No abstract

In order to elucidate the sterilization mechanism underlying the explosive decompression system, baker's yeast was pressurized with COz, NzO, N z , or Ar gas at 40atm and 40°C for 4h, and then explosively discharged. The survival ratio was markedly decreased only by the treatments with COz and NzO, which are relatively soluble gases in water, suggesting that the microorganisms' death may be highly correlated with gas absorption by the cells. Lower decompression rates to atmospheric pressure, however, led to neither any lower reduction of remaining cells nor any smaller release of total cellular proteins. Furthermore, operating with a longer treatment time and smaller number of repetitions was usually more lethal than with a shorter time and more frequent repetition. From these results, most of the yeast cells appear to have been sterilized during the pressurization process. The spore cells of B. megaterium are considered to have been killed in a somewhat different manner, because of their distinct sensitivity to the applied gases.Key words: sterilization; high-pressure carbon dioxide; food microorganism; survival ratio; decompression rateAlthough heat sterilization is the traditional and most popular method for protecting foods from microbial spoilage, it may often cause hostile changes in the nutritional and sensory properties of the products because of the high temperature involved. To develop and establish a novel and effective alternative to heat treatment, the lethal action of high-pressure COz on microorganisms, with no or only a minimal heating process, has recently received a great deal of attention.Among the methods, the explosive decompression system consists of a pressurization stage to induce penetration of the applied gas into the microbial cells and a subsequent explosive discharge that results in rapid gas expansion within the cells. With this system, the greater part of the microbial cells are believed to be mechanically ruptured like a popped ballon at the moment of flash depressurization.1 6) This technique was first reported by Fraser in 1951, and he found that E. coli cells were significantly disrupted by the quick release of COz gas from 5001bf/in 2 (about 34.0 atm) to atmospheric pressure. 1) This method has subsequently been developed mainly for unit operation to recover intracellular enzymes, recombinant-DNA proteins and nucleic acids from microbial cell cultures.1 5) We have recently demonstrated with a similar system that 10 8 cells/ml of baker's yeast could be thoroughly killed after COz saturation at 40 atm and 40°C for 3 h, and proposed this as a new method for sterilizing food microorganisms. 7.8) Other approaches involving no flash pressure drop, using high-pressure COz treatment alone, have also been widely studied to provide the required lethal action on various kinds of microbes. 9 -17) 15 ) From these investigations, most microbial cells appear to be sterilized mainly due to the inactivating effect of COz under pressure; this may be in conflict with the foregoing explanatio...

No abstract

No abstract

This apparatus takes in a signal containing speech mixed with certain nonspeech audio components and separates the speech from the nonspeech, producing both separated portions on the two output lines. A cepstral analysis is performed on frames of the input signal and two measures, the peak cepstral value and the average cepstral value, are extracted. A cepstral peak indicates a vowel sound. Frames with a high average and a low peak are further discriminated based on the rate of change of the spectral structure. It is assumed, in other words, that the nonspeech sounds exhibit slower spectral changes than the noise portions of speech.--DLR 5,185,848 43.72.Ew NOISE REDUCTION SYSTEM USING NEURAL NETWORK Toshiyuki Aritsuka et aL, assignors to Hitachi, Limited 9 February 1993 (Class 395/2); filed in Japan 14 December 1988 This patent presents an implementation of a neural network technique for performing a vector quantization of speech analysis feature vectors. The method is said to result in noise reduction in the speech signal and to be applicable to speech coding or recognition systems. The NEURON __ INPUT HIDDEN OUTPUT LAYER LAYERS LAYER discussion on training the network is primarily concerned with the management of noisy and noise-free speech data to construct training and test data sets. There is no explicit discussion of either network training algorithms or internal network structure.--DLR with a time mark relative to the most recent change of voicing. This enables the receiver to determine quickly when a new speech event has started and to fill in, if necessary, for late arriving packets.--DLR 5,148,431 43.72.Gy SPEECH SIGNAL DECODING APPARATUS WITH ERROR DETECTION AND CORRECTION Ryoji Hayashi, assignor to Mitsubishi Denki Kabushiki Kaisha 15 September 1992 (Class 371/5.1); filed in Japan 15 May 1989 Error-correcting coding (ECC) methods introduce an interesting sort of modularity in the number of errors which are detectable and the number which are correctable. If too many errors occur, the received signal passes the test as if nothing were wrong. This condition can result in bursts of severe noise when ECC is used in packet speech transmission. This patent proposes a way of detecting the condition by accumulating the numbers of detectable and correctable errors in the surrounding packets.--DLR 5,148,487 43.72.Gy AUDIO SUBBAND ENCODED SIGNAL DECODER Kiyotaka Nagai and Yasushi Nakajima, assignors to Matsushita Electric Industrial Company 15 September 1992 (Class 381/46); filed in Japan 26 February 1990This patent applies to subband vocoders in which the subband signals are packetized, each being separately encoded using an errorcorrecting code. When an uncorrectable number of errors occurs in a subband signal, the previous approach would have been to reset that band to zero, resulting in noticeable disruption of the speech output. As disclosed here, an erroneous subband signal can, instead, be reconstructed by interpolation of the preceding and following frames.--DLR 5,148,429 43.72.Gy VOICE DATA TRANSMISSION...