Nonresonant background suppression in wide-field Coherent anti-Stokes Raman scattering microscopy with transport of intensity equation based phase imaging

Abstract

Coherent anti-Stokes Raman scattering (CARS) microscopy is a valuable tool for label-free imaging of biological samples, since it enables providing contrast via vibrational resonances of a specific chemical bond. However, in a conventional CARS image the Raman resonant anti-Stokes radiation is often superimposed by a nonresonant contribution arising from the electronic part of the polarization. The situation becomes worse if a sample is composed of a significant amount of water, where a strong nonresonant background over the whole image is obtained.To date, various approaches including Epi, polarization sensitive, time-resolved, and CARS phase imaging have been implemented to suppress the undesirable nonresonant background in CARS microscopy. Notably, optical heterodyne based phase imaging schemes are of particular interest due to their intrinsic ability to retrieve Im(χ(3)), which is proportional to the Raman resonant signal. Nevertheless, all the reported phase imaging methods that require an independent reference wave lead to an increase in the setup complexity, thus making the measurement sensitive to external perturbations. In order to simplify the setup, single-beam scheme has also been utilized for vibrational CARS imaging by using wave-front sensors to acquire the phase of the complex anti-Stokes amplitude. However, this method demands highly accurate wave-front sensors.In this paper we present a reference-less CARS phase imaging technique to suppress nonresonant CARS background based on transport of intensity equation (TIE). Resonant CARS radiation ECARSR can be obtained when the frequency difference between the pump and Stokes beams is tuned to match a molecular vibration frequency (Raman resonant mode). In contrast, the nonresonant background ECARSNR can be obtained when the frequency difference between the pump and Stokes beams does not match a molecular vibration frequency (Raman resonant mode). Considering the fact that there is a phase shift of π/2 between the resonant and non-resonant CARS field, the phase imaging of both resonant and nonresonant CARS field can provide a background-free image. In implementation, three intensity images of the CARS field under resonant mode are recorded at three neighboring planes by moving the CCD camera along the axial direction. In the meantime, three images of the CARS field under non-resonant mode are also recorded. Considering the fact that the TIE links the intensity distributions in three neighboring planes (through which a beam transverses) with the phase distribution of the field, the phase images of the CARS field under both resonant and nonresonant modes are reconstructed from the recorded intensity images. The phase difference φχ between the resonant CARS field and the non-resonant CARS field is calculated. Eventually, the CARS background is efficiently suppressed by using the relation ICARSbf≅ICARSR·sin2φχ.Compared with conventional CARS background suppression techniques, the proposed method is robust against environmental disturbance, since it does not require an additional reference beam. Furthermore, the proposed method is easy to incorporate in a conventional CARS configuration. Therefore, the proposed method has the potential to become a versatile technique to image deep tissue with low background signal.