ImaChek – Why Image Analysis Matters

Proper image processing and Utilization of the Latest Tools

By iGroup Japan

Introduction

The Alarming Rise of Retracted Papers and Image Problems

The number of retracted academic papers is increasing. Strikingly, a small number of authors are responsible for many retractions, with Japanese authors accounting for five of the top ten globally. Fields like biology, chemistry, and medicine, which rely heavily on image data, are most affected. Image-related issues are a top cause of retractions, taking the longest to discover and rectify, leading to flawed research being cited for years and eroding scientific credibility.

The True Cost of Image Misconduct

Retracted papers cause more than financial loss; they damage reputations and erode trust within the scientific community. The longer misconduct goes undetected, the greater the harm to scientific reliability.

The Crucial Need for Proper Image Processing Education

A primary cause of image misconduct is the lack of training in image processing and analysis. Unlike some European institutions where specialists handle images, Japanese researchers often do it themselves, increasing errors. The “Guide to Proper Image Processing Methods and Journal Submission Guidelines (2nd Edition)” aims to counter this by promoting ethical practices, preventing fraud, and ensuring research transparency. Key takeaways from the guide include:

• Documenting image processing methods and preserving original data.
• Applying proper image adjustments (trimming, scaling, contrast) without distorting information.
• Creating gel images accurately.

The guide also lists unacceptable practices: arbitrary cutting/pasting/deletion, S-curve contrast manipulation, changing aspect ratios, “decorative” processing, and inconsistent processing across comparative images. A common, unintentional error is using “Auto” contrast adjustment for multiple images, which can lead to misleading comparisons. Manual adjustment is recommended for comparative studies.

AI and Image Integrity

While AI tools can create diagrams, they must not be used to generate artificial research data. Users must document AI tools used and be mindful of copyright infringement. The rise of AI in image processing necessitates clearer guidelines for its ethical use.

ImaChek: An AI-Powered Solution

ImaChek, a new AI tool, automatically detects image misconduct. The upcoming version 2 offers improved accuracy and speed. It identifies duplication (reused images) and manipulation (altered images). ImaChek can analyze files internally or compare them against a large repository of over 2 million (soon 10 million) open-access PubMed papers. It supports various image types and offers a user-friendly interface with detailed analysis, note-taking, and report generation.

The Path Forward: Education and Tools