AI-driven projection tomography with multicore fibre-optic cell rotation.

Optical tomography has emerged as a non-invasive imaging method, providing three-dimensional insights into subcellular structures and thereby enabling a deeper understanding of cellular functions, interactions, and processes. Conventional optical tomography methods are constrained by a limited illumination scanning range, leading to anisotropic resolution and incomplete imaging of cellular structures. To overcome this problem, we employ a compact multi-core fibre-optic cell rotator system that facilitates precise optical manipulation of cells within a microfluidic chip, achieving full-angle projection tomography with isotropic resolution. Moreover, we demonstrate an AI-driven tomographic reconstruction workflow, which can be a paradigm shift from conventional computational methods, often demanding manual processing, to a fully autonomous process. The performance of the proposed cell rotation tomography approach is validated through the three-dimensional reconstruction of cell phantoms and HL60 human cancer cells. The versatility of this learning-based tomographic reconstruction workflow paves the way for its broad application across diverse tomographic imaging modalities, including but not limited to flow cytometry tomography and acoustic rotation tomography. Therefore, this AI-driven approach can propel advancements in cell biology, aiding in the inception of pioneering therapeutics, and augmenting early-stage cancer diagnostics.

K-CapsNet: K-Nearest Neighbor Based Convolution Capsule Network for Cerenkov Luminescence Tomography Reconstruction.

Cerenkov luminescence tomography (CLT) has received significant attention as a promising imaging modality that can display the three-dimensional (3D) distribution of radioactive probes. However, the reconstruction of CLT suffers from severe ill-posed problem. It is difficult for traditional model-based method to obtain satisfactory result. Recently, deep learning-based method have shown great potential for accurate and efficient CLT reconstruction. In this study, a KNN-based convolution capsule network, named K-CapsNet, is proposed for cerenkov luminescence tomography. In K-CapsNet, the surface photon intensity is encoded in capsule form. The KNN-based convolution and K-means clustering are proposed for efficient encoding. Numerical simulation experiments have been carried out to verify the performance of K-CapsNet, and the results show that it performs superior in source localization and morphological restoration compared with existing methods.

FISTA-NET: Deep Algorithm Unrolling for Cerenkov luminescence tomography.

Cerenkov luminescence tomography (CLT) is a highly sensitive and promising imaging technique that can be used to reconstruct the three-dimensional distribution of radioactive probes in living animals. However, the accuracy of CLT reconstruction is limited by the simplified radiative transfer equation and ill-conditioned inverse problem. To address this issue, we propose a model-based deep learning network that combines the neural network with a model-based approach to enhance the performance of CLT reconstruction. The Fast Iterative Shrinkage Thresholding Algorithm (FISTA), a traditional model-based approach, is expanded into a deep network (known as FISTA-NET). Each layer in the network represents an iteration of the algorithm steps, and connecting these layers can form a deep neural network. In addition, different from the traditional FISTA, the key parameters in FISTA, such as gradient step size and threshold value, can be learned through training data without manual production. To evaluate the performance of FISTA-NET, numerical simulation experiments were conducted, which demonstrate its excellent positioning and shape recovery abilities.Clinical Relevance-This indicates that FISTA-NET strategy can significantly improve the quality of CLT reconstruction, which is further beneficial to the assessment of disease activity and treatment effect based on CLT.

One-dimensional convolutional neural network for Jacobian in Diffuse Optical Tomography.

Non-linear least square minimization algorithms are often employed to solve Diffuse Optical Tomography (DOT) inverse problem. However, it is time-consuming to calculate the Jacobian matrix. This work has proposed a data-driven neural network method to improve computational efficiency. The singular value decomposition is employed to compute the updated Jacobian and a mapping from boundary measurements to the singular values based on a convolutional neural network (CNN) is learned to obtain the singular values. The method is validated with 3D numerical simulation data. We have demonstrated that the approach can save computation time compared to Adjoint method, and reconstructed absorption coefficient close to Adjoint method.Clinical Relevance- These results are not focused on clinical relevance currently, but in the future may be helpful to accelerant DOT reconstruction in clinic.

