Introduction:

Scientometrics analyzes scientific publications through bibliometric and computational techniques, whereby productivity and impact indicators are generated.

Objective:

To propose a multidimensional methodology in order to obtain the scientometric profile of the National Cancer Institute (INCan), Mexico, and rank it with regard to other national health institutions.

Method:

Using the LabSOM software and the ViBlioSOM methodology based on artificial neural networks, the INCan scientific production indexed in the Web of Science from 2007 to 2017 was analyzed. The multidimensional scientometric profile of the Institute was obtained and compared with that of other national health institutions.

Results:

In terms of productivity, INCan ranks fourth among the 10 Mexican public health institutions indexed in the Web of Science; in the normalized impact ranking, it ranks sixth. Although out of 1323 articles 683 (51.62 %) did not receive citations, 11 articles classified as excellent (0.83 %) obtained 24 % of 11,932 citations and, consequently, INCan normalized impact rate showed a mean productivity higher than the world mean.

Conclusion:

Multidimensional analysis with the proposed neural network enables obtaining a more reliable and comprehensive absolute and relative institutional scientiometric profile than that derived from measuring isolated variables.