Related Experiment Video
Updated: Aug 17, 2026
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
[Automatization of a hospital-based tumor registry]
Josepa Ribes1, Jordi Gálvez, Angels Melià
1Servei d'Epidemiologia i Registre del Càncer, Institut Català d'Oncologia, 08907 L'Hospitalet de Llobregat, Barcelona, España. j.ribes@iconcologia.net
Introduction:
To increase data reliability and reduce the costs associated with the HTR, the Catalan Institute of Oncology programmed the manual procedures of data collection from databases by means of a computer application (ASEDAT).
Material And Method:
ASEDAT detects the incident tumors of the registry from the databases of the pathology records (PR) and discharge records (DR) and selects the basic information from both databases. Data from the HTR data was collected for the period 1999-2000 by means of 2 procedures: manual and automatized collection and the results obtained were compared.
Results:
10,498 cancer patients were detected. Manual resolution detected 8,309 incident tumors and 2,374 prevalent tumors. ASEDAT automatically detected 8,901 patients (84.8%), in whom 8,367 incident tumors were detected (58 more tumors than the manual procedure). Validation of agreement was performed in the incident tumors detected by both methods (7,063 tumors). In 6,185 tumors (87.6%) the information agreed in all the variables. Of the discordant tumors, 692 (9.8%) were obtained by the RHT staff using manual resolution, and the remainder (186; 2.6%) were obtained by the application (automatic resolution).
Conclusions:
Cancer registry automatization is feasible when PR and DR databases are available, coded and automatized.
