Radio Frequency Identification Gate System to Identify Misused Personal Dosimeters.

The use of two personal dosimeters, one worn over and one worn under a protective apron, provides the best estimate of effective dose. However, inappropriate positioning of dosimeters is a common occurrence, resulting in abnormally high or low radiation exposure records. Although such incorrect positioning can be identified by radiation exposure records, doing so is time-consuming and labor-intensive for administrators. Therefore, a system that can identify incorrect locations of dosimeters without burdening administrators must be developed. In this study, we developed a radio frequency identification (RFID) gate system that can differentiate between two RFID-tagged dosimeters placed over and under a metal apron and identify misused dosimeters. To simulate the position of the RFID-tagged dosimeters, we designed four dosimeter-wearing classes, including "proper use" and three types of "misuse" (i.e., "reversed," "both under," and "both over"). When the system predicts "misuse" based on the tag reading, the worker is alerted with lights and alarms. The system performance was evaluated using a confusion matrix, with an overall accuracy of 97.75%, demonstrating high classification performance. The safety of the system against life support devices was also investigated, demonstrating that they were not affected by the electric field at 0.3 m or more from the antenna of the system under any transmit powers tested. This RFID gate system is highly capable of identifying incorrectly positioned dosimeters, enabling real-time monitoring of dosimeters to manage their positioning.

[Computerized classification of pneumoconiosis radiographs based on grey level co-occurrence matrices].

Pneumoconiosis is diagnosed as categories 0-4 according to the Pneumoconiosis Law. Physicians have difficulty precisely categorizing many chest images. Therefore, we have developed a computerized method for automatically categorizing pneumoconiosis from chest radiographs. First, we extracted the rib edge regions from lung ROIs. Second, texture features were extracted using a dot enhancement filter, line enhancement filter, and grey level co-occurrence matrix. Third, the rib edge regions were removed from these processed images. Finally, we used a support vector machine for feature analysis. In a consistency test, 56 cases (69.7%) were classified correctly, and 45 cases (61.8%) were classified correctly in a validation test. These results show that the proposed features and removal of the rib edge are effective in classifying the profusion of opacities that indicate pneumoconiosis.