Compensation for non-uniform illumination and optical fluence attenuation in three-dimensional optoacoustic tomography of the breast

Optoacoustic tomography (OAT) is a promising modality for breast imaging that provides high resolution, detection sensitivity and diagnostic specificity for vascularized breast tumors. In OAT systems employing an arc- shaped illuminator, irregular overlaps of light beams can yield a non-uniform illumination throughout the entire volume of the breast. The imbalance in optical fluence leads to intensity loss in the reconstructed OAT images. Additionally, because optical fluence decreases with depth from breast skin surface, i.e., optical attenuation, deep breast tissues are diminished in the reconstructed images. For qualitative enhancement in 3D OAT imaging, we propose an image processing method to estimate, and compensate for, both the non-uniform incident optical fluence and the optical attenuation. We approximate the non-uniform illumination via maximum intensity extraction for polar angles in a spherical coordinate system. The location of the breast surface is estimated by detecting blood vessels nearest to the breast skin layer that appear with relatively high intensities in the reconstructed image. The breast depth is computed as the minimum distance between each voxel and the detected breast surface. The depth-dependent optical attenuation in the breast is estimated using the Beer– Lambert law down to the maximum penetration depth determined from an analysis of noise and artifacts in the reconstructed image. At each polar angle, the reciprocals of the estimated attenuation is used to compensate for the loss in intensity. The results are that previously invisible structures near the chest wall are revealed, and visible penetration depth was increased by 67% over the conventional, non-compensated volumes.

[1]  Chao Huang,et al.  Accelerating image reconstruction in three-dimensional optoacoustic tomography on graphics processing units. , 2013, Medical physics.

[2]  Lihong V. Wang,et al.  Single-breath-hold photoacoustic computed tomography of the breast , 2018, Nature Communications.

[3]  O. Scherzer Handbook of mathematical methods in imaging , 2011 .

[4]  Mark A. Anastasio,et al.  Full-view 3D imaging system for functional and anatomical screening of the breast , 2018, BiOS.

[5]  D. F. Swinehart,et al.  The Beer-Lambert Law , 1962 .

[6]  M. Elad,et al.  $rm K$-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation , 2006, IEEE Transactions on Signal Processing.

[7]  Lihong V. Wang,et al.  Biomedical Optics: Principles and Imaging , 2007 .

[8]  S. Jacques Optical properties of biological tissues: a review , 2013, Physics in medicine and biology.

[9]  A. Bruckstein,et al.  K-SVD : An Algorithm for Designing of Overcomplete Dictionaries for Sparse Representation , 2005 .

[10]  Mark A. Anastasio,et al.  Photoacoustic and Thermoacoustic Tomography: Image Formation Principles , 2015, Handbook of Mathematical Methods in Imaging.

[11]  Lihong V. Wang,et al.  Universal back-projection algorithm for photoacoustic computed tomography. , 2005 .

[12]  T. Gregorkiewicz,et al.  Efficient carrier multiplication in CsPbI3 perovskite nanocrystals , 2018, Nature Communications.

[13]  Weimin Zhou,et al.  Generation of anatomically realistic numerical phantoms for photoacoustic and ultrasonic breast imaging , 2017, Journal of biomedical optics.

[14]  Kevin Skadron,et al.  Scalable parallel programming , 2008, 2008 IEEE Hot Chips 20 Symposium (HCS).