Image Reconstruction Using Deep Learning for Near-Infrared Optical Tomography: Generalization Assessment.

Time is one of the most critical factors in preventing brain lesions due to hypoxic ischemia in preterm infants. Since early detection of low oxygenation is vital and the time window for therapy is narrow, near-infrared optical tomography (NIROT) must be able to process the high-dimensional data provided by today's advanced systems in the shortest possible time. Deep learning approaches are attractive because they can exploit such high information density while reducing inference time. The aim of this study was to evaluate the performance of a hybrid convolutional neural network, designed for NIROT image reconstruction and trained on synthetic data. Generalization capability was assessed using measurements on phantoms of a surface topology more divergent than the range of variation in the geometries of the in-silico data, with unseen, non-spherical inclusion shapes, and with source and detector arrangements different from those used for data generation. Substantial gains in speed, localization accuracy, and high image quality were achieved even under the highly varied measurement conditions.

Time Domain Near-Infrared Optical Tomography Utilizing Full Temporal Data: A Simulation Study.

The analysis of full temporal data in time-domain near-infrared optical tomography (TD NIROT) measurements enables valuable information to be obtained about tissue properties with good temporal and spatial resolution. However, the large amount of data obtained is not easy to handle in the image reconstruction. The goal of the project is to employ full-temporal data from a TD NIROT modality. We improved TD data-based 3D image reconstruction and compared the performance with other methods using frequency domain (FD) and temporal moments. The iterative reconstruction algorithm was evaluated in simulations with both noiseless and noisy in-silico data. In the noiseless cases, a superior image quality was achieved by the reconstruction using full temporal data, especially when dealing with inclusions at 20 mm and deeper in the tissue. When noise similar to measured data was present, the quality of the recovered image from full temporal data was no longer superior to the one obtained from the analysis of FD data and temporal moments. This indicates that denoising methods for TD data should be developed. In conclusion, TD data contain richer information and yield better image quality.

Imaging Cerebral Blood Vessels Using Near-Infrared Optical Tomography: A Simulation Study.

Cerebral veins have received increasing attention due to their importance in preoperational planning and the brain oxygenation measurement. There are different modalities to image those vessels, such as magnetic resonance angiography (MRA) and recently, contrast-enhanced (CE) 3D gradient-echo sequences. However, the current techniques have certain disadvantages, i.e., the long examination time, the requirement of contrast agents or inability to measure oxygenation. Near-infrared optical tomography (NIROT) is emerging as a viable new biomedical imaging modality that employs near infrared light (650-950 nm) to image biological tissue. It was proven to easily penetrate the skull and therefore enables the brain vessels to be assessed. NIROT utilizes safe non-ionizing radiation and can be applied in e.g., early detection of neonatal brain injury and ischemic strokes. The aim is to develop non-invasive label-free dynamic time domain (TD) NIROT to image the brain vessels. A simulation study was performed with the software (NIRFAST) which models light propagation in tissue with the finite element method (FEM). Both a simple shape mesh and a real head mesh including all the segmented vessels from MRI images were simulated using both FEM and a hybrid FEM-U-Net network, we were able to visualize the superficial vessels with NIROT with a Root Mean Square Error (RMSE) lower than 0.079.

3D autofluorescence imaging of hydronephrosis and renal anatomical structure using cryo-micro-optical sectioning tomography.

