My research interests include bio-photonics and medical imaging techniques specifically, Optical coherence tomography (OCT) and functional OCT (Doppler OCT, Optical coherence elastography), biomedical optics, optical image processing. A brief overview of some of my published research work is given below.
Research Lab: Biophotonics Imaging and Sensing lab, NJIT
Research advisor: Dr. Xuan Liu, Assistant professor, ECE department of NJIT, https://ece.njit.edu/faculty/xliu
Quantitative Optical Coherence Elastography (qOCE)
Nonlinear Characterization of Elasticity using Quantitative Optical Coherence Elastography
Optical coherence elastography (OCE) is used to perform mechanical characterization on biological tissue at the microscopic scale. In this work, we used quantitative optical coherence elastography (qOCE) to study the nonlinear elastic behavior of biological tissue. The qOCE system had a fiber-optic probe to exert a compressive force to deform tissue under the tip of the probe. Using the space-division multiplexed optical coherence tomography (OCT) signal detected by a spectral domain OCT engine, we were able to simultaneously quantify the probe deformation that was proportional to the force applied, and to quantify the tissue deformation. Our qOCE system allowed us to establish the relationship between mechanical stimulus and tissue response to characterize the stiffness of biological tissue. Most biological tissues have nonlinear elastic behavior, and the apparent stress-strain relationship characterized by our qOCE system was nonlinear an extended range of strain, for a tissue-mimicking phantom as well as biological tissues. Our experimental results suggested that the quantification of force in OCE was critical for accurate characterization of tissue mechanical properties and the qOCE technique was capable of differentiating biological tissues based on the elasticity of tissue that is generally nonlinear.
Figure 1: qOCE probe
- Y. Qiu, F. Zaki, N. Chandra, S. A. Chester, and X. Liu, “Nonlinear characterization of elasticity using quantitative optical coherence elastography,” Biomedical Optics Express, vol. 7, no. 11, pp. 4702-4710. https://doi.org/10.1364/BOE.7.004702
Quantitative Optical Coherence Elastography for Robust Stiffness Assessment
In this work, we demonstrated the capability of quantitative optical coherence elastography (qOCE) for robust assessment of material stiffness under different boundary conditions using the reaction force and displacement field established in the sample. We presented a method to achieve robust stiffness assessment using qOCE data (displacement field and reaction force) and validated the method using experimental data and fitted the result with an analytical model to extract the elastic modulus. The capability to measure stiffness under different boundary conditions is crucial for intraoperative assessment of tumor margin in situ where the boundary condition is usually not known.
Figure 2: Extraction of elastic modulus using qOCE data
- X. Liu, F. Zaki, and Y. Wang, “Quantitative optical coherence elastography for robust stiffness assessment,” Applied Science, vol. 8, no. 8, pp. 1255, 2018. https://doi.org/10.3390/app8081255
Adaptive Doppler Analysis for Robust Handheld Optical Coherence Elastography
In this work, we presented the development of a handheld OCE instrument to conveniently interrogate the localized mechanical properties of in vivo tissue. Handheld OCT imaging device is an attractive detecting and surgical tool for many clinical applications, including guiding vitreous-retinal surgery, delineating tumor margin for surgical excision, and guiding tissue biopsy for the diagnosis of breast or prostate cancer. During handheld OCE characterization, the handheld probe compresses the sample and quantifies the displacement of the sample by OCT signal analyzer. However, the major challenge for manual OCE characterization of tissue is the unpredictable and unstable time varying hand maneuver generated during compression. In addition, the sample deforms under compression, implying spatial variation of motion characteristics. We have described a temporally and spatially adaptive Doppler analysis method for a robust motion tracking method for manual OCE measurement. The method selects the time interval (δt) between signals through Doppler analysis to track the motion speed v(z,t) that varies temporally in a manual compression process and spatially in a deformed sample volume. The method is validated in OCE system with a handheld single fiber probe and real-time signal processing software based on GPU. The method performed an online estimation of the motion speed, selected an optimal δt adaptively and then accomplished robust motion tracking for OCE measurement. The results are obtained from phantom experiments and in vivo tissue characterization demonstrated the effectiveness of the adaptive Doppler analysis for motion tracking in a dynamic manual loading process.
