AI-powered dual-imaging approach enhances skin cancer detection accuracy

Skin cancer is the most common form of cancer worldwide, and early detection is key to successful treatment. While traditional methods rely on visual inspection and biopsy, researchers are developing new technologies that can provide more detailed information without cutting into the skin. A recent study published in the Journal of Biomedical Optics introduces a compact, noninvasive imaging system that combines two advanced techniques to examine both the structure and chemical composition of skin cancers. This approach could improve how doctors diagnose and classify skin cancer, and how they monitor treatment responses.

Developed by researchers at the Saint-Étienne University Hospital and Paris-Saclay University in collaboration with Damae Medical (France), the system merges two types of imaging: line-field confocal optical coherence tomography (LC-OCT), which captures high-resolution images of skin tissue at the cellular level, and confocal Raman microspectroscopy, which analyzes the chemical makeup of specific areas identified in those images. Together, these tools allow researchers to not only see the shape and structure of cancerous cells but also understand their molecular characteristics.

Over the course of a year, the system was tested in a clinical setting on more than 330 skin cancer samples, specifically nonmelanoma types like basal cell carcinoma and squamous cell carcinoma. The researchers used LC-OCT to locate suspicious structures and then applied Raman microspectroscopy to gather over 1,300 chemical spectra from those areas. To interpret the data, they trained an artificial intelligence (AI) model to recognize patterns associated with cancerous tissues.

The AI model performed well, achieving a classification accuracy of 95 percent for basal cell carcinoma and 92 percent when both types of cancer were included. These results suggest the system can reliably distinguish cancerous structures based on their chemical signatures. Further analysis revealed distinct chemical differences between various cancer types, offering new insights into how these cancers develop and behave.

This dual-imaging approach could lead to more precise, less invasive skin cancer diagnostics in the future. By combining structural and chemical information, clinicians may be able to make faster and more informed decisions about treatment, potentially improving outcomes for patients.