Abstract: Diagnostic tasks require determining the differences between a model of an artifact and the artifact itself. The differences between the manifested behavior of the artifact and the predicted behavior of the model guide the search for the differences between the artifact and its model. The diagnostic procedure presented in this paper is model-based, inferring the behavior of the composite device from knowledge of the structure and function of the individual components comprising the device. The system (GDE — General Diagnostic Engine) has been implemented and tested on examples in the domain of troubleshooting digital circuits. This research makes several novel contributions: First, the system diagnoses failures due to multiple faults. Second, failure candidates are represented and manipulated in terms of minimal sets of violated assumptions, resulting in an efficient diagnostic procedure. Third, the diagnostic procedure is incremental, reflecting the iterative nature of diagnosis. Finally, a clear separation is drawn between diagnosis and behavior prediction, resulting in a domain (and inference procedure) independent diagnostic procedure.

Abstract: For suppressing cardiovascular artifact from a respiration signal derived from a patient's transthoracic impedance, an adaptive filtering device determines the patient's heart rate, converts the respiration signal from an analog to digital representation at a rate proportional to the heart rate, and then filters the digital respiration signal such that that portion of the respiration signal having a frequency content at or above the heart rate is most greatly attenuated.

Abstract: This paper presents a critical discussion of methods used for the recording and the automated detection of sleep rapid eye movements (REMs) via electro-oculographic (EOG) methodologies. Specifically, the paper elaborates on: 1) various EOG montages for sleep eye movement recording, 2) AC vs. DC coupling for EOG signal amplification, 3) artifact presence and elimination thereof via filtering of the EOG signal, and 4) available EOG-based automated methods for REM detection. Recommendations for the development of an “optimum’ automatic sleep REM detection system based on EOG recordings are presented as well.

Abstract: A boundary artifact in MR images due to truncation of the infinite Fourier series necessary to encode tissue discontinuities was investigated by using doped water phantoms and normal volunteers. All images were obtained on 0.3-T permanent and 0.6-T superconducting MR imagers with varying phase and frequency sampling rates. The artifact appeared in both the phase and frequency encoding direction as parallel lines or ringing adjacent to borders or tissue discontinuities. This was unlike motion artifacts, which occur predominantly in the phase direction, and chemical shift misregistration errors, which are most pronounced in the frequency direction. Increasing the sampling frequency from 128 to 512 resulted in higher frequency ringing and more rapid drop-off in amplitude. Low-pass digital filtering also decreased the ringing at the expense of fine detail. The truncation of the infinite Fourier series necessary to encode edges to the 128-512 terms used for most MR imaging produces the artifact. It is important to recognize this common artifact and not mistake it for patient motion or disease.

Abstract: Pseudomonilethrix is a microscopic hair shaft abnormality characterized by small beaded globular swellings. It is an artifact produced while preparing hairs for microscopic examination and occurs when overlapped hairs are pressed together between glass slides. Evidence documenting this finding is presented.

Abstract: A high-speed multichannel signal processing system is described which is capable of performing automated detection of epileptogenic sharp transients (ST) in the electroencephalogram (EEG). The system is implemented with individually programmable microprocessors on the input channels, followed by a single-board microcomputer which correlates results obtained from each channel, and can process data played back from a tape recorder at a speed eight times the realtime recording speed. A multichannel correlation algorithm is used to enhance the performance of the system in the presence of muscle artifact (EMG). Results are presented showing that the multichannel correlation is capable of reducing, in some cases, both missed detections due to poorly defined ST's and false alarms due to EMG.

Abstract: Design is, perhaps, the toughest problem solving activity that engineers perform. Design can be viewed as (Mostow, 1985) constructing an artifact or structure that: 1. satisfies a given functional specification, 2. conforms to the limitations of the resources, and 3. satisfies implicit and explicit design criteria on the form of the artifact.

Abstract: A phantom consisting of a human skull fixed in a cylinder containing water was used to study the occurrence of artifacts and their influence on image quality in CT of the posterior fossa. Only minor differences in interpetrous bone artifact and in deviation in CT numbers in other parts of the posterior fossa were found between eight tested scanner models. Interpetrous bone artifact was not significantly affected by slice thickness but it could be reduced to some extent by proper selection of the scan angle and almost eliminated by application of a dual-energy technique. Other bone-related artifacts, however, were found to be dependent on slice thickness.

