A photoacoustic tomography image reconstruction method based on forward imaging model
-
摘要:
为了解决在光声层析成像（photo acoustic tomography，PAT）中，由于不均匀光通量分布、组织复杂的光学和声学特性以及超声探测器的非理想特性等因素所致的重建图像质量下降的问题。本文提出一种PAT图像重建方法，建立考虑不均匀光通量、非定常声速、超声探测器的空间脉冲响应和电脉冲响应、有限角度扫描和稀疏采样等因素的前向成像模型，通过交替优化求解成像模型的逆问题，实现光吸收能量分布图和声速分布图的同时重建。仿真、仿体和在体实验结果表明，与反投影法、时间反演法和短滞后空间相干法相比，该方法重建图像的结构相似度和峰值信噪比可分别提高约83%、56%、22%和80%、68%、58%。与传统方法相比，对非理想成像场景采用该方法重建的图像质量有显著提高。
Abstract:The aim of this study is to address the issue of degraded image quality in photoacoustic tomography (PAT) caused by the inhomogeneous light fluence, complex optical and acoustic properties of biological tissues, and non-ideal properties of ultrasonic detectors. This paper proposes a model-based method for reconstructing PAT images. A comprehensive forward imaging model is proposed to describe the physical process of imaging in non-ideal scenarios. The model takes into account variables such as the heterogeneity of light fluence, unsteady speed of sound, spatial and electrical impulse responses of ultrasonic transducers, limited-view scanning, and sparse sampling. The inverse problem of the imaging model is solved by alternate optimization, and images representing optical absorption and SoS distributions are reconstructed simultaneously. The study outcomes indicate that utilizing this method for image reconstruction enhances the structural similarity of the reconstructed images by about 83%, 56%, and 22%, in comparison with back projection, time-reversal, and short-lag spatial coherence techniques, respectively. Additionally, the peak signal-to-noise ratio can be improved by approximately 80%, 68% and 58%, respectively. This method considerably enhances the image quality of non-ideal imaging scenarios when compared to traditional techniques.
-
图 3活体小鼠光声层析成像实验系统示意图
Figure 3.Schematic diagram of PAT experimental setup forin vivomice
图 4根据全角度密集采样的仿真光声信号重建的图像及其评价指标 (a) 标准图像和重建图像；(b)评价指标
Figure 4.Reconstructed distribution maps and their evaluation metrics based on simulated photoacoustic signals that are densely-sampled and collected at a full-angle (a) standard and reconstructed images; (b) evaluation metrics
图 5根据有限角度稀疏采样仿真光声信号重建的光吸收能量分布图及其评价指标 (a) 重建图像；(b) 评价指标
Figure 5.Reconstructed light absorption energy distribution maps and their evaluation metrics based on limited-view sparse sampling simulated data. (a) reconstructed images; (b) evaluation metrics.
图 6仿体图像重建结果及评价指标 (a) 光吸收能量分布图；(b) 声速分布图；(c) 评价指标
Figure 6.Reconstructed distribution maps and their evaluation metrics of phantoms. (a) AOED distribution map; (b) SoS distribution map; (c) evaluation metrics.
图 7小鼠胸腹切片图像重建结果及评价指标 (a) 光吸收能量分布图；(b) 声速分布图；(c) 光吸收能量分布图评价指标
Figure 7.Reconstructed thoracic and abdominal slice images ofin vivomice and evaluation metrics. (a) AOED distribution map; (b) SoS distribution map; (c) evaluation metrics of distribution map.
图 8采用不同迭代初始值时的重建图像和迭代次数 (a) 重建图像；(b) 重建模型1中不同位置处的AOED所需的迭代次数
Figure 8.Reconstructed distribution maps using different iterative initial values and the number of iterations. (a) reconstructed images; (b) number of iterations required for AOED at different locations in the reconstructed model 1
图 9采用不同优化算法重建的AOED分布图及其评价指标 (a) 重建图像；(b) 评价指标
Figure 9.Reconstructed distribution maps and their evaluation metrics of AOED map reconstructed using different optimization algorithms. (a) reconstructed images; (b) evaluation metrics.
图 10不同固定声速条件下，根据仿真光声信号重建的AOED分布图和评价指标 (a) 重建图像；(b) 评价指标
Figure 10.Reconstructed AOED distribution map and evaluation metrics based on simulated photoacoustic signals at different fixed sound speeds. (a) reconstructed images; (b) evaluation metrics.
图 11优化光通量对重建图像质量的影响 (a) 重建的AOED图像；(b) 评价指标
Figure 11.Effect of optimized luminous flux on reconstructed image quality (a) reconstructed AOED image; (b) evaluation metrics.
图 12采用不同的考虑超声探测器响应的方法重建的光吸收能量分布图和评价指标 (a) 重建图像；(b) 评价指标
Figure 12.Reconstructed AOED distribution maps and their evaluation metrics using different methods of considering ultrasonic detector response. (a) reconstructed images; (b) evaluation metrics.
