No abstract

Although it is well known that the slopes of target strength (TS) and length relationships vary widely, it is common in fisheries acoustics to force the TS–length regression through a slope of 20. Is it time to abandon this practice? The theoretical justification was that TS should be proportional to cross-sectional area, and that area should scale as the square of the linear dimension (fish length). There are now many species other than gadoids that are the subject of acoustic surveys, and many of them do not have the same morphology as the gadoid fishes. The slope of the TS–length regressions deviates significantly from 20. The empirical slope should be used wherever it can be shown to be more appropriate than the 20 log10 L model. Using the data from swimbladder models, it is shown that Macrourids, a merluccid hake and Oreosomatidae have a different relationship should between swimbladder size and fish size compared with that of gadoids. It is demonstrated that the 20 log10 L model is not appropriate for these deep-water fish and that deviations from the model arise, to a considerable degree, from variation in fish morphotypes. The TS of deep-water Macrourids, a merluccid hake and Oreosomatidae are lower than that of gadoids. This is related to the swimbladder size–fish size relationship in different morphotypes, although not much evidence can be found to support the concept that swimbladder sizes are generally smaller in deep-sea fishes.

No abstract

No abstract

We have been investigating the use of electrical impedance methods for the study of cerebral haemodynamics in the newborn for many years. Of particular interest has been the early detection of intraventricular haemorrhage, which is a major cause of death or handicap in low birthweight infants. Several problems exist in obtaining representative impedance measurements from the newborn, most notably movement artefact, respiratory-based modulation and a blood-flow related pulsatile component. Movement artefact is by far the most significant problem; in order to perform long-term monitoring we have developed an effective algorithm for rejection of corrupted data. To remove the respiratory component, which can be 1-2% of the impedance measurement, we link to the output of a respiratory monitor (standard in intensive care units) as a means of synchronising measurements to a fixed point in the respiratory cycle. Similarly, to remove the blood-flow pulsatile component, measurements are gated from the R-wave of the recorded ECG. Our total system consists of a front-end measurement system that passes image data over a serial link to a host computer. Data storage, image reconstruction and display is performed on the host, which can be any of a wide range of personal computers. The front-end contains the impedance measurement circuit, electrode switching electronics and a microprocessor for control, and is of a sufficiently small size to fit into the incubator next to the baby. Incorporating a microprocessor into the front-end produces a very flexible system that has many benefits, including: asynchronous operation from host, 'intelligent' pre-processing of measurements, command driven operation from host, etc. Software development for the front-end is performed on the host with program down-load for interactive debugging. Details of the front-end electronics and software, system performance, preliminary clinical results and other application areas of the impedance imaging technique to the care of the newborn are presented.