Johns Hopkins photoacoustics tells healthy and cancerous breast tissues apart

18 Feb 2025

Multispectral comparison technique could allow real-time determination during biopsy or surgery.

The ability of photoacoustic (PA) methods to image hemoglobin and hence map blood vessels without needing added markers has made it attractive for medical applications since the technique was first developed.

Its potential as a way to accurately show breast cancer malignancies is particularly attractive for clinicians, with early trials over a decade ago showing that the high degree of PA contrast in near-infrared light absorption between benign and malignant tissues could be put to good use.

A range of other trials and experiments have since applied PA to further clinical tasks, with examples including the use of PA alongside robotic surgery in cardiac procedures and the prospect of using PA instead of fluoroscopy during surgical investigations.

Those two examples came from the Johns Hopkins University Photoacoustic & Ultrasonic Systems Engineering Lab (PULSE), and that group has now developed a novel PA technique intended to assist identification of breast cancer tumors.

Published in Biomedical Optics Express, the findings could point to the determination of tumor margins in real time during biopsy or surgery.

"In contrast to histological analysis, multispectral photoacoustic breast imaging has the potential to provide a rapid, label-free, non-invasive alternative that could significantly expedite the diagnostic process and facilitate immediate clinical decision-making for breast cancer," noted the project in its paper.

"This is the first known application of multispectral photoacoustic imaging to differentiate between healthy and cancerous breast tissues using wavelengths up to 2000 nanometers."

Compare and contrast

The PULSE method involved measuring photoacoustic contrast as a function of wavelength with particular interest in the response of hemoglobin and lipids. This was followed by computational comparison between the image data and existing representative spectra of healthy and malignant areas.

Calculation of the similarity between the spectrum of each pixel in the test and the representative data provided a reliable way to determine the nature of the tissues being examined.

In tests, healthy tissue regions had a 0.967 mean correlation with the representative healthy tissue spectrum, and a lower correlation of 0.801 with a cancer tissue spectrum. Conversely, a test cancer tissue region had a 0.954 mean correlation with the cancer tissue spectrum and a lower mean correlation (0.762) with the healthy tissue spectrum.

These differences stem from physiological differences in the malignant areas, noted the project. The PA signals from healthy tissue are primarily correlated with the optical absorption of deoxyhemoglobin, while the response of invasive ductal carcinoma breast cancer tissue stemmed instead primarily from the optical absorption of lipids. The different PA response offers a reliable way to tell the tissues apart.

Future work at PULSE will include employing 266-nanometer light in the same platform, since this is a wavelength at which cell nuclei have higher absorption coefficients than cytoplasm or the extracellular matrix. This could make the technique's pathological examinations more accurate, bridging the gap between in vivo imaging and traditional histopathology.

"Our approach utilizing multispectral photoacoustic imaging has the potential to enable real-time tumor margin determination during biopsy or surgery," said the project.