Abstract: A circuit is described for automatic on-line real-time elimination of electrode-pop artifact from EEG recordings, which can occur occasionally from a sudden change in the dc contact potential at an electrode. The circuit, which functions irrespective of polarity and amplitude (over a wide range) of the artifact, is intended primarily for eliminating this artifact from EEG's for computer analysis, from which spurious results can otherwise emerge. Examples of operation of the circuit are included.

Abstract: A mathematical model for a sonogram containing a circular object with a sonic speed less than that of its surroundings was constructed. The circular object acts as a lens to the ultrasound beam, which explains the refractive artifacts often observed. An unexpected artifact is also predicted, a region of increased echogenicity, which the authors named the retrolenticular afterglow, since it occurs behind a lens-like structure. An experiment was designed to show the appearance of this artifact occurring distal to an ethyl alcohol-filled cylindrical well in a tissue equivalent phantom.

Abstract: High-intensity reverberation artifacts frequently occur within the fetal cranium during obstetric ultrasound examinations. One hundred ninety-six static ultrasonograms obtained from 100 consecutive obstetric patients were reviewed to determine the incidence, shape, and location of high-intensity reverberation artifacts within the fetal head. Such artifacts occurred in one-third of the fetuses examined. The most common appearance was an arc-shaped artifact that occurred most frequently in the fetal hemicranium farthest from the maternal skin surface. Such an artifact can mimic epidural fluid collections and cause confusion for examiners not familiar with this finding. A discussion of this "pseudoepidural" artifact, its cause, and ways of excluding underlying intracranial pathology is presented.

Abstract: An artifact (2) which is being intentionally covered with a material (10) of low magnetic permeability has a permanently magnetic element (14) atop or integral with the artifact. Subsequent to the covering, the artifact is located by passing a compass (16) or compass-like element over the covering material (10) until the magnetic field of the magnetic element (14) associated with the artifact (2) is detected by movement of the compass (16) or compass-like element caused by said magnetic field. The magnetic element may be integral with a cap (12) or plug put upon or attached to the artifact, or the magnetic element may be placed upon or adhesively affixed to the artifact.

Abstract: Planview area is a useful measurement in lithic artifact analysis. The use of a compensating polar planimeter to obtain this measurement from artifact photocopies is rapid and accurate. The use of planview area in analyzing biface reduction trajectories demonstrates the value of this simple measurement.

Abstract: The phenomenon known as the rare-male advantage has recently become a controversial issue in population biology. On the one hand, negative reports on the presence of rare-male mating advantage in a number of Drosophila strains have questioned the generality of the phenomenon (Anderson and McGuire 1978; Markow 1978; Markow et al. 1980; Pot et al. 1980; Partridge and Gardner 1983). On the other hand, models presented in Kence and Bryant (1978) and Kence (1981) have predicted that a rare-male advantage can arise as an experimental artifact of pooling the results of runs in which marking is rotated between rare and common strains. Since the rotation of marking between strains in replicate runs of experiments on frequency-dependent mating has been a rather common practice to avoid a possible bias caused by marking only one of the strains (see Ehrman et al. 1965; Ehrman 1966; Ehrman and Petit 1968; Spiess 1968; Spiess and Spiess 1969; Tardif and Murnik 1975; Fontdevilla and Mendez 1979; Molin 1979), the possibility that the rare-male advantage could be an artifact of experimental design casts serious doubts on the biological reality of the phenomenon. Recently Knoppien (1984) questioned the validity of predictions derived from the mating-behavior models of Kence (1981) and Kence and Bryant (1978) as they pertain to frequency-dependent mating. Here, I reply to those criticisms. First, I show that his criticism of the algebraic model of Kence (1981) is not valid. Second, I present simulation results to answer his criticism with respect to the computersimulation model of Kence and Bryant (1978). When marking is rotated between the strains in successive runs of a mating experiment, the overall mating frequency (x*) of the rare strain can be computed by averaging x' and x", the mating frequencies of rare males in rare-marked and common-marked runs, respectively. The overall mating frequency for rare males is thus larger than expected if marking reduces the mating success of males (Kence 1981). In criticizing Kence (1981), Knoppien (1984) argued that x* should be computed from the formula x* = NR/(NR + Nc), where NR and Nc are the total numbers of matings over all experimental runs by rare and common males, respectively. He then attempted to estimate mating frequencies using the total number of matings over all runs (N = NR + Nc). For example, the number of matings by the rare strain when it is marked was computed by Knoppien as NR' = Nx(1 s), where x is the frequency of the rare strain and s is the reduction in the