Rationale: Mesoscopic visualization of the main anatomical structures of the whole kidney in vivo plays an important role in the pathological diagnosis and exploration of the etiology of hydronephrosis. However, traditional imaging methods cannot achieve whole-kidney imaging with micron resolution under conditions representing in vivo perfusion. Methods: We used in vivo cryofixation (IVCF) to fix acute obstructive hydronephrosis (unilateral ureteral obstruction, UUO), chronic spontaneous hydronephrosis (db/db mice), and their control mouse kidneys for cryo-micro-optical sectioning tomography (cryo-MOST) autofluorescence imaging. We quantitatively assessed the kidney-wide pathological changes in the main anatomical structures, including hydronephrosis, renal subregions, arteries, veins, glomeruli, renal tubules, and peritubular functional capillaries. Results: By comparison with microcomputed tomography imaging, we confirmed that IVCF can maintain the status of the kidney in vivo. Cryo-MOST autofluorescence imaging can display the main renal anatomical structures with a cellular resolution without contrast agents. The hydronephrosis volume reached 26.11 ± 6.00 mm3 and 13.01 ± 3.74 mm3 in 3 days after UUO and in 15-week-old db/db mouse kidneys, respectively. The volume of the cortex and inner stripe of the outer medulla (ISOM) increased while that of the inner medulla (IM) decreased in UUO mouse kidneys. Db/db mice also showed an increase in the volume of the cortex and ISOM volume but no atrophy in the IM. The diameter of the proximal convoluted tubule and proximal straight tubule increased in both UUO and db/db mouse kidneys, indicating that proximal tubules were damaged. However, some renal tubules showed abnormal central bulge highlighting in the UUO mice, but the morphology of renal tubules was normal in the db/db mice, suggesting differences in the pathology and severity of hydronephrosis between the two models. UUO mouse kidneys also showed vascular damage, including segmental artery and vein atrophy and arcuate vein dilation, and the density of peritubular functional capillaries in the cortex and IM was reduced by 37.2% and 49.5%, respectively, suggesting renal hypoxia. In contrast, db/db mouse kidneys showed a normal vascular morphology and peritubular functional capillary density. Finally, we found that the db/db mice displayed vesicoureteral reflux and bladder overactivity, which may be the cause of hydronephrosis formation. Conclusions: We observed and compared main renal structural changes in hydronephrosis under conditions representing in vivo perfusion in UUO, db/db, and control mice through cryo-MOST autofluorescence imaging. The results indicate that cryo-MOST with IVCF can serve as a simple and powerful tool to quantitatively evaluate the in vivo pathological changes in three dimensions, especially the distribution of body fluids in the whole kidney. This method is potentially applicable to the three-dimensional visualization of other tissues, organs, and even the whole body, which may provide new insights into pathological changes in diseases.

Myopic Vascular Changes Revealed by Optical Tomography Angiography and Their Association with Myopic Fundus Manifestations.

INTRODUCTION: We aimed to quantify and evaluate fundal vascular changes at different severities of myopia using optical tomography angiography (OCTA) and explore their association with fundus changes captured by ultra-widefield (UWF) fundus cameras. METHODS: Seventy-four participants with myopia were enrolled in the study and underwent basic ophthalmic examination, OCTA, and UWF fundus photography. Multiple parameters were obtained using OCTA (flow area, structure thickness, and vessel density) and UWF fundus cameras (tessellation and parapapillary atrophy [PPA]). RESULTS: The right eye of 30 participants with low and moderate myopia and 44 participants with high myopia (HM) were included. Patients with HM had a larger flow area of the outer retina (FA-OR) and a smaller thickness of choroid (TC). Axial length was significantly correlated with retinal and choroidal flow area and thickness in the different zones. The PPA area was positively correlated with FA-OR and negatively correlated with TC. Tessellation exhibited different levels of correlation with OCTA parameters regarding the flow area, thickness, and vessel density of the fundal layers, mainly in the inner retina. CONCLUSION: FA-OR and TC exhibited sensitive changes in patients with HM and axial elongation; therefore, they could serve as predictive OCTA biomarkers. The PPA and tessellation were connected to the vascular and structural changes revealed by OCTA.

Cross-scale tracing of nanoparticles and tumors at the single-cell level using the whole-lung atlas.

Currently, the effectiveness of oncotherapy is limited by tumor heterogeneities, which presents a huge challenge for the development of nanotargeted drug delivery systems (DDSs). Therefore, it is important to resolve the spatiotemporal interactions between tumors and nanoparticles. However, targeting evaluation has been limited by particle visualization due to the gap between whole-organ scale and subcellular precision. Here, a high-precision three-dimensional (3D) visualization of tumor structure based on the micro-optical sectioning tomography (MOST) system and fluorescence MOST (fMOST) system is presented to clarify 3D spatial distribution of nanoparticles within the tumor. We demonstrate that through the MOST/fMOST system, it is possible to reveal multidimensional and cross-scale correlations between the tumor structure and nanoparticle distribution to remodel the tumor microenvironment and explore the structural parameters of vasculature. This visualization methodology provides an accurate assessment of the efficacy, distribution, and targeting efficiency of DDSs for oncotherapy compared to available approaches.

