Subtype-Specific Detection in Stage Ia Breast Cancer - Integrating Raman Spectroscopy, Machine Learning, and Liquid Biopsy for Personalised Diagnostics

Aug 3, 2026

This study integrates Raman spectroscopy with machine learning to enable early detection and subtype classification of breast cancer using blood plasma samples. Detailed spectral analyses identify key patterns linked to cancer biomarkers. Results support accurate discrimination of the four major breast cancer subtypes at stage Ia, achieving an average sensitivity of 90% and specificity of 95%, with a cross-validated macro-averaged AUC of 0.98. The work advances non-invasive, personalised diagnostics and monitoring by combining vibrational signals with learning-based models.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Introduction

Liquid biopsy and clinical motivation

Study approach: Raman spectroscopy and machine learning

Results: subtype classification performance

Significance and personalised diagnostics

Edinburgh Research Explorer Subtype-Specific Detection in Stage Ia Breast Cancer: Integrating Raman Spectroscopy, Machine Learning, and Liquid Biopsy for Personalised Diagnostics Citation for published version: Tipatet, K, Hanna, K, Davison-Gates, L, Kerst, M & Downes, A 2025, 'Subtype-Specific Detection in Stage Ia Breast Cancer: Integrating Raman Spectroscopy, Machine Learning, and Liquid Biopsy for Personalised Diagnostics', Journal of biophotonics, vol. 18, no. 1, e202400427 . [https://doi.org/10.1002/jbio.202400427](https://doi.org/10.1002/jbio.202400427) Digital Object Identifier (DOI): 10.1002/jbio.202400427 Link: Link to publication record in Edinburgh Research Explorer Document Version: Publisher's PDF, also known as Version of record Published In: Journal of biophotonics General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please [contact openaccess@ed.ac.uk](contact openaccess@ed.ac.uk) providing details, and we will remove access to the work immediately and investigate your claim. Download date: 29. Nov. 2025 Journal of Biophotonics RESEARCH ARTICLE OPEN ACCESS Subtype-Specific Detection in Stage Ia Breast Cancer: Integrating Raman Spectroscopy, Machine Learning, and Liquid Biopsy for Personalised Diagnostics Kevin Saruni Tipatet1,2 | Katie Hanna3 | Liam Davison-Gates1 | Mario Kerst4 | Andrew Downes1 1Institute for BioEngineering, School of Engineering, University of Edinburgh, Edinburgh, UK | 2Promotionskolleg NRW, Bochum, Germany | 3Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK | 4Faculty of Life Sciences, Rhine-Waal University of Applied Sciences, Kleve, Germany Correspondence: Kevin Saruni Tipatet ([k.saruni.tipatet@ed.ac.uk](k.saruni.tipatet@ed.ac.uk)) Received: 24 September 2024 | Revised: 5 November 2024 | Accepted: 8 November 2024 Keywords: breast cancer subtyping | early detection | liquid biopsy | machine learning | personalised cancer care | Raman spectroscopy ABSTRACT This study explores the integration of Raman spectroscopy (RS) with machine learning for the early detection and subtyping of breast cancer using blood plasma samples. We performed detailed spectral analyses, identifying significant spectral patterns associated with cancer biomarkers. Our findings demonstrate the potential for classifying the four major subtypes of breast cancer at stage Ia with an average sensitivity and specificity of 90% and 95%, respectively, and a cross-validated macro-averaged area under the curve (AUC) of 0.98. This research highlights efforts to integrate vibrational spectroscopy with machine learning, enhancing cancer diagnostics through a non-invasive, personalised approach for early detection and monitoring disease progression. This study is the first of its kind to utilise RS and machine learning to classify the four major breast cancer subtypes at stage Ia. 1 | Introduction Breast cancer (BC) in women is the most frequently diagnosed malignancy worldwide and ranks among the leading causes of cancer-related fatalities [1] . Despite significant advancements in the prevention, diagnosis, and treatment of BC over recent decades [2], it remains a formidable global health challenge. The 5-year survival rate for stage I breast cancer is approximately 99%, compared with 86% for stage II, 57% for stage III, and just 27% for stage IV, where the cancer has spread to distant organs [3, 4] . Often, tumours are detected only

FAQs

What does the study integrate to improve breast cancer detection?

It integrates Raman spectroscopy with machine learning, using blood plasma samples to support early detection and breast cancer subtyping.

How is liquid biopsy used in the study context?

The research is motivated by liquid biopsy’s ability to provide non-invasive, real-time insight into tumour progression, and it aims to address sensitivity limits at early stages.

What performance metrics are reported for classifying stage Ia subtypes?

The model classifies the four major subtypes at stage Ia with average sensitivity of 90% and specificity of 95%, and a cross-validated macro-averaged AUC of 0.98.

Aug 3, 2026

Upload by Theodora