“Considerable overlap” leads to retraction of medical imaging paper

We have poked fun at Pattern Recognition Letters before for failing to catch blatant plagiarism. We probably should have held off on those jokes for this post.

A group of IT researchers from India has suffered the retraction of a paper in PRL for heavily basing the piece on at least four previous papers written by one of the co-authors without proper attribution (not that such attribution likely would have absolved the sin).

The paper, titled “A robust kernelized intuitionistic fuzzy c-means clustering algorithm in segmentation of noisy medical images,” was published in January of this year by Prabhjot Kaur and colleagues.

Here’s the retraction notice:

Seeing double in Pattern Recognition Letters leads to retraction

You’d think this sort of thing would be, well, obvious to the editors of a journal called Pattern Recognition Letters — could a fox get away with publishing in Henhouse News? — but a group of Lithuanian researchers managed to get a duplicate article into the pages of PRL.

The paper, titled “Application of Bayes linear discriminant functions in image classification,” appeared in the February 2011 issue of PRL. But a very similar version already had been published in a special meeting issue of another journal, Procedia Environmental Sciences. Both are Elsevier titles.

According to the retraction notice: