上海理工大学 光电信息与计算机工程学院，上海 200093
细胞显微成像是生物学研究中进行细胞表型检测、获取细胞特征信息的重要手段。传统荧光成像技术是目前主要的细胞成像手段，但是荧光成像系统结构复杂、成本较高，而且特异性染色会对细胞造成损伤。针对此问题，研究了一种虚拟染色技术，使用多模态配准算法执行严格配准明场和荧光图像数据集，改进网络架构、损失函数、后处理、硬件适应性用于训练优化，并且通过虚拟染色评价标准对染色转换偏差进行验证。该方法可以降低荧光成像对荧光成像设备的依赖，无需各种复杂的染色操作，将会减轻生物研究、病理分析、疾病诊断流程的负担。
深度学习 细胞成像 虚拟染色 deep learning cellular imaging virtual staining
Author Affiliations
Abstract
1 Stephenson Research and Technology Center, University of Oklahoma Norman, Oklahoma 73019, USA
2 School of Electrical and Computer Engineering, University of Oklahoma Norman,Oklahoma 73019, USA
Photoacoustic imaging (PAI), also known as optoacoustic imaging, is a rapidly growing imaging modality with potential in medical diagnosis and therapy monitoring. This paper focuses on the techniques of prostate PAI and its potential applications in prostate cancer detection. Transurethral light delivery combined with transrectal ultrasound detection overcomes light scattering in the surrounding tissue and provides optimal photoacoustic signals while minimizing invasiveness. While label-free PAI based on endogenous contrast has promising potential for prostate cancer detection, exogenous contrast agents can further enhance the sensitivity and specificity of prostate cancer PAI. Further in vivo studies are required in order to achieve the translation of prostate PAI to clinical implementation. The minimal invasiveness, relatively low cost, high specificity and sensitivity, and real-time imaging capability are valuable advantages of PAI that may improve the current prostate cancer management in clinic.
Photoacoustic/optoacoustic imaging prostate cancer cancer therapy monitoring prostate endoscopy cellular imaging Journal of Innovative Optical Health Sciences
2017, 10(4): 1730008
