Ripening Study Based on Multi-Structural Inversion of Cherry Tomato qMRI.

Abstract

This study introduces a non-destructive, quantitative method using low-field MRI to assess moisture mobility and content distribution in cherry tomatoes. This study developed an advanced 3D non-local mean denoising model to enhance tissue feature analysis and applied an optimized TransUNet model for structural segmentation, obtaining multi-echo data from six tissue types. The structural T2 relaxation inversion was refined by integrating an ACS-CIPSO algorithm. This approach addresses the challenge of low signal-to-noise ratios in multi-echo MRI images from low-field equipment by introducing an innovative solution that effectively reduces voxel noise while retaining structural relaxation variability. The study reveals that there are consistent patterns in the changes in moisture mobility and content across different structures of cherry tomatoes during their ripening process. Mono-exponential analysis reveals the patterns of changes in moisture mobility (T2) and content (A) across various structures. Furthermore, tri-exponential analysis elucidates the patterns of changes in bound water (T21), semi-bound water (T22), and free water (T23), along with their respective contents. These insights offer a novel perspective on the changes in moisture mobility throughout the ripening process of tomato fruit, thereby providing a research pathway for the precise assessment of moisture status and ripening expression in fruits.

Plant phenotyping relevance

低磁場MRIのノイズ低減、構造セグメンテーション、T2緩和反転を開発し、トマト果実の水分状態と成熟表現型を定量化する方法が研究の中心である。

Code and data availability

The article describes cherry tomato qMRI measurements, segmentation, and T2 inversion analysis, but contains no public phenotype dataset, image deposit, author code repository, or trained model release. The Data Availability Statement only offers contact with the corresponding authors, which does not constitute a paper

No evidence-backed public reproduction asset is currently recorded.