Non-Negative Iterative Convex Refinement Approach for Accurate and Robust Reconstruction in Cerenkov Luminescence Tomography

Fibroblast activation protein peptide-targeted NIR-I/II fluorescence imaging for stable and functional detection of hepatocellular carcinoma

PURPOSE

Cancer-associated fibroblasts (CAFs) are the primary stromal component of the tumor microenvironment in hepatocellular carcinoma (HCC), affecting tumor progression and post-resection recurrence. Fibroblast activation protein (FAP) is a key biomarker of CAFs. However, there is limited evidence on using FAP as a target in near-infrared (NIR) fluorescence imaging for HCC. Thus, this study aims to develop a novel NIR fluorescent imaging strategy targeting FAP+ CAFs in HCC.

METHODS

The ICG-FAP-TATA probe was synthesized by conjugating a novel cyclization anti-FAP peptide with an indocyanine green derivative (ICG-NH2) as fluorophore, capable for NIR window I (NIR-I, 700-900 nm) and II (NIR-II, 1000-1700 nm) imaging. Its efficacy in lesion localization and other potential applications was evaluated.

RESULTS

In vivo imaging of subcutaneous HCC models revealed that ICG-FAP-TATA specifically targeted FAP+ CAFs in the stroma and detected differences in CAFs loading within lesions. The fluorescence intensity/tumor-to-background ratio (TBR) positively correlated with FAP expression (R2 > 0.8, p < 0.05). Ex vivo incubation of tumor tissues with ICG-FAP-TATA provided stable fluorescence imaging of tumors in subcutaneous and orthotopic HCC models, including different cell line co-culture systems (LM3-luc, MHCC97H-luc, HepG2-luc + LX2), and various liver backgrounds (healthy/fibrotic) (n = 5 per group). TBR of the tumor mice models was higher for NIR-II than NIR-I imaging (3.89 ± 1.27 vs. 2.64 ± 0.64, p < 0.05). Moreover, NIR-I/II imaging of fresh tissues from seven patients with HCC undergoing surgery incubated with ICG-FAP-TATA visually provided the spatial distribution heterogeneity of CAFs. The targeted fluorescence was relatively enriched more in the blood flow direction and at the tumor edge, both of which were associated with tumor metastasis (all p < 0.05).

CONCLUSION

This study presents a rapid and effective method for detecting HCC lesions, locating FAP+ CAFs, and visualizing high-risk areas for tumor metastasis at the macroscopic level. It offers a new promising approach with translational potential for imaging HCC.

A new EGFR and c-Met bispecific NIR-II fluorescent probe for visualising colorectal cancer and metastatic lymph nodes

BACKGROUND

The aim of the study was to increase the specificity and targeting of tumour imaging, targeting molecules that enable the simultaneous recognition and binding of multiple tumour-associated receptors. We constructed a NIR-II fluorescence probe based on a bispecific antibody to epidermal growth factor receptor (EGFR) and cellular mesenchymal-epithelial transition factor (c-Met) for visualising colorectal cancers (CRCs) and metastatic lymph nodes.

METHODS

The expression of EGFR and c-Met in tumour and metastatic lymph node specimens from patients with CRC was examined using immunohistochemistry. The EGFR and c-Met bispecific antibody (Rybrevant) was labelled, and its cell-specific binding ability was assessed using laser confocal microscopy. Subcutaneous CRC and orthotopic tumour models were constructed to evaluate the fluorescence imaging of the probe in vivo. To assess the performance of Rybrevant-IRDye800CW in the differential diagnosis of metastatic lymph nodes, a CRC lymph node metastasis model was constructed using human CRC cells implanted in mouse claw pads. Finally, surgically resected CRC tumours and lymph node specimens were incubated with Rybrevant-IRDye800CW for fluorescence NIR-II imaging to evaluate the efficacy of Rybrevant-IRDye800CW for preclinical visualisation.

