DNA-CTMF: Reconstruct high quality image from lossy DNA storage via Pixel-Base codebook and median filter.

Area of Science:

• Biotechnology
• Bioinformatics
• Image Processing

Background:

• DNA data storage offers high density but is susceptible to base errors.
• Existing error correction methods can introduce redundancy or amplify errors, degrading image quality.
• Reconstructing high-fidelity images from error-prone DNA sequences remains a significant challenge.

Purpose of the Study:

• To develop a novel method, DNA-CTMF, for reconstructing high-quality images from DNA sequences with substantial base errors and indels.
• To address the limitations of traditional error correction codes by employing image processing techniques.
• To provide an interdisciplinary solution for reliable image storage in DNA.

Main Methods:

• Utilizing a Pixel-Base codebook and a chaotic system to ensure DNA sequences adhere to biological constraints.
• Implementing a codebook to adjust indel-affected base groups to their original positions.
• Applying a median filter to eliminate salt-and-pepper noise originating from base errors.

Main Results:

• DNA-CTMF successfully reconstructs high-quality images with minimal variation across different error compositions.
• Achieved high image quality (PSNR ~23, MS-SSIM > 0.9) even at a 5% error rate with indels comprising 2/3 of errors.
• Demonstrated superiority over other methods across 4000 tested images in simulations and validation through wet experiments.

Conclusions:

• DNA-CTMF effectively reconstructs images from error-prone DNA sequences using image processing, offering a novel approach.
• The method shows robustness against high error rates and indels, outperforming traditional error correction strategies.
• This interdisciplinary technique provides a new perspective for DNA-based image storage solutions.