Validation of threshold method for myocardial control database by use of clinical data.

A database is an important factor in the statistical analysis of myocardial scintigraphy. Our aim in this study was to verify the validity of the threshold method using phantoms and to create a clinical database using this method. Since this method involves artificially excluding a low count area on a polar map, we created a myocardial phantom with defects. Then, we applied this method to the construction of a control database (CDB) for which we used stress-rest scans of 152 male and 52 female Japanese patients. The clinical relevance of this database was investigated by comparison of the values between the CDB and a Japanese normal database. In the study evaluation, we mainly used the summed extent score (SES) and a severity map (severity). Data from the phantom with defects demonstrated that the threshold method could compensate for defective areas, enabling the use of data for the creation of the CDB. Comparison of the CDB with the Japanese normal database showed a good relationship with respect to the SES and severity (Initial post-stress: SES: r = 0.978; severity: r = 0.997, Redistribution: SES: r = 0.944; severity: r = 0.993). The threshold method facilitates the effective creation of a database by use of clinical data. This enables individual institutions to build their own databases, taking into account differences in collection and processing conditions between institutions as well as the characteristics of individual equipment.