FINDINGS

The combined expression rate of EGFR and c-Met in CRC and metastatic lymph nodes was significantly higher than the single-target expression rate. The bispecific probe Rybrevant-IRDye800CW was successfully synthesised, and its fluorescence signal could be extended up to 1600 nm using NIR-II imaging. Cell incubation experiments showed that the fluorescence intensity of Rybrevant-IRDye800CW was strongly correlated with EGFR and c-Met overexpression of the cells. NIR-II in vivo fluorescence imaging showed that double-positively expressing subcutaneous tumours significantly uptook Rybrevant-IRDye800CW after tail vein injection of the probe, which rapidly accumulated within the tumours in about 6 h. In EGFR and or c-Met blockade assays, subcutaneous tumours showed weaker uptake of Rybrevant-IRDye800CW. Similarly, Rybrevant-IRDye800CW was specifically identified in orthotopic CRC and lymph node metastasis models, with all orthotopic tumours showing high tumour-to-background ratios in NIR-II imaging. In a NIR-II preclinical study, Rybrevant-IRDye800CW could specifically identify fresh human CRC and its metastatic lymph node tissue.

INTERPRETATION

This study confirmed the complementary EGFR and c-Met expression in CRC and its metastatic lymph nodes. Compared to single-target probes, EGFR and c-Met dual-specific fluorescent probes identified CRC and its metastatic lymph nodes using NIR-II imaging. Thus, NIR-II-guided R0 surgery was performed to resect the CRC and metastatic lymph nodes.

FUNDINGS

This study was supported by the Beijing Natural Science Foundation (Grant numbers: L222054, 7244517, 4232058, L248026, L232020), National Natural Science Foundation of China (NSFC) (92059207, 92359301, 92259303, 62027901, 81930053, 81227901, U21A20386), CAS Youth Interdisciplinary Team (JCTD-2021-08), and the Fundamental Research Funds for the Central Universities (Grant no. JK2024-2-35-02).

GAICN: Graph Attention Iterative Contraction Network for Bioluminescence Tomography

Bioluminescence tomography (BLT) can provide non-invasive quantitative three-dimensional tumor information which has been widely applied in pre-clinical studies. Meanwhile, in recent years, deep learning methods have significantly improved the reconstruction resolution and speed by establishing a non-linear mapping relationship between surface-measured bioluminescence and light source distribution. However, this mapping relationship only works for specific biological tissues and light transmission processes under fixed wavelengths, resulting in poor stability and generalizability. To meet the requirements of diverse practical scenarios and inspired by more effective sparse regularization and graph representation theory, we propose a novel Graph Attention Iterative Contraction Network (GAICN) to conduct a finite element mesh spatial representation study. In the GAICN framework, two learnable spatial topological transforms based on the graph attention mechanism and an iterative contraction activation function were devised to achieve non-local feature aggregation and dynamic adjustment of weights between first-order neighboring nodes in the mesh. As a deep unrolling method, GAICN naturally inherits the coherence of surface bioluminescence with the light source in Forward-Backward Splitting (FBS), thus enhancing the generalizability, stability and interpretability of the network. Both simulation and in-vivo experiments further indicated that GAICN achieved superior reconstruction performance in terms of spatial location, dual light source resolution, stability, generalizability, as well as in-vivo practicability.

TSPE: Reconstruction of multi-morphological tumors of NIR-II fluorescence molecular tomography based on positional encoding

BACKGROUND AND OBJECTIVE

Fluorescence molecular tomography (FMT) is a noninvasive and highly sensitive imaging modality, which can display 3D visualization of tumors by reconstructing fluorescence probes' distribution. However, existing methods mostly ignore positional information, which includes spatial structure information crucial for the reconstruction of light sources. This limits the reconstruction accuracy of light sources with multiple morphologies. Therefore, to our best knowledge, we for the first time integrated positional encoding into the FMT task, enabling the incorporation of high-frequency spatial structure information.

METHODS

We proposed a three-stage network embedded with a positional encoding module (TSPE) to perform high reconstruction accuracy of tumors with multiple morphologies. Additionally, our study focused on NIR-II which had less severe scattering problems and higher imaging accuracy than NIR-I.

RESULTS

The simulation experiments demonstrated that TSPE achieved high reconstruction accuracy in NIR-II FMT, with the barycenter error (BCE) for single-tumor reconstruction reaching 0.18 mm, representing a 14 % reduction compared to other methods. TSPE more accurately distinguished adjacent multi-morphological tumors with a minimal edge-to-edge distance (EED) of 0.3 mm. In vivo experiments also showed that TSPE could achieve more accurate reconstruction of tumors compared with other methods.

