Evaluación del rendimiento de algoritmos de reconstrucción de imágenes para tomografía fotoacústica

Abstract

Photoacoustic tomography (PAT) is an emerging biomedical imaging modality with great clinical potential that combines the use of optical and acoustic techniques for high-resolution imaging of biological tissues. The technique consists in illuminating a tissue using a pulsed laser beam causing local heating of the tissue. Due to the photoacoustic effect, in which optical energy is absorbed and converted into acoustic energy by thermoelastic expansion of the tissue, the rarefaction of the tissue produces a pulsed ultrasonic signal that can be measured with a phased array transducer and converted into images using reconstruction algorithms. To reconstruct the images, beamforming algorithms such as the delay and sum beamformer are used, which is computationally efficient and simple to implement, although it offers low image resolution and contrast. To overcome these limitations, improvements have been made to obtain better image quality. These improvements are based on combining the signals from the different elements, windowing the signal, or even calculating the optimal apodization for each pixel based on the signals received by the ultrasonic elements. This work is a numerical and experimental study that uses lateral and axial resolution and signal-to-noise ratio to evaluate the performance of the main reconstruction algorithms used in PAT. The results show how the adaptive algorithms allow to improve the axial and lateral resolution with respect to the delay and sum beamformer, but in certain situations the size of the objects in the image does not correspond to their real size, which can limit their effectiveness in the quantification of the size of biological structures.

@article{PiSEA2022,
 author = {Irene Pi-Martín and Juan J. García-Garrigós and Alejandro Cebrecos and Noé Jiménez and Francisco Camarena},
 doi = {http://www.sea-acustica.es/fileadmin/publicaciones/02_25.pdf},
 issn = {0210-3680},
 journal = {Revista de Acústica},
 month = {07},
 number = {1-2},
 pages = {17-25},
 title = {Evaluación del rendimiento de algoritmos de reconstrucción de imágenes para tomografía fotoacústica},
 url = {http://www.sea-acustica.es/fileadmin/publicaciones/02_25.pdf},
 volume = {53},
 year = {2022}
}