Construction of the knowledge file for an image understanding system.

To enable an image understanding system to provide an automated interpretation of diagnostic imagery it must have access to all of the concepts, procedures and methods used by human experts. The paper describes information elicitation from experts of different domains and the construction of a knowledge file. Uncertainty management is based on Bayesian belief network methods.     

Diffuse peripheral odontogenic fibroma: report of 3 cases

Since peripheral odontogenic fibroma (POF) is characteristically described as a solitary lesion and no diffuse POF had been reported in the literature, our cases should be considered as extremely unusual. Three diffuse cases of POF are described of which one case was seen in association with ocular and skin lesions. The question arises whether POF should be considered as a true odontogenic tumor rather than a diffuse hamartomatous lesion caused by uncontrolled induction of the gingiva. It is also possible that such lesions could be part of a yet undescribed syndrome.     

Treatment and results in severe craniocerebral trauma

Treatment of severe craniocerebral traumas in medium and major hospitals not provided with special neurosurgical equipment requires particularly close interdisciplinary cooperation between surgeons, neurologists and anaesthesiologists. CT facilities are an essential prerequisite for best possible patient care. Patient safety during the posttraumatic and postoperative phases is improved by measuring the intracranial pressure. Whereas corticosteroid treatment may be arguable, barbiturate treatment should presently not be a routine procedure because of its side effects and high rate of complications.     

Metabolic, molecular genetic and toxicological aspects of the acetylation polymorphism in inbred mice.

Over the past 10 years, much fascinating information has been obtained concerning the biochemistry, genetics, toxicological implications and molecular genetics of the N-acetylation polymorphism in mice. Using C57BL/6J (B6) mice as representative of rapid acetylation and A/J (A) mice as representing slow acetylation, it has been shown that the polymorphism observed in N-acetyltransferase (NAT) activity in liver also occurs in kidney, bladder, blood, and other tissues. The development of congenic acetylator mouse lines derived from B6 and A, have provided the necessary tools to study the role of the acetylation polymorphism, on either the B6 or A genetic background, free of nearly all other genetic differences between these strains. Eliminating genes which modify and complicate the differences due to the acetylator genes make the congenic lines very useful in toxicology studies, particularly those involving carcinogenesis. The molecular genetic basis of the acetylator polymorphism in B6 and A mice involves two Nat genes. Nat-1 encodes a protein termed NAT1 which is identical in rapid and slow acetylator strains. Nat-2, however, differs between rapid and slow strains by a single nucleotide change in the coding region. The corresponding NAT2 proteins differ by a single change at amino acid 99: an hydrophilic asparagine in rapid acetylator NAT2 to an hydrophobic isoleucine in NAT2 from slow acetylators. The mechanistic basis for the differences between rapid and slow acetylation in mice appears to be that NAT2 from the rapid B6 strain is 15-fold more stable at 37 degrees C and is transcribed/translated with a maximal efficiency twice that of the enzyme from slow acetylator A mice. Results discussed in this review indicate that mice provide an excellent system for studying the N-acetyltransferase polymorphism and also are useful for modelling several aspects of the human N-acetyltransferase polymorphism.     

Comparison of questionnaire screening and urine toxicology for detection of pregnancy complicated by substance use.

OBJECTIVES: To compare the efficacy of structured questionnaire screening and prenatal urine toxicology for the detection of substance use by pregnant women, and to describe substance use patterns in a group of women presenting to a university-based obstetric clinic. METHODS: All patients presenting to our obstetric clinic for their first prenatal visit were evaluated for evidence of current use of alcohol or any illicit substances. Nursing personnel administered an extensive questionnaire, which included detailed questions about past and current substance use patterns. Urine samples were examined by a commercial laboratory for alcohol and a number of illicit substances. Current users were defined as those who admitted use within the past 30 days or who had positive urine toxicology. RESULTS: Fifty-eight of 302 patients (19.2%) were identified as current users of alcohol or illicit substances. Only 17 of 41 women (41.5%) with toxicologic evidence of recent use admitted to current use. Only 17 of 34 (50.0%) admitting to current use had toxicology positive for any substance. Patients with a history of multiple past substance use were significantly more likely to have positive urine toxicology than those without such a history (26.1 versus 7.4%; P < .005). Among current users, multiple substance use (34.5%) and cigarette smoking (52%) were common. CONCLUSION: A screening combination of structured questionnaire and universal urine toxicology identifies more current users than either technique alone, and neither is clearly superior to the other. A history of multiple substance use may be an important indicator of current use.