Journal of Biomedical Optics最新文献

Significance: Melanoma's rising incidence demands automatable high-throughput approaches for early detection such as total body scanners, integrated with computer-aided diagnosis. High-quality input data is necessary to improve diagnostic accuracy and reliability.

Aim: This work aims to develop a high-resolution optical skin imaging module and the software for acquiring and processing raw image data into high-resolution dermoscopic images using a focus stacking approach. The obtained hyperfocus images should significantly enhance the diagnostic performance of total body scanners in clinical settings.

Approach: We employed focus stacking to merge multiple images, each with a limited depth of field, into a single hyperfocus image, ensuring every part of the skin is in clear focus. The method was implemented in the high-resolution imaging module using an electrically tunable liquid lens to quickly capture a series of differently focused images in vivo. Algorithms were developed for image alignment, focus measurement, and fusion, with the addition of deep learning-based super-resolution techniques to further enhance image quality. A classification model was trained to provide an artificial intelligence (AI)-based lesion classification.

Results: The developed optical imaging system successfully produced noncontact dermoscopic images with complete focus across all skin topographies. The hyperfocus images obtained demonstrated high resolution of $28\mu m$ and captured focus stacks at 50 frames per second, ensuring rapid acquisition and patient comfort, however, with some variance in resolution of individual lesions compared with contact-based dermoscopy standards.

Conclusions: The focus stacking-based approach for noncontact dermoscopy improves the quality of diagnostic images by ensuring an all-in-focus view of differently shaped skin lesions, essential for early melanoma detection. Although the approach marks a substantial improvement in noninvasive skin imaging, the use of super-resolution techniques requires careful consideration to avoid compromising the authenticity of the raw data. This work enables the usage of advanced imaging and AI techniques in total body scanners for early melanoma detection in clinical practice.

Significance: The spatial and temporal distribution of fluorophore fractions in biological and environmental systems contains valuable information about the interactions and dynamics of these systems. To access this information, fluorophore fractions are commonly determined by means of their fluorescence emission spectrum (ES) or lifetime (LT). Combining both dimensions in temporal-spectral multiplexed data enables more accurate fraction determination while requiring advanced and fast analysis methods to handle the increased data complexity and size.

Aim: We introduce two methods, a phasor and a feedforward neural network (FNN) analysis, to extract fluorophore fractions from temporal-spectral data. These methods aim to handle the increased data complexity and size of temporal-spectral multiplexed data and therefore enable access to a more accurate and fast fraction determination.

Approach: The phasor analysis determines the fraction in each dimension and combines them, whereas the FNN is trained using artificially mixed data. Both methods are compared with the reference method using linear combination-based curve fitting (FIT). The methods are tested in a two-component scenario of exogenous fluorophores with different ES and LT and in a three-component scenario of endogenous fluorophores with similar ES and different LT.

Results: In this case, the phasor analysis showed the lowest absolute errors in the fraction determination (1.4% two-component, 4.7% three-component), outperforming the FNN (6.3%) and FIT (8.7%) analysis, which are both not able to recognize all fluorophores in the three-component scenario. The computational effort was reduced by roughly a factor of 6 (Phasor/FNN) compared with FIT.

Conclusions: Both methods demonstrate substantial advantages over common fitting, offering a faster and more accurate determination of fluorophore fractions. These advancements make temporal-spectral multiplexed data more accessible and practical, particularly for high-speed applications.

Significance: Terahertz (THz) waves have gained significant attention as an imaging technology due to their ability to provide physical and chemical information in a label-free, noninvasive, and nonionizing manner. Notably, their low energy enables nondestructive inspection of internal structures without damaging samples, making them well-suited for biomedical applications. However, the use of THz imaging has been constrained by limited spatial resolution due to the diffraction limit.

Aim: This study introduces an approach using THz scattering-type scanning near-field optical microscopy, an advanced technique capable of overcoming these limitations and enabling single-cell scale measurements to image and distinguish individual bacterial cells, specifically Escherichia coli and Bacillus subtilis, representing Gram-negative and Gram-positive bacteria, respectively.

Approach: We utilized tungsten vertical nanoprobes in an apertureless setup to achieve high-resolution imaging.

Results: In our experiments, bacteria were measured on a hydrophilic gold substrate with a spatial resolution of 50 nm, demonstrating excellent resolution and image contrast. In addition, quantitative analysis using the line dipole image method allowed calculation of the complex refractive indices, revealing clear differences between the two bacterial species.

Conclusions: This technique offers a nonlabel, noninvasive method for bacterial identification, with promising implications for advanced biomedical applications.

Significance: Conventional fringe projection profilometry (FPP) requires multiple image acquisitions and therefore long acquisition times that make it slow for high-speed dynamic measurements. We propose and demonstrate a deep-learning-based single-shot FPP system utilizing a single endoscope for surgical guidance.

Aim: We aim to achieve real-time depth map generation of target tissues with high accuracy for robotic surgical guidance.

Approach: We proposed an endoscopic single-shot FPP system based on a deep learning network to generate real-time accurate tissue depth maps for surgical guidance. The system utilizes a dual-channel endoscope, where one channel projects fringe patterns from a projector and the other channel collects images using a camera. In addition, we developed a data synthesis method to generate a large number of diverse training datasets. The network consists of MaskNet, which segments the tissue from the background, and DepthNet, which predicts the depth map of the image. The results from both networks are combined to generate the final depth map.

Results: We tested our algorithm using fringe patterns with different frequencies and found that the optimal frequency for single-shot FPP in our setup is 20 Hz. The algorithm has been tested on both synthetic and experimental data, achieving a maximum depth prediction error of $\sim 2\mathrm{mm}$ and a processing time of about 12.75 ms per frame.

Conclusion: A deep-learning-based single-shot FPP endoscopic system was shown to be highly effective in real-time depth map generation with millimeter-scale error. Implementing such a system has the potential to improve the reliability of image-guided robotic surgery.

Significance: Bacterial biofilm agglomerates are the cause of hard-to-treat implant-associated infections but currently can only be distinguished using sophisticated microbiological or molecular biological methods. Optical methods can potentially provide a label-free, noncontact approach to detect the presence of bacterial species associated with implant infections that could aid in the early diagnosis of implant-associated diseases.

Aim: Our aim is to measure the polarization signal from implant-associated bacteria biofilms using Mueller matrix polarimetry. Furthermore, we present an analysis of the Mueller matrix element to detect and distinguish the different bacterial biofilm species.

Approach: Several biofilms formed by bacterial species associated with orthopedic (Staphylococcus aureus and Staphylococcus epidermidis) and dental implants (Streptococcus oralis, Streptococcus mutans, and Porphyromonas gingivalis) were grown on titanium, a typical implant material. Polarization signals were acquired in a reflection mode using a calibrated polarimetry setup.

Results: The results show that different biofilms could be qualitatively distinguished using the Mueller matrix element analysis. The values derived from bacterial species measurements were distinctly different from those of the bare titanium discs. From the Lu-Chipman decomposition, parameters such as polarizance and diattenuation were calculated for each of the species.

Conclusions: The results provide deeper insight into the interaction of polarized light with bacterial microcolonies. The physiologically growing biofilms form the basis of their polarimetric response signal. Our approach has potential for fast and nondestructive investigation for implant infection detection, potentially in situ and in vivo.

Significance: Laser speckle contrast imaging (LSCI) is widely used for intraoperative blood flow monitoring, but traditional methods have limitations in imaging low blood flow velocities and small vessels. An improved LSCI method, termed the fluent imaging technique, is proposed to enhance imaging sensitivity and accuracy, providing real-time and high-resolution blood flow assessment for neurosurgical applications.

Aim: We aim to validate the performance of the fluent imaging technique in imaging small vessels with low blood flow velocities and assess its application in cerebrovascular surgical procedures, including carotid artery clamping, reperfusion, and ferric chloride ${(\mathrm{FeCl}}_3)$ -induced thrombosis.

Approach: The fluent imaging technique was validated in vivo using male Sprague-Dawley rats, with three types of experiments: (1) ear vein vessel imaging, (2) proximal common carotid artery blood flow intervention (stenosis and clamping), and (3) ${\mathrm{FeCl}}_3$ -induced thrombosis. Blood flow changes were monitored in real time using an LSCI system, and signal-to-background ratio (SBR) analysis was conducted to assess image quality improvements.

Results: The fluent imaging technique improved image quality, particularly for small vessels and low-velocity blood flow, compared with traditional LSCI methods. In capillary regions, it achieved up to 189% improvement in SBR over spatial contrast (SK) and 37% over AWSDK. In a selected region of interest, the SBR increased from 0.53 (SK) and 1.12 (AWSDK) to 1.53 with the fluent imaging method. In carotid artery interventions, the method effectively captured dynamic blood flow changes, including early Relative Blood Flow Index (RBFI) recovery after clamp release. In ${\mathrm{FeCl}}_3$ -induced thrombosis experiments, it detected vascular occlusion and collateral perfusion.

Conclusions: The fluent imaging technique enhances the accuracy and sensitivity of LSCI for blood flow monitoring in neurosurgery. It provides reliable real-time intraoperative assessment of vascular conditions, improving surgical safety and efficacy. We establish a foundation for its broader clinical application and further optimization.

Significance: Understanding the structural organization of biological tissues is critical for studying their function and response to physiological and pathological conditions. In vivo imaging techniques, such as multiphoton microscopy, enable high-resolution visualization of tissue architecture. However, automated orientation analysis remains challenging due to imaging noise, complexity, and reliance on manual annotations, which are time-consuming and subjective.

Aim: We present a Radon transform-based algorithm for robust, annotation-free structural orientation analysis across multimodal imaging datasets, aiming to improve objectivity and efficiency without introducing preprocessing artifacts.

