Abstract: Recently, low-dose computed tomography (CT) has become highly desirable due to increasing attention to the potential risks of excessive radiation. For low-dose CT imaging, ensuring image quality while reducing radiation dose is a major challenge. To facilitate low-dose CT imaging, we propose an improved statistical iterative reconstruction scheme based on the Penalized Weighted Least Squares (PWLS) standard combined with total variation (TV) minimization and sparse dictionary learning (DL) to improve reconstruction performance. We call this method "PWLS-TV-DL". In order to evaluate the PWLS-TV-DL method, we performed experiments on digital phantoms and physical phantoms, respectively. The experimental results show that our method is in image quality and calculation. The efficiency is superior to other methods, which confirms the potential of its low-dose CT imaging.
Abstract: In electromagnetic imaging, because of the diffraction limited system, the pixel values could change slowly near the edge of the image targets and they also change with the location in the same target. Using traditional digital image segmentation methods to segment electromagnetic gradient images could result in lots of errors because of this change in pixel values. To address this issue, this paper proposes a novel image segmentation and extraction algorithm based on Mean-Shift and dictionary learning. Firstly, the preliminary segmentation results from adaptive bandwidth Mean-Shift algorithm are expanded, merged and extracted. Then the overlap rate of the extracted image block is detected before determining a segmentation region with a single complete target. Last, the gradient edge of the extracted targets is recovered and reconstructed by using a dictionary-learning algorithm, while the final segmentation results are obtained which are very close to the gradient target in the original image. Both the experimental results and the simulated results show that the segmentation results are very accurate. The Dice coefficients are improved by 70% to 80% compared with the Mean-Shift only method.
Abstract: In this paper, a data-driven dictionary approach is proposed for the automatic detection and classification of cardiovascular abnormalities. Electrocardiography (ECG) signal is represented by the trained complete dictionaries that contain prototypes or atoms to avoid the limitations of pre-defined dictionaries. The data-driven trained dictionaries simply take the ECG signal as input rather than extracting features to study the set of parameters that yield the most descriptive dictionary. The approach inherently learns the complicated morphological changes in ECG waveform, which is then used to improve the classification. The classification performance was evaluated with ECG data under two different preprocessing environments. In the first category, QT-database is baseline drift corrected with notch filter and it filters the 60 Hz power line noise. In the second category, the data are further filtered using fast moving average smoother. The experimental results on QT database confirm that our proposed algorithm shows a classification accuracy of 92%.
Abstract: Speaker recognition is performed in high Additive White Gaussian Noise (AWGN) environments using principals of Computational Auditory Scene Analysis (CASA). CASA methods often classify sounds from images in the time-frequency (T-F) plane using spectrograms or cochleargrams as the image. In this paper atomic decomposition implemented by matching pursuit performs a transform from time series speech signals to the T-F plane. The atomic decomposition creates a sparsely populated T-F vector in “weight space” where each populated T-F position contains an amplitude weight. The weight space vector along with the atomic dictionary represents a denoised, compressed version of the original signal. The arraignment or of the atomic indices in the T-F vector are used for classification. Unsupervised feature learning implemented by a sparse autoencoder learns a single dictionary of basis features from a collection of envelope samples from all speakers. The approach is demonstrated using pairs of speakers from the TIMIT data set. Pairs of speakers are selected randomly from a single district. Each speak has 10 sentences. Two are used for training and 8 for testing. Atomic index probabilities are created for each training sentence and also for each test sentence. Classification is performed by finding the lowest Euclidean distance between then probabilities from the training sentences and the test sentences. Training is done at a 30dB Signal-to-Noise Ratio (SNR). Testing is performed at SNR’s of 0 dB, 5 dB, 10 dB and 30dB. The algorithm has a baseline classification accuracy of ~93% averaged over 10 pairs of speakers from the TIMIT data set. The baseline accuracy is attributable to short sequences of training and test data as well as the overall simplicity of the classification algorithm. The accuracy is not affected by AWGN and produces ~93% accuracy at 0dB SNR.
Abstract: Clusters of microcalcifications in mammograms are an
important sign of breast cancer. This paper presents a complete
Computer Aided Detection (CAD) scheme for automatic detection of
clustered microcalcifications in digital mammograms. The proposed
system, MammoScan μCaD, consists of three main steps. Firstly
all potential microcalcifications are detected using a a method for
feature extraction, VarMet, and adaptive thresholding. This will also
give a number of false detections. The goal of the second step,
Classifier level 1, is to remove everything but microcalcifications.
The last step, Classifier level 2, uses learned dictionaries and sparse
representations as a texture classification technique to distinguish
single, benign microcalcifications from clustered microcalcifications,
in addition to remove some remaining false detections. The system
is trained and tested on true digital data from Stavanger University
Hospital, and the results are evaluated by radiologists. The overall
results are promising, with a sensitivity > 90 % and a low false
detection rate (approx 1 unwanted pr. image, or 0.3 false pr. image).
Abstract: Sparse representation has long been studied and several
dictionary learning methods have been proposed. The dictionary
learning methods are widely used because they are adaptive. In this
paper, a new dictionary learning method for audio is proposed. Signals
are at first decomposed into different degrees of Intrinsic Mode
Functions (IMF) using Empirical Mode Decomposition (EMD)
technique. Then these IMFs form a learned dictionary. To reduce the
size of the dictionary, the K-means method is applied to the dictionary
to generate a K-EMD dictionary. Compared to K-SVD algorithm, the
K-EMD dictionary decomposes audio signals into structured
components, thus the sparsity of the representation is increased by
34.4% and the SNR of the recovered audio signals is increased by
20.9%.