The utilization of xeroradiography for radiography of cross-section of thin cadaveric slices.

Although xeroradiography has had many medical applications, its utilization and the advantages it offers in the study of cadaveric sections has not previously been fully explored. It was our purpose to determine possible advantages or disadvantages this modality might have for this purpose. The edge enhancement which xeroradiography allows is a distinct advantage since the various anatomic structures become more clearly defined. On the other hand, the hollow viscera and the lumina of blood vessels are not enhanced, but rather blend with the walls of these "hollow" structures. Also, there is some loss of detail in the depiction of some of the glandular structures, such as the seminal vesicles. An artist's assistance to define these by comparison with the gross specimen gives us an optimum result. An x-ray enhanced image is thereby attained which exceeds the usual line drawing or artist's depiction of the cross-sectional specimen, and both can be similarly labeled for study.

Predictive value of a diagnosis of colonic polyp on the double-contrast barium enema.

The positive predictive value of a diagnosis of colonic polyp on the double-contrast barium enema was determined by analyzing 160 polypoid lesions reported in 105 patients. Of the 160 polyps, 139 were confirmed by endoscopy, surgery, or repeat barium enema for a positive predictive value of 87%. Twenty-one lesions were not confirmed despite repeat endoscopy, repeat barium enema, review of the original study, or surgery, giving a false positive rate of 13%. The individual false positive error rates of 5 radiologists ranged from 8% to 26%. False positive errors were due to stool, diverticula, air bubbles, misinterpretation of the ileocecal valve, and haustration. Many of the false positive errors appeared to have been avoidable if the original examination had been more meticulously interpreted.

The radiologic detection of duodenal ulcers: effects of examiner variability, ulcer size and location, and technique.

Study of 91 endoscopically verified duodenal ulcers compared the effects of examiner variability, ulcer size and location, and radiographic technique on ulcer detection. Radiologic sensitivity of 61.5% was found for the entire group of 91 ulcers. Examiner variability was the strongest determinant of success; sensitivities for individual examiners ranged from 44.4% to 80%. Ulcer size was a second factor in radiologic detectability; ulcers 5 mm or larger were detected at a higher rate (80.0%) than those less than 5 mm (64.5%). Sensitivities of 65.9% and 57.4% were recorded for single- and double-contrast examinations, respectively, a statistically insignificant difference.

Positive predictive value and examiner variability in diagnosing duodenal ulcer.

The positive predictive value of a diagnosis of duodenal ulcer was studied in 255 patients with 274 ulcers diagnosed radiographically. Retrospective analysis of the radiographic examinations and correlation with endoscopy assessed examiner variability, ulcer size, duodenal bulb deformity, and the confidence level of the ulcer diagnoses. For all diagnoses, the overall positive predictive value was 57% with individual predictive values of 47%, 54%, and 70% among three examiners. The individual predictive values correlated directly with the average level of confidence of the diagnoses for each examiner. The sensitivity of the individual examiner for ulcer detection had an inverse relation to the examiner's positive predictive value, confirming that higher sensitivities are achieved at the expense of an increased false-positive rate.