Reorganizing Data from Transverse Section Scans of the Brain Using Digital Processing

In section scanning, counts from a radioactive object are collected from many different views and then represented as a rearrangement that causes correspondence of images on a depth-specific plane (1–5). In 1963 we introduced transverse section scanning as a means of portraying the radioactivity in body organs as a crosssection picture (Fig. 1). Two radiation detectors make a sequence of tangential scans at regular angular intervals around the patient so as to view any radioactive structure in the selected cross-section from many different directions. In earlier reports (1–4) we used film exposure summation to reorganize these scan data into a section picture. Originally we made the section picture during the scanning procedure by exposing film with a thin line of light that was moved on the screen of a cathode ray tube (CRT) with speed and orientation matched to the detectors' line of view (Fig. 2). The detectors made tangential scans at 24 different angles (every 7.5°) and 24 separate exposures of the film produced a section picture. With this method, however, we were not always able to prejudge line brightness correctly so that the final picture would have proper contrast. We abandoned making the picture during the scan and instead recorded the data on perforated tape during the study, then used the tape to control the beam of a CRT for picture generation later. We were successful in applying this combined technic of tape recording and film exposure summation to clinical section scanning. For example, we used it to learn that transverse section brain scanning gives additional information about brain tumor distribution and improves detection accuracy for lesions situated at the base of the cranial vault where ordinarily their images can be obscured (4). We suspected, however, that film exposure summation, with its inherent problems of nonlinear addition, might not be the best possible way to make section pictures (5). This paper is a report of our subsequent experience with an alternative method of making transverse section pictures—digital summation. Methods Recording the Scan Data Two opposed scintillation detectors (A and B) scan over paths tangential to the head, rotate 7.5°, rescan along a second pair of tangent paths, rotate another 7.5°, and so on to complete a full revolution and collect data from a total of 24 pairs of scan paths (Fig. 2). Counts detected during each 0.25-cm increment of detector displacement are recorded in 6-bit binary code on standard 8-column perforated paper tape, along with codes specifying the beginning of the scan sequence (RS), the start of each tangent scan (LS), and the end of the scan sequence (RE). An additional code distinguishes data characters of detector A from those of detector B. The counts from each detector (A or B) may be recorded separately on the tape. Alternatively, the counts from the two detectors may be added together during scanning, and the sum (A + B) may be recorded.