ML in NMR
Spectroscopy_

// project 1 — denoising ¹³C NMR spectra

Monitoring chemical reactions by ¹³C NMR at natural abundance is challenging due to low sensitivity. We developed a GAN-based framework operating in the frequency–time dual domain to denoise and enhance ¹³C NMR spectra, enabling quantitative reaction monitoring without isotope labeling.

// project 2 — diagonal suppression in COSY

Phase-sensitive COSY spectra suffer from strong diagonal artifacts that mask cross-peaks near the diagonal, hiding short-range scalar couplings. We applied machine learning to suppress the diagonal without attenuating the cross-peaks.

// links