Two-stage classification strategy for breast cancer diagnosis using ultrasound-guided diffuse optical tomography and deep learning.

Significance: Ultrasound (US)-guided diffuse optical tomography (DOT) has demonstrated great potential for breast cancer diagnosis in which real-time or near real-time diagnosis with high accuracy is desired. Aim: We aim to use US-guided DOT to achieve an automated, fast, and accurate classification of breast lesions. Approach: We propose a two-stage classification strategy with deep learning. In the first stage, US images and histograms created from DOT perturbation measurements are combined to predict benign lesions. Then the non-benign suspicious lesions are passed through to the second stage, which combine US image features, DOT histogram features, and 3D DOT reconstructed images for final diagnosis. Results: The first stage alone identified 73.0% of benign cases without image reconstruction. In distinguishing between benign and malignant breast lesions in patient data, the two-stage classification approach achieved an area under the receiver operating characteristic curve of 0.946, outperforming the diagnoses of all single-modality models and of a single-stage classification model that combines all US images, DOT histogram, and imaging features. Conclusions: The proposed two-stage classification strategy achieves better classification accuracy than single-modality-only models and a single-stage classification model that combines all features. It can potentially distinguish breast cancers from benign lesions in near real-time.

Application of fluorescence micro-optical sectioning tomography in the cerebrovasculature and applicable vascular labeling methods.

Fluorescence micro-optical sectioning tomography (fMOST) is a three-dimensional (3d) imaging method at the mesoscopic level. The whole-brain of mice can be imaged at a high resolution of 0.32 × 0.32 × 1.00 µm3. It is useful for revealing the fine morphology of intact organ tissue, even for positioning the single vessel connected with a complicated vascular network across different brain regions in the whole mouse brain. Featuring its 3d visualization of whole-brain cross-scale connections, fMOST has a vast potential to decipher brain function and diseases. This article begins with the background of fMOST technology including a widespread 3D imaging methods comparison and the basic technical principal illustration, followed by the application of fMOST in cerebrovascular research and relevant vascular labeling techniques applicable to different scenarios.

Single-exposure ultrasound-modulated optical tomography with a quaternary phase encoded mask.

Ultrasound-modulated optical tomography (UOT) is a deep-tissue imaging modality that provides optical contrast with acoustic resolution. Among existing implementations, camera-based UOT improves modulation depth through parallel detection but suffers from a low camera frame rate. The condition prohibits this technique from being applied to in vivo applications where speckles decorrelate on a time scale of 1 ms or less. To overcome this challenge, we developed single-exposure camera-based UOT by employing a quaternary phase encoded mask (QPEM). As a proof of concept, we demonstrated imaging of an absorptive target buried inside a dynamic scattering medium with a speckle correlation time as short as 0.49 ms, typical of living biological tissues. Benefiting from the QPEM-enabled single-exposure wavefront measurement (5.5 ms) and GPU-assisted wavefront reconstruction (0.97 ms), the point scanning and result update speed can reach up to 150 Hz. We envision that the QPEM-enabled single-exposure scheme paves the way for in vivo UOT imaging, which holds promise for a variety of medical and biological applications.

Diffuse optical tomography spatial prior for EEG source localization in human visual cortex.

Electroencephalography (EEG) and diffuse optical tomography (DOT) are imaging methods which are widely used for neuroimaging. While the temporal resolution of EEG is high, the spatial resolution is typically limited. DOT, on the other hand, has high spatial resolution, but the temporal resolution is inherently limited by the slow hemodynamics it measures. In our previous work, we showed using computer simulations that when using the results of DOT reconstruction as the spatial prior for EEG source reconstruction, high spatio-temporal resolution could be achieved. In this work, we experimentally validate the algorithm by alternatingly flashing two visual stimuli at a speed that is faster than the temporal resolution of DOT. We show that the joint reconstruction using both EEG and DOT clearly resolves the two stimuli temporally, and the spatial confinement is drastically improved in comparison to reconstruction using EEG alone.

