M Maass1, M Kosonen, M Kormano. 1. Department of Diagnostic Radiology, University of Turku and Turku University Central Hospital, Turku, Finland.
Abstract
PURPOSE: To identify costs deriving from data migration of obsolete digital archives by measuring the workload, and to analyse migration-associated problems. MATERIAL AND METHODS: Two digital archives were used (DTL and MOD) and the capacity of these archives could no longer support the needs of the Medical Imaging Centre. The entire content of the DLT archive and selected data from the MOD archive were transferred to the current higher capacity (17 TB) tape archive. The running time of work processes was measured by self-reporting, and the cost of work was calculated. RESULTS: The transfer of 43,096 studies required 314 working hours over the course of 15 months in total. The work was partly manual, partly automatic. The percentage of non-retrievable MOD images was 35. Less than 0.2% of the DLT image transfers failed due to incorrect patient or image data. The MOD - DLT transfer cost was six times higher per study than the DLT - DLT transmission cost. CONCLUSION: At present, data migration may be inevitable as the amount of data increases and technology advances. The data transfer proved to be labour intensive, with high fault sensitivity regarding the MOD archive. The cost of work of data migration was 0.4% of estimated digital archiving total yearly cost. Automated data migration is preferable.
PURPOSE: To identify costs deriving from data migration of obsolete digital archives by measuring the workload, and to analyse migration-associated problems. MATERIAL AND METHODS: Two digital archives were used (DTL and MOD) and the capacity of these archives could no longer support the needs of the Medical Imaging Centre. The entire content of the DLT archive and selected data from the MOD archive were transferred to the current higher capacity (17 TB) tape archive. The running time of work processes was measured by self-reporting, and the cost of work was calculated. RESULTS: The transfer of 43,096 studies required 314 working hours over the course of 15 months in total. The work was partly manual, partly automatic. The percentage of non-retrievable MOD images was 35. Less than 0.2% of the DLT image transfers failed due to incorrect patient or image data. The MOD - DLT transfer cost was six times higher per study than the DLT - DLT transmission cost. CONCLUSION: At present, data migration may be inevitable as the amount of data increases and technology advances. The data transfer proved to be labour intensive, with high fault sensitivity regarding the MOD archive. The cost of work of data migration was 0.4% of estimated digital archiving total yearly cost. Automated data migration is preferable.
Authors: Peter M A van Ooijen; Kadek Yota Aryanto; André Broekema; Steven Horii Journal: Int J Comput Assist Radiol Surg Date: 2014-10-28 Impact factor: 2.924