Research conducted by PULSE Lab visiting PhD student, Guilherme S. P. Fernandes, and co-advisor, Prof. Muyinatu Bell, is featured on CNN.
The topic of the feature is an algorithm that Prof. Bell invented as a graduate student, which is now revealing capabilities to address skin tone bias in photoacoustic imaging. When using this technique to image through skin with light, darker skin tones produce more acoustic clutter than lighter skin tones, which introduces unwanted biases (e.g., Black patients have worse images than white patients, leading to a disparity in the ability to see the important content needed to make an accurate diagnosis). The novel and innovative algorithm, termed short-lag spatial coherence (SLSC) beamforming, makes clearer pictures for all patients, regardless of skin tone. Prof. Bell collaborated with Prof. Theo Pavan and colleagues at the University of São Paulo in Brazil to test this algorithm on multiple volunteers, and the PULSE Lab hosted Guilherme over the past year to finalize this work, leading to the publication Mitigating skin tone bias in linear array in vivo photoacoustic imaging with short-lag spatial coherence beamforming.
SLSC beamforming was previously shown to reduce acoustic clutter in cardiac ultrasound images and in abdominal ultrasound images of patients with higher body mass indexes. The technique and its derivatives can also be used to clarify the fluid vs. solid content of indeterminate breast masses surrounded by dense breast tissue, thereby reducing unnecessary procedures and follow-ups when trying to detect breast cancer. Additional use cases include clarifying anatomical details in photoacoustic-guided surgery. These multiple examples demonstrate the expansive power, potential, and capabilities of designing more equitable imaging and healthcare technologies that serve a wider range of our global patient population.