- X. Liu, F. Zaki, H. Wu, C. Wang, and Y. Wang, “Temporally and spatially adaptive Doppler analysis for robust handheld optical coherence elastography,” Biomedical Optics Express, vol. 9, no. 7, pp. 3335-3353, 2018. https://doi.org/10.1364/BOE.9.003335
Optical Coherence Tomography (OCT)
Noise Adaptive Wavelet Thresholding (NAWT) for Speckle Noise Reduction in Optical Coherence Tomography
In this work, we have implemented a noise adaptive wavelet thresholding (NAWT) algorithm to mitigate the speckle noise in OCT images. Speckle noise randomly alters the magnitude of OCT signal and therefore, hinders to identify the subtle features from the sample images which in turn, reduces the effectiveness of OCT system for the clinical applications. Conventional wavelet thresholding algorithms are capable of reducing the speckle noise while conserving the image sharpness. In wavelet thresholding, the magnitude of wavelet coefficients determines if a coefficient is noise or signal. A wavelet coefficient with larger amplitude carries signal information whereas a wavelet coefficient with smaller amplitude is noise. However, speckle noise in OCT images has different characteristics in different spatial scales, which is not considered in conventional wavelet domain thresholding. In our NAWT algorithm, the noise variance (σw2) in individual wavelet sub-band is determined and then the optimal threshold for individual sub-band.is calculated using σw2. The algorithm is simple, fast, effective and is closely related to the physical origin of speckle noise in OCT image. We have also presented a number of examples (homogeneous scattering sample, IR viewing card, ex vivo human fingertip) to mitigate speckle noise by NAWT algorithm in OCT imaging. NAWT algorithm results clearly have demonstrated better performance to adaptively remove speckle noise while preserving the structure features in OCT image compared to conventional wavelet domain thresholding and linear filtering. Moreover, NAWT improves the visual appearance of OCT image and shows better SNR. NAWT algorithm takes approximately 0.2s to process a 512×1024 image using CPU in MATLAB environment. The main steps of NAWT are- wavelet decomposition, soft thresholding and wavelet reconstruction. All these steps can be parallelized using GPU. Therefore, NAWT algorithm can be implemented in GPU for real-time speckle noise removal.
Figure 4: Signal processing flow-chart for NAWT
- F. Zaki, Y. Wang, H. Su, X. Yuan, and X. Liu, “Noise adaptive wavelet thresholding for speckle noise removal in optical coherence tomography,” Biomedical Optics Express, vol.8, no. 5, pp. 2720-2731, 2017. https://doi.org/10.1364/BOE.8.002720
Assessment and Reduction of Additive Noise in a Complex Optical
Coherence Tomography Signal based on Doppler Variation Analysis
In this work, we developed a denoising technique to mitigate the additive noise from the complex OCT signal. Conventional denoising algorithms reflect only the magnitude of OCT and also takes into account the additive Gaussian noise. However, Phase of OCT plays a vital role in several motion tracking applications through Doppler analysis, such as, vascular visualization, blood flow measurement, OCE, cellular motion detection, etc. Moreover, OCT signal affects from both additive Gaussian and multiplicative speckle noises. In our denoising algorithm, we have first mapped and analyzed the characteristics of additive noise through Doppler variation. Next, with the help of local adaptive Weiner filter, we have processed and suppressed the additive noise from the real and imaginary parts of the complex OCT as independent signal channels. The denoising algorithm takes approximately 0.3s to generate a 1024×1024 denoised image using CPU in MATLAB environment. Our denoising algorithm shows the SNR and sensitivity improvement for the structural images (i.e. human finger-tip, IR viewing card), maintains the spatial resolution of OCT without any additional upgradation of SD-OCT imaging setup and data acquisition protocol.
