18 Jun 2025
Liu Yiheng, a final-year student at the Department of Biosciences and Bioinformatics, School of Science, has developed an innovative deep learning model to classify breast cancer cells using Raman spectroscopy. Published as first author in Computational and Structural Biotechnology Journal, his work could pave the way for faster, more accurate cancer diagnosis.
AI Diagnostic Tool
Raman spectroscopy, as a non-destructive molecular "fingerprinting" technique, has demonstrated significant potential for early cancer diagnosis. However, effectively analyzing complex spectral data remains a major challenge in the field. Under the guidance of Dr. Xia Huang, final-year student Yiheng Liu innovatively proposed a Random Splicing Convolutional Neural Network (RS-CNN) model for machine learning-based analysis of tumor Raman spectra. The model achieved an impressive classification accuracy of 98.63% across six breast cancer cell lines, offering a novel technological approach for clinical cancer diagnosis.The research team notes that the RS-CNN model is not only applicable to breast cancer diagnosis but can also be extended to rapid identification of other cancer types, as well as pathogens like bacteria and fungi. With the increasing adoption of Raman spectroscopy in clinical settings, this AI-driven analytical method promises to provide robust support for precision medicine and personalized treatment.
The Undergraduate Research Journey: From Beginner to Independent Researcher
This research experience has given Yiheng Liu a clearer vision for his future. "Although my academic record doesn't qualify me for direct Ph.D. admission, I plan to pursue a master's degree at XJTLU and continue my research with Dr. Huang," he shared. "Scientific research is truly fascinating—it helped me discover my passion."
From an undergraduate with ordinary grades to the first author of an internationally published paper, Liu's transformation is remarkable. "The greatest takeaway isn’t just the published paper, but finding a career path I truly love. I’ve learned that even if your grades aren’t perfect, passion and hard work can still lead to meaningful achievements in research." Yiheng said.
Liu’s journey would not have been possible without the strong support for undergraduate research from Prof. John Moraros, Dean of the School of Science, as well as XJTLU’s distinctive philosophy in nurturing young researchers. The SURF program offers invaluable opportunities for students like Liu—those with a passion for research but not necessarily top-tier grades—to engage in cutting-edge projects under faculty mentorship.
"Undergraduates are fully capable of producing high-quality research," Dr. Huang emphasized. "The key is giving them opportunities, trust, and proper guidance. We look forward to more students like Yiheng discovering their potential through SURF, and we warmly welcome curious, self-driven undergraduates to join our team."
Content from Dr Xia Huang
Review：Professor John Moraros
18 Jun 2025