Reconstructing Super-Resolution Raman Spectral Image Using a Generative Adversarial Network-Based Algorithm

• Published in: Analytical chemistry (Washington) Vol. 97; no. 31; pp. 17121 - 17131
• Main Authors: Xu, Jie, An, Haorui, Kong, Xiangtao, Zhang, Zixuan, Liu, Qidong, Li, Jie, Qin, Jie, Bratchenko, Ivan A., Wang, Shuang
• Format: Journal Article
• Language: English
• Published: United States American Chemical Society 12.08.2025
• Subjects: Algorithms; Biochemical analysis; Chemical fingerprinting; Cluster analysis; Clustering; Data acquisition; Deep learning; Generative Adversarial Networks; Image acquisition; Image Processing, Computer-Assisted - methods; Image reconstruction; Image resolution; Information processing; Machine learning; Neural Networks, Computer; Performance evaluation; Real time; Root-mean-square errors; Signal to noise ratio; Spatial discrimination; Spatial distribution; Spatial resolution; Spectrum Analysis, Raman - methods; Transfer learning; Vector quantization
• ISSN: 0003-2700; 1520-6882; 1520-6882
• DOI: 10.1021/acs.analchem.5c02934

Raman imaging utilizes molecular fingerprint information to visualize the spatial distribution of a substance within the scanned area. Subject to its scanning mechanism, it usually costs a prolonged data acquisition duration for achieving high-resolution Raman images. In this study, we propose a generative adversarial network (GANs) based algorithm to significantly enhance both the Raman spectral imaging speed and spatial resolution. The proposed method was trained and evaluated on 186 hyperspectral Raman datasets acquired from unlabeled cells, and its reconstruction performance was quantitatively evaluated by the parameters of peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and root-mean-square error (RMSE). Univariate imaging and K-means clustering analysis (KCA) were both adopted to evaluate the preservation of biochemical information after image reconstructing. The results demonstrated that the proposed method effectively enhances spatial resolution by a factor of 2–4 while accelerating imaging speed by a factor of 4–16. Furthermore, transfer learning was utilized to adapt the pretrained model to different objects, validating its generalization capabilities and extending its universalities. This study highlighted the potential of deep learning for super-resolution Raman imaging, providing a promising pathway for high-throughput and real-time biochemical analysis.