Figure 5: Flow-chart for removal of additive noise from complex OCT
- X. Liu, F. Zaki, and D. Renaud, “Assessment and removal of additive noise in a complex optical coherence tomography signal based on Doppler variation analysis,” Applied Optics, vol. 57, no. 11, pp. 2873-2880, 2018. https://doi.org/10.1364/AO.57.002873
Secure Fingerprint Identification based on Structural and Microangiographic Optical Coherence Tomography
Optical coherence tomography (OCT) allows noncontact acquisition of fingerprints and hence is a highly promising technology in the field of biometrics. OCT can be used to acquire both structural and microangiographic images of fingerprints. Microangiographic OCT derives its contrast from the blood flow in the vasculature of viable skin tissue, and microangiographic fingerprint imaging is inherently immune to fake fingerprint attack. Therefore, dual-modality (structural and microangiographic) OCT imaging of fingerprints will enable more secure acquisition of biometric data, which has not been investigated before. Our study on fingerprint identification based on structural and microangiographic OCT imaging is, we believe, highly innovative. In this study, we performed OCT imaging study for fingerprint acquisition and demonstrated the capability of dual-modality OCT imaging for the identification of fake fingerprints.
- X. Liu, F. Zaki, Y. Wang, Q. Huang, X. Mei and J. Wang, “Secure fingerprint identification based on structural and microangiographic optical coherence tomography,” Applied Optics, vol. 56, no. 8, pp. 2255-2259, 2017. https://doi.org/10.1364/AO.56.002255
Optical Coherence Tomography for Non-invasive Examination and
Conservation of Cultural Heritage Objects
Cultural heritage works, such as ancient murals and historical paintings, are examined routinely for the conservation purpose. In conventional optical coherence tomography (OCT), which is a three-dimensional (3D) microscopic imaging modality in the field of heritage works conservation, data acquisition by OCT offers both 3D surface information of the object and structural information beneath the surface. Therefore, cross-sectional information on the object can be employed to study the artist’s painting layer structure and brush stroke techniques. However, conventional OCT has limited capability in high-definition (HD) examination of paintings or murals that are in macroscopic scale. HD examination of heritage works needs to scan large areas and process huge amounts of data, while OCT imaging has a limited field of view and processing power. In this work, we developed a novel high-speed, large field-of-view (FOV) OCT imaging platform to advance the OCT application in the conservation of heritage works. Our results suggest that this OCT platform has the potential to become a nondestructive alternative for the analysis and conservation of paintings and murals.
- F. Zaki, I. Hou, D. Cooper, D. Patel, Y. Yang, and X. Liu, “High-definition optical coherence tomography imaging for noninvasive examination of heritage works,” Applied Optics, vol. 55, no. 36, pp. 10313-10317, 2016. https://doi.org/10.1364/AO.55.010313
Previous Research Work
Research Institute: Department of Electrical and Electronic Engineering (EEE), Bangladesh University of Engineering and Technology (BUET)
Research advisor: Dr. Mohammad Faisal, Professor, EEE department of BUET
Analysis of Third Order Dispersion in Ultra-High Speed Optical Fiber Communication System and its Compensation Technique
In this work, we presented a comprehensive investigation on pulse distortions due to the third-order dispersion (TOD) on ultra-high speed long-haul single channel optical fiber communication system using OptiSystem. The optical communication system consists of dispersion-managed line with periodic amplification by Er-doped fiber amplifiers. The presence of the TOD introduces broadening and an additional temporal shift on the propagating pulse. The impact of TOD is observed at the receiving end of transmission line considering the variation of different factors such as transmission reach, bit rate, duty cycle, pulse shape and fiber type. Only self-phase modulation, second and third order dispersion, fiber loss, and amplified spontaneous emission (ASE) noise are considered here. BER performance is also observed considering receiver noise.
The numerical result shows that temporal effect on pulse center decreases in the case of RZ Gaussian pulse while using SSMF-DCF system and when both group velocity dispersion and TOD effects are considered.
Figure 8: System model
- F. Zaki and M. Faisal, “Impact of third-order dispersion in ultra-high speed long-haul optical fiber communication system,” in 2nd International Conference on Informatics, Electronics & Vision (ICIEV), 2013, pp.1-5, 17-18 May 2013.