PubMed Health⌕ Search

Biomedical subjects

J R Hallberg

Publications and source records attributed to J R Hallberg.

3 recordsLinked to original sources
PubMedDate not supplied

Perception of simulated lung lesions.

Two series of test radiographs were produced, each containing a single nodular lesion. Each series was identical except that the lesions were spatially shifted with respect to adjacent structures, resulting in matched pairs of chest radiographs simulating a stereoscopic examination. Observation of each set singly and in pairs (non-stereoscopically) demonstrated a marked improvement in observer performance when the shifted pairs were viewed side-by-side.

Diagnostic Errors↗
PubMedDate not supplied

Some effects of method on the measured conspicuity of chest lesions.

Both the magnitude and the precision of the conspicuity of a series of solitary chest nodules depends on the number and spacing of the measurement points. Observer accuracy in detecting 1.0 cm and 1.5 cm psuedo chest lesions does not correlate well with conspicuity.

Diagnosis, Differential↗
PubMedDate not supplied

Computational analysis and dosimetric evaluation of a commercial irregular-fields computer program.

The proper evaluation of the accuracy of a new computer program for radiation-therapy dosimetry requires consideration of both the mathematical algorithm used in the program and the performance required in clinical applications. As an example, our evaluation of the irregular-fields dosimetry program currently marketed by SHM for their Rad-8 system is described. The evaluation begins with an explanation of the mathematical computation described. The evaluation begins with an explanation of the mathematical computation method, with emphasis on the points where the calculation differs from previous methods. Next, the procedure for setting up the beam data file is discussed. Finally, a step-by-step procedure is described in which calculated doses are compared with measured doses, using a Varian Clinac-4 with lead flattening filter, and the limits within which a +/- 5% accuracy is attainable are estimated. Some sources of error and areas for possible improvements are mentioned.

Computers↗