Dicentric chromosome analysis is the gold standard for biological dose assessment. To enhance the efficiency of biological dose assessment in large-scale radiation catastrophes, automatic identification of dicentric chromosome images is a promising and objective method. In this paper, an automatic identification method for dicentric chromosome images using two-stage convolutional neural network is proposed based on Giemsa-stained automatic microscopic imaging. To automatically segment the adhesive chromosome masses, ak-means based adaptive image segmentation and watershed segmentation algorithm is applied. The first-stage CNN is used to identify the dicentric chromosome images from all the images and the second-stage CNN works to specifically identify the dicentric chromosome images. This two-stage CNN identification method can effectively detects chromosome images with concealed centromeres, poorly expanded and long-armed entangled chromosomes, and tricentric chromosomes. The novel two-stage CNN method has a chromosome identification accuracy of 99.4%, a sensitivity of 85.8% sensitivity, and a specificity of 99.6%, effectively reducing the false positive rate of dicentric chromosome. The analysis speed of this automatic identification method can be 20 times quicker than manual detection, providing a valuable reference for other image identification situations with small target rates.

A method is proposed for the estimation of the time difference of arrival (TDOA) from a sound source to a pair of microphones. Given noisy observations of the source, the magnitude spectrum of the source is first estimated via smoothing across time frames. Then, a probability density function (PDF) for the phase conditioned upon the magnitude is constructed based on the signal-to-noise ratio (SNR) at every frequency. Subsequently, the conditional PDF for the inter-channel phase difference (IPD) is calculated, and the computation can be accelerated via Gaussian curve fitting. Finally, by combining the information from all frequencies, the TDOA can be estimated in the maximum a posteriori (MAP) sense. Results from "anechoic" simulation showed that, at various SNRs (0 to 40 dB), the proposed method consistently produced more accurate estimation than the well known GCC-PHAT [1] and a recent method that was also based on IPD modeling [2]. Experiments conducted in an office environment are also reported, using speech and footsteps as test materials.