Three-dimensional label-free visualization of the interactions of PM2.5 with macrophages and epithelial cells using optical diffraction tomography.

Particulate matter ≤ 2.5 µm (PM2.5) poses health risks related to various diseases and infections. However, the interactions between PM2.5 and cells such as uptake and cell responses have not been fully investigated despite advances in bioimaging techniques, because the heterogeneous morphology and composition of PM2.5 make it challenging to employ labeling techniques, such as fluorescence. In this work, we visualized the interaction between PM2.5 and cells using optical diffraction tomography (ODT), which provides quantitative phase images by refractive index distribution. Through ODT analysis, the interactions of PM2.5 with macrophages and epithelial cells, such as intracellular dynamics, uptake, and cellular behavior, were successfully visualized without labeling techniques. ODT analysis clearly shows the behavior of phagocytic macrophages and nonphagocytic epithelial cells for PM2.5. Moreover, ODT analysis could quantitatively compare the accumulation of PM2.5 inside the cells. PM2.5 uptake by macrophages increased substantially over time, but uptake by epithelial cells increased only marginally. Our findings indicate that ODT analysis is a promising alternative approach to visually and quantitatively understanding the interaction of PM2.5 with cells. Therefore, we expect ODT analysis to be employed to investigate the interactions of materials and cells that are difficult to label.

Mapping cortical activations underlying covert and overt language production using high-density diffuse optical tomography.

Gold standard neuroimaging modalities such as functional magnetic resonance imaging (fMRI), positron emission tomography (PET), and more recently electrocorticography (ECoG) have provided profound insights regarding the neural mechanisms underlying the processing of language, but they are limited in applications involving naturalistic language production especially in developing brains, during face-to-face dialogues, or as a brain-computer interface. High-density diffuse optical tomography (HD-DOT) provides high-fidelity mapping of human brain function with comparable spatial resolution to that of fMRI but in a silent and open scanning environment similar to real-life social scenarios. Therefore, HD-DOT has potential to be used in naturalistic settings where other neuroimaging modalities are limited. While HD-DOT has been previously validated against fMRI for mapping the neural correlates underlying language comprehension and covert (i.e., "silent") language production, HD-DOT has not yet been established for mapping the cortical responses to overt (i.e., "out loud") language production. In this study, we assessed the brain regions supporting a simple hierarchy of language tasks: silent reading of single words, covert production of verbs, and overt production of verbs in normal hearing right-handed native English speakers (n = 33). First, we found that HD-DOT brain mapping is resilient to movement associated with overt speaking. Second, we observed that HD-DOT is sensitive to key activations and deactivations in brain function underlying the perception and naturalistic production of language. Specifically, statistically significant results were observed that show recruitment of regions in occipital, temporal, motor, and prefrontal cortices across all three tasks after performing stringent cluster-extent based thresholding. Our findings lay the foundation for future HD-DOT studies of imaging naturalistic language comprehension and production during real-life social interactions and for broader applications such as presurgical language assessment and brain-machine interfaces.

Surgical Treatment of Blow-out Fracture: A New Perspective.

The present work aims to report 2 clinical cases of the use of current technologies for the treatment of orbital fractures. The cases are of patients who were victims of car accidents and who developed a blow-out orbital fracture. Clinically, they presented periorbital ecchymosis, blepharoedema, enophthalmos, and ophthalmoplegia and, therefore, underwent surgical reconstructive treatment. For both cases, preoperative computed tomography and biomodel impression of the orbits were performed. The modeling of the titanium mesh covering the defect in the biomodel that would be used in the surgery was performed. In the intraoperative period of reduction and fixation of the fracture with the titanium mesh, optics were used to better visualize the posterior defect, as well as computed tomography to ensure that the entire affected area was reconstructed. Both patients were followed up in the postoperative period and evolved without clinical and functional complaints.