Approach: The algorithm employs a patch-based Radon transform approach to detect oriented structures in noisy images. By analyzing projection peaks in Radon space, it enhances small structures' visibility while minimizing noise and artifact influence. The method was evaluated using synthetic and in vivo datasets, comparing its performance with human annotations.

Results: The algorithm achieved strong agreement with human annotations, with detection accuracy exceeding 88% across different imaging modalities. Variability among trained raters emphasized the benefits of an objective, mathematically driven approach.

Conclusions: The proposed method provides a robust and adaptable solution for structural orientation analysis in biological images. Its ability to quantify tissue component orientation without preprocessing artifacts makes it valuable for high-resolution, dynamic studies in tissue architecture and biomechanics.

Significance: The expanding use of fluorescence in surgery necessitates standardized characterization methods to facilitate reproducibility and regulatory review of imaging devices. Current guidelines suggest the use of optical phantoms as tools to quantify optical system performance, yet measurements of uniformity and spatial accuracy or distortion remain challenging and are performed in an ad hoc manner or not collected at all.

Aim: We introduce a photostable solid phantom, the reference uniformity and distortion (RUD) phantom, and accompanying analysis code for characterizing fluorescence uniformity and geometric distortion (GD). In addition, the concept of fluorescence flat-field correction is explored using this phantom.

Approach: The RUD phantom was imaged on a custom fluorescence imaging device, as well as five commercial systems. The analysis code characterized uniformity and distortion in these systems. Flat-field correction was explored on the custom device by imaging solid fluorescent reference phantoms at different locations within the field of view.

Results: Successful characterization of the imaging systems' uniformity and GD was achieved. Flat-fielding experiments showed that although it qualitatively improves the appearance of images, it could negatively impact quantitative analyses.

Conclusions: The RUD addresses the need for standardized characterization of fluorescence uniformity and GD. Although fluorescence flat-field correction qualitatively enhances image uniformity, caution is advised as it may adversely affect quantitative accuracy.

Significance: Surgery is a common intervention for patients with pituitary adenomas, particularly those experiencing endocrine symptoms or mass effect. Persistent challenges in pituitary surgery include the detection of small microadenomas, difficulty in discerning residual tumor from normal gland, and infiltrative adenomas. Although standard perioperative diagnostics include magnetic resonance imaging (MRI), computed tomography, ultrasound imaging, and neuronavigation, some centers employ intraoperative MRI, ultrasound, and fluorescence-guided endoscopy to increase the rate of gross total resection and preserve pituitary function. However, these techniques are often limited by availability, time requirements, cost, and inability to provide histological diagnosis.

Aim: This review addresses opportunities to optimize both the extent of resection and gland preservation in pituitary adenoma procedures. We discuss the existing constraints faced in pituitary surgery and showcase the current and emerging detection techniques employed in clinical practice, as well as their limitations. We also discuss newer probing approaches such as elastography and Raman spectroscopy.

Approach: We outline key attributes for an ideal optical tool, considering surgical theater functionality, ergonomics, and result reliability and accuracy.

Results: A case study is presented describing the recent development of a fiber-optics instrument specifically designed for endonasal applications based on clinical requirements, along with preliminary data supporting the feasibility of intraoperative implementation.

Conclusions: Current imaging and navigation tools, although invaluable, have inherent limitations in resolution, integration, and molecular specificity. Raman spectroscopy offers a promising, label-free method for real-time tissue identification, especially when integrated into fiber-optic probes for endonasal use. As a complementary tool, it could enhance intraoperative decision-making and surgical precision. Further clinical validation is needed to support its integration into standard workflows.

Significance: Monte Carlo simulation of light propagation in turbid media is important in biomedical optics. Most existing platforms simulate light-tissue interactions in backscattering and planar geometries and are voxel-based, which limits their ability to model curved boundaries accurately. Few platforms incorporate Doppler shifts from flowing media, and they allow limited customization of flow profiles and scattering properties. Although laser Doppler flowmetry (LDF) is common in backscattering-based tissue measurements or low-scattering through-transmission setups, the intermediate case of through-transmission measurements in more scattering samples is underexplored. This case is relevant for applications such as flow quantification in lab-on-a-chip systems and inline flow sensors for biological fluids.

Aim: To study flow in highly scattering samples (1 to $10{\mathrm{mm}}^{-1}$ ), we developed a voxel-free Monte Carlo simulation platform for through-transmission LDF: MC-Doppler. We compare simulated and experimental Doppler power spectra.

Approach: MC-Doppler uses unit vectors and ray tracing to model light propagation, with fully customizable scattering phase functions and flow fields. It was tested with various suspensions of differently sized polystyrene beads, at flow rates ranging from 0 to $15\mathrm{mL}/\min$ , within a 1 mm diameter glass tube.

Results: Simulated and measured Doppler power spectra matched well for scattering coefficients up to $5{\mathrm{mm}}^{-1}$ . Mismatches between the spectra were found near $10{\mathrm{mm}}^{-1}$ .

Conclusions: MC-Doppler accurately simulates light propagation for through-transmission laser Doppler up to moderate scattering coefficients.