表 1数值仿真模型的组织特性参数
Table 1.Tissue property parameters of numerical phantoms
组织
名称组织
成分折射率 吸收系数
(cm‒1)散射系数
(cm‒1)各向异
性因子声速
(m/s)密度
(kg/L)心脏 肌肉组织 1.37 0.78 132 0.96 1580 1.060 肺 肌肉组织 1.37 0.72 114 0.95 1561 1.043 肾 结缔组织 1.36 0.76 205 0.90 1560 1.050 肝脏 肌肉组织 1.37 0.75 103 0.91 1595 1.060 胸骨 钙质 1.37 0.05 150 0.96 1580 1.050 
下载:
导出CSV
-
[1] YAO J J, WANG L V. Recent progress in photoacoustic molecular imaging[J].Current Opinion in Chemical Biology, 2018, 45: 104-112.doi:10.1016/j.cbpa.2018.03.016 [2] 孙正, 王新宇. 深度学习在光声成像中的应用现状[J]. 计算机科学,2020,47(6A):148-152,156.SUN ZH, WANG X Y. Application of deep learning in photoacoustic imaging[J].Computer Science, 2020, 47(6A): 148-152,156. (in Chinese). [3] COX B T, LAUFER J G, BEARD P C,et al. Quantitative spectroscopic photoacoustic imaging: a review[J].Journal of Biomedical Optics, 2012, 17(6): 061202.doi:10.1117/1.JBO.17.6.061202 [4] JAVAHERIAN A, HOLMAN S. Direct quantitative photoacoustic tomography for realistic acoustic media[J].Inverse Problems, 2019, 35(8): 084004.doi:10.1088/1361-6420/ab091e [5] XU M H, WANG L V. Universal back-projection algorithm for photoacoustic computed tomography[J].Proceedings of SPIE, 2005, 5697: 251-254.doi:10.1117/12.589146 [6] SUN ZH, HAN D D, YUAN Y. 2-D image reconstruction of photoacoustic endoscopic imaging based on time-reversal[J].Computers in Biology and Medicine, 2016, 76: 60-68.doi:10.1016/j.compbiomed.2016.06.028 [7] SHAN H M, WIEDEMAN C, WANG G,et al. Simultaneous reconstruction of the initial pressure and sound speed in photoacoustic tomography using a deep-learning approach[J].Proceedings of SPIE, 2019, 11105: 1110504. [8] LOU Y, WANG K, ORAEVSKY A A,et al. Impact of nonstationary optical illumination on image reconstruction in optoacoustic tomography[J].Journal of the Optical Society of America A, 2016, 33(12): 2333-2347.doi:10.1364/JOSAA.33.002333 [9] 孟琪, 孙正. 生物光声层析成像中不均匀和不稳定照明解决方法[J]. 中国光学,2021,14(2):307-319.doi:10.37188/CO.2020-0142MENG Q, SUN ZH. Solutions to inhomogeneous and unstable illumination in biological photoacoustic tomography[J].Chinese Optics, 2021, 14(2): 307-319. (in Chinese).doi:10.37188/CO.2020-0142 [10] CHO M H, KANG L H, KIM J S,et al. An efficient sound speed estimation method to enhance image resolution in ultrasound imaging[J].Ultrasonics, 2009, 49(8): 774-778.doi:10.1016/j.ultras.2009.06.005 [11] NAPOLITANO D, CHOU C H, MCLAUGHLIN G,et al. Sound speed correction in ultrasound imaging[J].Ultrasonics, 2006, 44 Suppl: e43-e46. [12] PETROSYAN T, THEODOROU M, BAMBER J,et al. Rapid scanning wide-field clutter elimination in epi-optoacoustic imaging using comb LOVIT[J].Photoacoustics, 2018, 10: 20-30.doi:10.1016/j.pacs.2018.02.001 [13] LEDIJU BELL M A, KUO N, SONG D Y,et al. Short-lag spatial coherence beamforming of photoacoustic images for enhanced visualization of prostate brachytherapy seeds[J].Biomedical Optics Express, 2013, 4(10): 1964-1977.doi:10.1364/BOE.4.001964 [14] NGUYEN H N Y, HUSSAIN A, STEENBERGEN W. Reflection artifact identification in photoacoustic imaging using multi-wavelength excitation[J].Biomedical Optics Express, 2018, 9(10): 4613-4630.doi:10.1364/BOE.9.004613 [15] 孙正, 闫向阳. 采用稀疏测量数据的有限角度光声层析成像的研究进展[J]. 