Stable tissue-mimicking phantoms for longitudinal multimodality imaging studies that incorporate optical, CT, and MRI contrast.

Significance: Tissue phantoms that mimic the optical and radiologic properties of human or animal tissue play an important role in the development, characterization, and evaluation of imaging systems. Phantoms that are easily produced and stable for longitudinal studies are highly desirable. Aim: A new type of long-lasting phantom was developed with commercially available materials and was assessed for fabrication ease, stability, and optical property control. Magnetic resonance imaging (MRI) and x-ray computed tomography (CT) contrast properties were also evaluated. Approach: A systematic investigation of relationships between concentrations of skin-like pigments and composite optical properties was conducted to realize optical property phantoms in the red and near-infrared (NIR) wavelength range that also offered contrast for CT and MRI. Results: Phantom fabrication time was < 1 h and did not involve any heating or cooling processes. Changes in optical properties were < 2 % over a 12-month period. Phantom optical and spectral features were similar to human soft tissue over the red to NIR wavelength ranges. Pigments used in the study also had CT and MRI contrasts for multimodality imaging studies. Conclusions: The phantoms described here mimic optical properties of soft tissue and are suitable for multimodality imaging studies involving CT or MRI without adding secondary contrast agents.

Optical imaging of the small intestine immune compartment across scales.

The limitations of 2D microscopy constrain our ability to observe and understand tissue-wide networks that are, by nature, 3-dimensional. Optical projection tomography (OPT) enables the acquisition of large volumes (ranging from micrometres to centimetres) in various tissues. We present a multi-modal workflow for the characterization of both structural and quantitative parameters of the mouse small intestine. As proof of principle, we evidence its applicability for imaging the mouse intestinal immune compartment and surrounding mucosal structures. We quantify the volumetric size and spatial distribution of Isolated Lymphoid Follicles (ILFs) and quantify the density of villi throughout centimetre-long segments of intestine. Furthermore, we exhibit the age and microbiota dependence for ILF development, and leverage a technique that we call reverse-OPT for identifying and homing in on regions of interest. Several quantification capabilities are displayed, including villous density in the autofluorescent channel and the size and spatial distribution of the signal of interest at millimetre-scale volumes. The concatenation of 3D imaging with reverse-OPT and high-resolution 2D imaging allows accurate localisation of ROIs and adds value to interpretations made in 3D. Importantly, OPT may be used to identify sparsely-distributed regions of interest in large volumes whilst retaining compatibility with high-resolution microscopy modalities, including confocal microscopy. We believe this pipeline to be approachable for a wide-range of specialties, and to provide a new method for characterisation of the mouse intestinal immune compartment.

Unrolled-DOT: an interpretable deep network for diffuse optical tomography.

Significance: Imaging through scattering media is critical in many biomedical imaging applications, such as breast tumor detection and functional neuroimaging. Time-of-flight diffuse optical tomography (ToF-DOT) is one of the most promising methods for high-resolution imaging through scattering media. ToF-DOT and many traditional DOT methods require an image reconstruction algorithm. Unfortunately, this algorithm often requires long computational runtimes and may produce lower quality reconstructions in the presence of model mismatch or improper hyperparameter tuning. Aim: We used a data-driven unrolled network as our ToF-DOT inverse solver. The unrolled network is faster than traditional inverse solvers and achieves higher reconstruction quality by accounting for model mismatch. Approach: Our model "Unrolled-DOT" uses the learned iterative shrinkage thresholding algorithm. In addition, we incorporate a refinement U-Net and Visual Geometry Group (VGG) perceptual loss to further increase the reconstruction quality. We trained and tested our model on simulated and real-world data and benchmarked against physics-based and learning-based inverse solvers. Results: In experiments on real-world data, Unrolled-DOT outperformed learning-based algorithms and achieved over 10× reduction in runtime and mean-squared error, compared to traditional physics-based solvers. Conclusion: We demonstrated a learning-based ToF-DOT inverse solver that achieves state-of-the-art performance in speed and reconstruction quality, which can aid in future applications for noninvasive biomedical imaging.