针对较厚的组织,普遍采用的自适应光学技术由于其单次校正视场范围有限,空间光调制器或可变形镜的刷新率有限,难以满足大视场范围波前畸变的快速校正,进而难以满足大视场高速成像的需求。结合共轭型自适应光学系统和相干光自适应校正技术,提出了一种并行的波前畸变校正算法,该算法可以在不增加空间光调制器等刷新次数的前提下,通过并行测量多个导引星的波前畸变,实现大视场范围内像差的一次性校正,为生物组织深处的高速、高分辨成像提供一种可行的参考方案。仿真结果表明:在采用9个导引星时,针对5层随机相位屏构成的薄散射介质,该算法单次校正的有效视场约为传统算法的4.7倍;对于120 μm厚的小鼠大脑组织切片样本,单次校正的有效视场约为传统算法的4.6倍。所提算法可以通过增加导引星的数量来进一步增大一次校正的视场范围,并且不会显著增加校正时间,在活体生物样本的大视场成像中具有广阔的应用前景。

As for deep tissues, the field of view of a single correction in the widely used adaptive optics is limited and the refresh rate of a spatial light modulator or a deformable mirror is also limited. Therefore it is difficult for them to satisfy the requirement of large field-of-view (FOV) rapid correction of wavefront distortion and thus that of high-speed imaging. A parallel wavefront correction method is proposed based on the conjugate adaptive optical correction system and the coherent optical adaptive technique. In this method, without increasing the number of refresh times of spatial light modulator, the large FOV of one-time correction can be realized by means of the parallel measurement of wavefront distortion of multiple guide stars, which provides a feasible reference solution for the high-speed and high-resolution imaging of deep tissues. The simulation results show that when 9 guiding stars are used, the effective FOV of a single correction by the proposed method is about 4.7 times than that of the conventional method for a thin scattering medium composed of 5 layers of random phase masks, and 4.6 times than that of the conventional method for 120-μm thick mouse brain tissue. Moreover, the proposed method can further improve the FOV of one-time correction by increasing the number of guide stars while the correction time does not significantly increase, which has broad application prospect in the large FOV imaging of in vivo biological tissues.