Intraoperative Tumor Histology to Improve Cancer Surgeries

Surgical removal of cancer remains the first-line treatment for many tumors, but ensuring that all cancerous tissue is removed while preserving healthy tissue is a major challenge. Surgeons currently rely on preoperative imaging and postoperative pathology to assess tumor margins, often learning only after surgery whether cancer cells remain. This delay frequently leads to repeat operations, particularly in breast-conserving procedures such as lumpectomy. A new imaging approach now allows excised tissue to be analyzed during surgery itself, enabling immediate decisions on whether further tissue removal is needed.

Researchers at the California Institute of Technology (Caltech, Pasadena, CA, USA), in collaboration with clinical cancer specialists, have created an imaging method called ultraviolet photoacoustic microscopy (UV-PAM) combined with artificial intelligence (AI) to analyze excised tissue without freezing, fixing, slicing, or chemical staining. The system is designed to fit within strict intraoperative time limits, allowing rapid assessment before a surgical incision is closed.

UV-PAM works by illuminating freshly excised tissue with a low-energy ultraviolet laser tuned to the absorption peak of nucleic acids in DNA and RNA. Cell nuclei absorb the light more strongly than surrounding tissue, creating natural contrast without dyes. The absorbed energy generates ultrasonic waves that are captured to form images at subcellular resolution, while AI algorithms transform the data into images that closely resemble standard hematoxylin and eosin (H&E) pathology slides.

The technique produces images with a resolution of 200–300 nanometers and can analyze large tissue areas within minutes. Deep-learning models trained on extensive histology datasets compare the images to known tissue patterns and provide rapid diagnostic guidance. The results, published in Science Advances, show that UV-PAM can distinguish cancerous from healthy tissue across multiple tissue types, including breast, bone, skin, and organs.

By eliminating time-consuming sample preparation and reducing dependence on subjective interpretation, the method could significantly reduce repeat surgeries caused by unclear margins. Its tissue-agnostic performance suggests broad applicability across cancer types. The research team is continuing validation studies and working toward the development of a commercial system suitable for routine use in operating rooms.

"We're confident we can image everything within 10 minutes, if not 5 minutes or shorter—fast enough to guide decisions before the surgeon closes the incision—and we can provide a lot more data than any single pathologist could read," said Lihong Wang, PhD, senior author of the study. “Importantly, the technique so far appears to be "tissue agnostic," meaning that it works equally well on breast, bone, skin, and organ tissues. We're still in the testing stage with this technology, but we hope to move toward a commercial product that can be widely used."

Related Links:
Caltech