声学技术,2020,39(1):1-10.doi:10.16300/j.cnki.1000-3630.2020.01.001SUN ZH, YAN X Y. Progress of limited-view photoacoustic tomography imaging based on sparse measurement[J].Technical Acoustics, 2020, 39(1): 1-10. (in Chinese).doi:10.16300/j.cnki.1000-3630.2020.01.001 [16] LI C H, WANG L V. Photoacoustic tomography and sensing in biomedicine[J].Physics in Medicine & Biology, 2009, 54(19): R59-R97. [17] HOCHULI R, POWELL S, ARRIDGE S,et al. Forward and adjoint radiance Monte Carlo models for quantitative photoacoustic imaging[J].Proceedings of SPIE, 2015, 9323: 93231P. [18] MOHAMMADI L, BEHNAM H, TAVAKKOLI J,et al. Skull’s photoacoustic attenuation and dispersion modeling with deterministic ray-tracing: towards real-time aberration correction[J].Sensors, 2019, 19(2): 345.doi:10.3390/s19020345 [19] WANG K, ERMILOV S A, SU R,et al. An imaging model incorporating ultrasonic transducer properties for three-dimensional optoacoustic tomography[J].IEEE Transactions on Medical Imaging, 2011, 30(2): 203-214.doi:10.1109/TMI.2010.2072514 [20] LIU D C, NOCEDAL J. On the limited memory BFGS method for large scale optimization[J].Mathematical Programming, 1989, 45(1-3): 503-528.doi:10.1007/BF01589116 [21] BECK A, TEBOULLE M. A fast iterative shrinkage-thresholding algorithm for linear inverse problems[J].SIAM Journal on Imaging Sciences, 2009, 2(1): 183-202.doi:10.1137/080716542 [22] HIRAKAWA M, NAGAKUBO D, KANZLER B,et al. Fundamental parameters of the developing thymic epithelium in the mouse[J].Scientific Reports, 2018, 8(1): 11095.doi:10.1038/s41598-018-29460-0 [23] LU T, WANG Y H, LI J,et al. Full-frequency correction of spatial impulse response in back-projection scheme using space-variant filtering for optoacoustic mesoscopy[J].Photoacoustics, 2020, 19: 100193.doi:10.1016/j.pacs.2020.100193 [24] SHENG Q W, WANG K, MATTHEWS T P,et al. A constrained variable projection reconstruction method for photoacoustic computed tomography without accurate knowledge of transducer responses[J].IEEE Transactions on Medical Imaging, 2015, 34(12): 2443-2458.doi:10.1109/TMI.2015.2437356 [25] ZANGERL G, MOON S, HALTMEIER M,et al. Photoacoustic tomography with direction dependent data: an exact series reconstruction approach[J].Inverse Problems, 2019, 35(11): 114005.doi:10.1088/1361-6420/ab2a30 [26] LI M L, WANG L V. A study of reconstruction in photoacoustic tomography with a focused transducer[J].Proceedings of SPIE, 2007, 6437: 64371E. [27] GAVIN H P.The Levenberg-Marquardt algorithm for nonlinear least squares curve-fitting problems[D]. Durham: Duke University, 2020: 1-19.(查阅网上资料, 未能确认文献类型, 请确认).GAVIN H P.The Levenberg-Marquardt algorithm for nonlinear least squares curve-fitting problems[D]. Durham: Duke University, 2020: 1-19.(查阅网上资料, 未能确认文献类型, 请确认). [28] 王倩, 蔡伟伟, 陶波. 基于层析成像的 强度分布测量方法[J]. 中国光学, 2023, 16(4): 743-752.WANG Q, CAI W W, TAO B. Laser intensity distribution measurement method based on tomographic imaging[J]Chinese Optics, 2023, 16(4): 743-752. (in Chinese). [29] HELOU E S, ZIBETTI M V W, HERMAN G T. Fast proximal gradient methods for nonsmooth convex optimization for tomographic image reconstruction[J].Sensing and Imaging, 2020, 21(1): 45.doi:10.1007/s11220-020-00309-z [30] BECK A, TEBOULLE M. Fast gradient-based algorithms for constrained total variation image denoising and deblurring problems[J].IEEE Transactions on Image Processing, 2009, 18(11): 2419-2434.doi:10.1109/TIP.2009.2028250 [31] BECK A, TEBOULLE M. A fast iterative shrinkage-thresholding algorithm for linear inverse problems[J].SIAM Journal on Imaging Sciences, 2009, 2(1): 183-202. (查阅网上资料, 本条与第21条文献重复, 请确认).doi:10.1137/080716542 BECK A, TEBOULLE M . A fast iterative shrinkage-thresholding algorithm for linear inverse problems[J]. SIAM Journal on Imaging Sciences,2009 ,2 （1 ）:183 -202 . (查阅网上资料, 本条与第21条文献重复, 请确认).doi:10.1137/080716542 -
下载:
点击查看大图
计量
- 文章访问数:37
- HTML全文浏览量:19
- PDF下载量:7
- 被引次数:0
