Frequency Domain Analysis of Dynamic Light Scattering in the Breast Cancer Risk Screening: A Proof of Concept

TL;DR

This study demonstrates a noninvasive, low-cost spectral analysis technique using dynamic light scattering to detect early biochemical changes in breast tissue, showing promising sensitivity and specificity for cancer risk screening.

Contribution

The paper introduces FEDSA, a novel spectral analysis method for breast cancer risk screening based on dynamic light scattering, validated through experiments with particles and human tissue.

Findings

Power spectra differ between normal and abnormal tissues in the 1-160 kHz range.

ROC analysis indicates 87.5% sensitivity and 68.1% specificity.

Significant spectral differences found in specific breast quadrants.

Abstract

We present a proof of concept for screening breast cancer risk by detecting biochemical alterations in breast tissue using a noninvasive and low-cost technique that uses dynamic light scattering for field effect detection by spectral analysis (FEDSA). This technique consists of a light source that illuminates the tissue, and the backscattering light by the tissue is acquired by two detectors; next, the signal goes to the acquisition system, and then with the designed software the power spectra are calculated. The power spectra contain the frequency contribution related to the size of tissue compounds. These frequency contributions change with the biochemical alterations that are amplified by the field effect on tissue. This implies that the initial alterations in the breast are not local. To test FEDSA, two experiments were performed: the first was with Alumina particles grouped in average sizes of $60-300nm$, $100 - 400nm$ and polystyrene nanoparticles in suspension ($315nm$). The second was with 24 women, 17 of whom had normal tissue and 7 abnormal; abnormalities were previously detected by mammogram or ultrasound. Power spectra were obtained for all particles, in agreement with the particle size. In the case of normal and abnormal tissues, the power spectra of the tissues show differences in the shape of the spectra in the range of 1 to 160 kHz. ROC analysis suggests a possible good sensitivity (87.5\%) and specificity (68.1\%) in classifying breast tissue conditions. Statistical analysis with $p<0.05$ revealed significant differences in two quadrants of the breast, the upper inner right and the upper outer left, which were consistent with previous diagnoses by mammography and ultrasound. This proof of concept opens the possibility of implementing and improving FEDSA in the detection of early anomalies in the breast as a low-cost technique that does not use ionizing radiation.