Label-free classification of breast cancer subtypes in ex vivo human tissues using Raman spectroscopy and machine learning
Raman spectroscopy (RS) can distinguish cancerous breast tissue from normal tissue with 97.84% sensitivity and 97.18% specificity, according to a study using a confocal Raman microscope. The technology also identified three specific cancer subtypes—invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC), and ductal carcinoma in situ (DCIS)—with in-class sensitivity between 83% and 96%.
Researchers measured 80 tissue samples from 71 individuals to evaluate the diagnostic performance of RS in ex vivo breast tissue classification. The samples included 46 normal and 34 cancerous tissues.
The study utilized supervised classification to analyze spectral signatures. This process involved both two-class tasks, comparing healthy versus cancerous tissue, and four-class tasks, which differentiated between healthy tissue and the three cancer subtypes.
How does Raman spectroscopy identify breast cancer subtypes?
The technology uses a confocal Raman microscope to investigate spectral signatures within the tissue. According to the study, this method captured clinically relevant differences between invasive and pre-invasive diseases.
For the four-class classification task, the system achieved specificity ranging from 93% to 99%. This allows the technology to separate normal tissue from IDC, ILC, and DCIS with high precision.
Why is this technology significant for breast conserving surgery?
Breast conserving surgery (BCS) aims to remove tumors while preserving the patient’s breast-related quality of life. However, the study notes that accurately identifying the margin between healthy and cancerous tissue remains a complication of the procedure.
The findings demonstrate that RS can accurately distinguish these margins. This capability supports the potential for intraoperative tissue characterization, which means surgeons could potentially identify tissue types during the operation itself.
What may happen next with this technology?
Based on these results, RS may be used as a tool for real-time tissue characterization during BCS. A possible next step could involve the application of these spectral signatures in a live surgical setting to improve margin accuracy.
Such an application is likely to assist in ensuring that cancerous tissue is fully excised while maximizing the preservation of healthy breast tissue.
Frequently Asked Questions
What was the accuracy of Raman spectroscopy in telling the difference between normal and cancerous tissue?
The technology showed 97.84% sensitivity and 97.18% specificity in differentiating cancerous from normal breast tissue.
Which specific types of breast cancer were included in the study?
The study investigated invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC), and ductal carcinoma in situ (DCIS).
What is the main goal of breast conserving surgery (BCS)?
BCS aims to excise breast tumors while preserving the patient’s breast-related quality of life.
Do you think real-time tissue analysis will become a standard part of cancer surgeries?