Abstracts:Bowl-shaped phased array therapeutic probes are becoming a standard module in magnetic resonance imaging-guided high-intensity focused ultrasound (MRgHIFU) systems because single beam focus can be electronically steered and multiple foci can be simultaneously synthesized. B-mode imaging provides only two dimensional image, therefore phased array HIFU probes are rarely used in ultrasound imaging-guided HIFU (USgHIFU) systems. An 144-element phased-array USgHIFU system was developed, and a B-mode probe was mounted in a rotating platform and harnessed to track the focus steered in three-dimensional volume. In this study, the targeting accuracy of this system was estimated experimentally using beef-embedded phantoms with four reference markers. The plane of treatment crossing the reference markers was reconstructed by acquired image sequence, and targeting area was predetermined as circular treated zone in the center of a square with corners in the middle of the markers. After sonication the contour of lesion was compared with the outline of planned targeting area. For three treated zones with a diameter of 5.8 mm, 9.7 mm, and 13.5 mm, the ratio of the thermal lesion area in the relevant treated zone to the planned targeting area was 89%, 92%, and 96%, respectively. The displacements between centers of the lesions and of the planned treated zones were no more than 1 mm. Moreover, the maximal crossing of borders outside the targeting area was lower than 2 mm. The targeting accuracy of this system was found to be around 1 mm being comparable to other image-guided HIFU systems. Therefore, the rotation-based 3D image reconstruction is effective for the targeting, treatment planning, and monitoring by using our system. Besides, such arrangement of treated zones is simple but useful in treatment planning.

Abstracts:Passive acoustic observation of whales is an increasingly important tool for whale research. Clicks are the predominant vocalizations of toothed whales, such as sperm whales and long-finned pilot whales. Classifying clicks of sperm whales and long-finned pilot whales is an essential task for the passive acoustic observation of the two whale species, especially in the case that both whale species vocalize in the same observed area. In this paper, we proposed a method performing the automated classification of clicks produced by sperm whales and long-finned pilot whales. First, the two types of whales’ original sounds were denoised using a wavelet denoising method. Then, a dual-threshold endpoint detection algorithm was utilized to detect and pick out all clicks from the denoised sounds. The continuous wavelet transform was applied to decompose the picked clicks, and a wavelet coefficient matrix can be obtained for each picked click. Focusing on the energy distribution and duration difference between the two types of whales’ clicks, we proposed a feature-vector extraction algorithm based on the wavelet coefficient matrix. For each picked click, scale (frequency) features and time feature were obtained respectively and they were used to form the feature vector. Finally, a back propagation (BP) neural network was designed as a classifier of feature-vector to output final classification result. The experiment results show the proposed method can obtain high classification performances. The effect of training dataset size, and the number of training features on the classification performance was also examined in the experiments.

Abstracts:An efficient integrated system for acoustic echo and noise cancellation in hands free devices is proposed in this paper. Adaptive filters are usually employed in such systems for the acoustic echo cancellation. In order to reduce the computational complexity, sub band adaptive filters are commonly used since the order of filter required will be less but at the cost of distortion due to aliasing effect. In this work, cross-band adaptive filters are developed in wavelet domain/wavelet packet domain in tune with psychoacoustic model, to reduce the above distortion and thereby effectively estimate the echo and cancel it. The system uses wavelet thresholding technique for smoothening the log multitaper spectrum which eliminates the annoying musical noise in the processed output. To achieve efficient echo and noise cancellation, a novel switching mechanism to switch between wavelet filter bank and perceptual wavelet packet filter bank with most matching wavelet basis, from a dictionary of wavelet bases, depending on the type of noise present in each frame of the noisy speech is designed and developed in this work. The performance of the proposed integrated system is evaluated in terms of Acoustic Interference Cancellation (AIC) and Noise Attenuation (NA) under different Signal to Noise Ratio (SNR) conditions. The system exhibits very good performance even at very low SNRs close to −10 dB when compared with other such competitive systems.

Abstracts:Cross-modal effects of noise and thermal conditions on subjective indoor environmental perception and speech recognition were investigated with 24 native Korean speakers in an indoor environmental chamber. Sixteen combinations of four Predicted Mean Votes (PMV) values (−1.53, 0.03, 1.53, 1.83), two noise levels (45 dBA, 60 dBA) and two noise types (babble, fan) were tested at the noise exposure times of 2 min and 55 min. Speech-to-noise ratios (SNRs) were set to 0, 5 and 10 dB for speech recognition tests based on the presented noise levels. It was found that noise had no effect on thermal sensation and thermal conditions had no effect on loudness and noisiness. However, effects of noise level and type on thermal comfort and effects of PMV on acoustic comfort and annoyance were statistically significant. Thermal comfort decreased with increased noise level. At thermoneutrality, acoustic comfort increased and annoyance decreased. Speech recognition was affected by not only SNR ratio but also by the PMV values according to the room acoustical conditions.