X-ray technologists' reproducibility from automated measurements of the medial tibiofemoral joint space width in knee osteoarthritis for a multicenter, multinational clinical trial.

OBJECTIVE: To determine the reproducibility of x-ray technologists, 26 in North America (NA), 24 in Europe (EU), in reliably repositioning patient's osteoarthritic (OA) knees, from computerized measurements of minimum joint space width (JSW) and reproducibility in joint repositioning, during their training for the clinical trial. METHODS: Technologists from 12 NA and 12 EU clinical radiology units received identical training, at one site on each continent, in performing the fluoroscopically assisted semiflexed knee examination and in quality control criteria (QCC) for film acceptance. Subjects recruited were 129 in NA and 70 in EU, with both knees radiographed for some subjects. Each technologist radiographed 5 OA knees and repeated the process on the same knees 2 to 7 days later. Minimum medial JSW was measured at a single center on digitized images with computer software that corrected for radiographic magnification. Technologists' reproducibility in joint repositioning and JSW measurement was determined from the difference between test and retest. RESULTS: In all, only 3/50 technologists failed qualification criteria with a repeat-film JSW difference > 0.50 mm. The mean, standard deviation (SD) of the difference in JSW between test/retest for 146 NA film-pairs of -0.020 (0.16) mm was not statistically different from that in 120 EU film-pairs: -0.001 (0.18) mm. In NA and EU 45% of examinations achieved high quality, i.e., JSW difference between repeat films < 0.1 mm, and 92% achieved excellent to good quality with a difference between repeat films < 0.3 mm. NA and EU technologists' reproducibility was unaffected by subject's sex, age, and degree of JSW loss. Reproducibility in joint reposition for all technologists was excellent. CONCLUSION: Between-continent precision of JSW measurements from all accepted pairs of semiflexed views was excellent to very good and similar to the high technical quality achieved in the authors' original report. The value of training incorporating both test/retest radiographs and film QCC is essential for the high technical quality required for multinational clinical trials.

Polyclonal Antibody-Based ELISA and HPLC Methods for the Determination of Fumonisins in Corn: A Comparative Study.

The performance of an experimental polyclonal antibody (PAb)-based competitive direct enzyme-linked immunosorbent assay (CD-ELISA) developed for the analysis of fumonisins in corn was assessed by comparison with an established high-performance liquid chromatography (HPLC) method. The comparative study was conducted using a series of 20 corn samples naturally contaminated with combined fumonisin levels ranging from <0.05 to >5 µg/g (ppm). Linear regression analysis between the results generated by HPLC and CD-ELISA provided correlation coefficients (r) and regression slopes (b) of r = 0.960, b = 1.493 (P < 0.001); r = 0.865, b = 3.903 (P < 0.001); and r = 0.832, b = 0.107 (P < 0.001) for the individual fumonisins B1 (FB1), B2 (FB2) and B3 (FB3), respectively, while corresponding values of r = 0.967, b = 1.059 (P < 0.001) were obtained for the combined FB1, FB2, and FB3 concentrations. In 3 of 18 fumonisin-positive corn samples, combined fumonisin levels determined by CD-ELISA were between 85 and 100% higher than those determined in the same extracts by HPLC, while in 13 other samples, CD-ELISA results were between 1.8 and 53% higher than those determined by HPLC. Conversely, in 2 of 18 samples, CD-ELISA results were lower than those determined by HPLC. The differences recorded between HPLC and the experimental PAb-based CD-ELISA were far less than those previously recorded for other mono- and polyclonal antibody-based CD-ELISA systems. The results indicate that the experimental PAb-based CD-ELISA may be effectively applied for the initial screening for fumonisins in corn.

Investigations of Syntactic Ambiguity.

Two experiments were conducted with male and female university students in an attempt to establish whether syntactically ambiguous words are better represented by independent lexical entries for each syntactic context in which the items occur, or whether the different usages could be adequately represented by single entries. Both experiments were attempts to influence recall through control of the coding of ambiguous items. Experiment one (N = 128) involved a free recall procedure in which ambiguous items were embedded in lists of nonambiguous nouns or verbs, and Experiment two (N = 144) involved a release from proactive inhibition procedure which has been shown to make syntactic features highly salient. The results of both experiments supported the single entry view; that is, the ambiguous items apparently were not coded according to one usage independent of other usages.