measure normalized mutual information is used, while optimization is performed by a genetic algorithm. The registration method is experimentally validated using an in-silico image phantom, and showed an accuracy of 5.4%. The model based on segmentation is evaluated in a seris of aortic outflow Doppler ultrasound images from 30 normal volunteers. Comparing the automated method to the manual delineation by an expert cardiologist the method proved accurate to 6.6%. The experimental results confirm the accuracy of the approach and shows that the method can be used for the segmentation of the clinically obtained aortic outflow velocity profiles.

Analysis of Doppler Ultrasound Outflow Profiles for the Detection of changes in Cardiac Function

Detecting changes in the contractility of the heart muscle, especially in the presence of coronary artery disease, is an important medical task. From isolated myocytes, it was suggested that chronic ischemia decreases but prolongs contraction. Additionally, severe aortic stenosis shows higher but often prolonged outflow velocities. From this, we hypothesize that there is a correlation between the morphology of the aortic outflow velocity profile and myocardial function. To test this hypothesis, we (semi-) automatically analyzed continuous wave Doppler aortic outflow traces from 112 individuals. The traces were segmented and a set of morphology features were extracted. Areas under ROC curve were used as a measure of quality of each feature for detecting changes in cardiac function related to coronary artery disease. Signal analysis has shown that the probabilistic distributions of the various features are different for normal individuals and patients with ongoing disease. This result shows that aortic outflow profiles provide information on cardiac function and that the presented signal analysis and feature extraction method might be used to provide additional diagnostic information in the clinical management of coronary artery disease.

Atlas construction for cardiac velocity profiles segmentation using a lumped computational model of circulatory system

Heart diseases are a leading cause of death worldwide, making a prompt and accurate diagnosis of cardiac functionality an important task. Recordings of cardiac outflow Doppler velocity profiles, obtained during an echocardiographic examination, are important to quantify hemodynamics and infer cardiac function. For automated segmentation and quantification of these images, a statistical atlas based approach has been proposed previously. Since acquiring a sufficient amount of data for an atlas can be a slow process in clinical practice and possibly result in a small and/or not representative dataset, we present an alternative approach for construction of the statistical atlas. This approach is based on simulating data from virtual patients, using a lumped computational model (CircAdapt), which incorporates knowledge of physiological processes in the human circulatory system under both normal and pathological conditions.

Absolute joint moments: a novel image similarity measure

In this paper, we propose a novel approach for estimating image similarity. This measure is of importance in assessing image correspondence or image alignment and plays an important role in image registration. Currently, this problem is approached rather one-dimensionally since most registration methods consider the problem as either mono- or multi-modal. This perspective leads to the selection of some form of either the correlation coefficient (CC) or mutual information (MI) as image similarity measure (ISM). We propose a more generic framework for ISM construction, based on absolute joint moments, which can be considered as a generalization of CC. Within this framework, we propose a specific ISM that provides a different trade-off between MI and CC in terms of performance and computational cost for general registration problems. To illustrate this, we compared CC and MI with the proposed ISM and performed extensive experiments with regard to accuracy, robustness and speed. The evaluation demonstrated that the proposed absolute joint moments is a good combination of properties of CC and MI, with respect to speed and performance.

Interpreting Self-Organizing Map errors in the classification of ocean patterns

Abstract The paper aims to introduce quality measures that can evaluate how well the Self-organizing Maps method performs in transitional stages. The errors have been computed with respect to the spatial and temporal properties of the data and in relation to the data gap significance. Temperature and salinity data collected in the central Adriatic Sea at six stations during 196 field cruises carried out between 1963 and 2011 have been used for the mapping of ocean patterns and computation of the respective errors. The errors resemble both the stability of ocean regimes and variability of patterns that are documented in the investigated region. As the data collection methodology and approach have changed over time, the errors may be a good indication for the presence of bad data in a series, which may then be controlled by other quality-check techniques.