Anthony Chung-Bin

Accuracy in patient setup and its consequence in dosimetry

Analysis of 5575 settings on a computer-monitered Theratron-80 60Co unit demonstrates that human error does occur in treating patients with radiation. The errors are due to inaccurate setting of such parameters as field size, gantry angle, collimator rotation, treatment time, etc. The error rate per parameter was found to be about 3%, and more than two-thirds of the patients monitored with the PDP 11/45 computer had at least one error at some stage during the full course of treatment. Both the dose and the dose distribution may be affected by these errors and have been studied in a few typical cases. The errors in timer setting have the largest effect on the prescribed dose and may change the probability of local control appreciably.

Accuracy in radiotherapy treatment

Radiotherapy has proved very effective in the management of cancer. There is increasing evidence that the dose-response curve is rather steep. Therefore, accurate doses are important clinically. At present there is not enough data to support that this degree of accuracy is achieved in clinical radiotherapy all the time. Our data on patient monitoring and verification suggest that timer or monitor setting has the largest effect on the dose and may be significant.

Can the AAPM task group 21 protocol lead to optimum ion chamber designs

The recently published AAPM Task Group 21 protocol for high-energy dosimetry is complicated in that it requires the physicist to obtain the values of about a dozen different physical variables by looking them up in tables or graphs. This should be compared with the procedure of earlier protocols using the concept of a single multiplier C lambda. We have investigated how the physical principles outlined in the improved AAPM protocol could be utilized for the redesign of the therapy-level ion chambers in such a way that one can reduce the number of factors that need to be looked up in tables or graphs for the calibration of high-energy teletherapy photon beams. In our analysis presented in this paper we found that one such design could be for an ion chamber having a wall acrylic or Bakelite of a thickness not exceeding 0.1 g/cm2 and having an inner diameter of 6 mm, and used in conjunction with a cobalt-60 buildup cap of thickness 0.35 g/cm2 made of acrylic, Bakelite, or Tufnol. If a chamber of such a design is used in a water phantom, the dosimetry practically reduces to the simplicity of the former protocols of depending on a single value of energy-dependent multiplier to be obtained from a table. With the above design parameters, it becomes possible to eliminate the explicit need to incorporate the factors Pwall, Prepl, Awall, beta wall, and the variable alpha, representing the fraction of ionization due to electrons from the wall material of the chamber.

Information System for Research and Operation in a University Radiology Department

Several relational database management systems have been implemented in Fortran/Assembler on a Digital PDP 11/45 computer to store the patient data of diagnostic radiology Nuclear Medicine, and Therapeutic Radiology at Rush-Presbyterian-St. Luke’s Medical Center. Each system involves its own on-line conversational input package and all systems utilize the same easy retrieval package. The former consists of various application programs for entering and updating the information. The latter is a general purpose MIRS (Medical Information Retrieval System) interpreter which executes the retrieving program to scan the data of interest and generates the summary listing. Data stored for each patient contains a set of general information and various sets of specific information for different kinds of studies which may only apply to selected group of patients. Data of each field within a set is defined by the type, length, and its offset to the beginning. Thus, new studies can easily be added to the system without moditying the existing programs. MIRS, a translator completely written in Assembler, uses the definitions in dictionary file to look up and convert the data that corresponds to the variable names in retrieving program, performs the relevant boolean operations, deciding whether criteria matched, and printing the results. Such system has been proven to be nighly satisfactory for its easy maintaining and fast retrieval.

Interactive Computer Program for the Selection of Optimum Pulse Sequences in NMR Imaging

Application of Nuclear Magnetic Resonance (NMR) in Medical Imaging is a relatively new approach to the diagnosis of tissue pathology. By analyzing contrast between signal intensities of two tissues, one can optimize the imaging technique. An interactive program has been implemented on a PDP 11/45 computer, manufactured by Digital Equipment Corporation, that allows for computation and plotting of signal intensities of various pulse sequences for tissues with different NMR parameters (T1, T2, Proton Density). The options for display of tissue contrast can be: (1) Contrast = Signal Intensity I-Signal Intensity II (2) Contrast = (Signal Intensity I-Signal Intensity II)/SQRT of repitition time (3) Contrast = (Signal Intensity I-Signal Intensity 2)/(Signal Intensity I + Signal Intensity 2)