Automatic detection of image manipulations in the biomedical literature

Enrico M Bucci

doi:10.1038/s41419-018-0430-3

Enrico M Bucci

Open Access

https://doi.org/10.1038/s41419-018-0430-3

Copy DOI

Abstract

Images in scientific papers are used to support the experimental description and the discussion of the findings since several centuries. In the field of biomedical sciences, in particular, the use of images to depict laboratory results is widely diffused, at such a level that one would not err in saying that there is barely any experimental paper devoid of images to document the attained results. With the advent of software for digital image manipulation, however, even photographic reproductions of experimental results may be easily altered by researchers, leading to an increasingly high rate of scientific papers containing unreliable images. In this paper I introduce a software pipeline to detect some of the most diffuse misbehaviours, running two independent tests on a random set of papers and on the full publishing record of a single journal. The results obtained by these two tests support the feasibility of the software approach and imply an alarming level of image manipulation in the published record.

Highlights

In a set of drawings dating 13th March 1610 published on the “Sidereus Nuncius”, Galileo represented the uneven curve of the sun’s light over the moon disc, as seen only in January of the same year using his telescope[1]
While this can be acceptable in principle—for example, intensity calibration of a digital image can be required for a quantitative analysis—it is true that image manipulation aiming to deceive the readers of a scientific paper became extremely easy
We considered the open source papers released by PubMed Central[5] (PMC) in January 2014

Summary

Introduction

In a set of drawings dating 13th March 1610 published on the “Sidereus Nuncius”, Galileo represented the uneven curve of the sun’s light over the moon disc, as seen only in January of the same year using his telescope[1]. Given the complexity of the subjects to be represented, in the field of life sciences only few scientists with excellent drawing skills (or having access to gifted artists) could successfully and universally propagate their findings using images— think for example of Haeckel’s embryos or of Darwin’s orchids It was only after “objective” photographic reproduction of experimental outcomes was routinely available, that using images to represent the outcome of a biological experiment became a method accessible to anyone; a method perceived to be as objective as any other experimental set-up, so that in many cases images produced by dedicated apparatuses became the results to be analyzed, qualitatively and quantitatively, to prove a given hypothesis. Official journal of the Cell Death Differentiation Association

Methods

Findings

Conclusion