CONCLUSIONS

The proposed method can achieve the best reconstruction performance. It has potential to promote the development of NIR-II FMT and its preclinical application.

NIR-II-excited off-on-off fluorescent nanoprobes for sensitive molecular imaging in vivo

Strong background interference signals from normal tissues have significantly compromised the sensitive fluorescence imaging of early disease tissues with exogenous probes in vivo, particularly for sensitive fluorescence imaging of early liver disease due to the liver's significant uptake and accumulation of exogenous nanoprobes, coupled with high tissue autofluorescence and deep tissue depth. As a proof-of-concept study, we herein report a near-infrared-II (NIR-II, 1.0-1.7 μm) light-excited "off-on-off" NIR-II fluorescent probe (NDP). It has near-ideal zero initial probe fluorescence but can turn on its NIR-II fluorescence in liver cancer tissues and then turn off the fluorescence again upon migration from cancer to normal tissues to minimize background interference. Due to its low background, a blind study employing our probes could identify female mice with orthotopic liver tumors with 100% accuracy from mixed subjects of healthy and tumor mice, and implemented sensitive locating of early orthotopic liver tumors with sizes as small as 4 mm. Our NIR-II-excited "off-on-off" probe design concept not only provides a promising molecular design guideline for sensitive imaging of early liver cancer but also could be generalized for sensitive imaging of other early disease lesions.

Comparative Study of Indocyanine Green Fluorescence Imaging in Lung Cancer with Near-Infrared-I/II Windows

BACKGROUND

We compare the application of intravenous indocyanine green (ICG) fluorescence imaging in lung cancer with near-infrared-I (NIR-I) and near-infrared-II (NIR-II) windows.

METHODS

From March to December 2022, we enrolled patients who received an intravenous injection of ICG (5 mg/kg) 1 day before the planned lung cancer surgery. The lung cancer nodules were imaged by NIR-I/II fluorescence imaging systems, and the tumor-to-normal-tissue ratio (TNR) was calculated. In addition, the fluorescence intensity and signal-to-background ratio (SBR) of capillary glass tubes containing ICG covered with different thicknesses of lung tissue were measured by NIR-I/II fluorescence imaging systems.

RESULTS

In this study, 102 patients were enrolled, and the mean age was 59.9 ± 9.2 years. A total of 96 (94.1%) and 98 (96.1%) lung nodules were successfully imaged with NIR-I and NIR-II fluorescence, and the TNR of NIR-II was significantly higher than that of NIR-I (3.9 ± 1.3 versus 2.4 ± 0.6, P < 0.001). In multiple linear regression, solid nodules (P < 0.001) and squamous cell carcinoma (P < 0.001) were independent predictors of a higher TNR of NIR-I/II. When capillary glass tubes were covered with lung tissue whose thickness was more than 2 mm, the fluorescence intensity and the SBR of NIR-II were significantly higher than those of NIR-I.

CONCLUSIONS

We verified the feasibility of NIR-II fluorescence imaging in intravenous ICG lung cancer imaging for the first time. NIR-II fluorescence can improve the TNR and penetration depth of lung cancer with promising clinical prospects.

Highly robust reconstruction framework for three-dimensional optical imaging based on physical model constrained neural networks

Objective. The reconstruction of three-dimensional optical imaging that can quantitatively acquire the target distribution from surface measurements is a serious ill-posed problem. Traditional regularization-based reconstruction can solve such ill-posed problem to a certain extent, but its accuracy is highly dependent ona priorinformation, resulting in a less stable and adaptable method. Data-driven deep learning-based reconstruction avoids the errors of light propagation models and the reliance on experience and a prior by learning the mapping relationship between the surface light distribution and the target directly from the dataset. However, the acquisition of the training dataset and the training of the network itself are time consuming, and the high dependence of the network performance on the training dataset results in a low generalization ability. The objective of this work is to develop a highly robust reconstruction framework to solve the existing problems.Approach. This paper proposes a physical model constrained neural networks-based reconstruction framework. In the framework, the neural networks are to generate a target distribution from surface measurements, while the physical model is used to calculate the surface light distribution based on this target distribution. The mean square error between the calculated surface light distribution and the surface measurements is then used as a loss function to optimize the neural network. To further reduce the dependence ona prioriinformation, a movable region is randomly selected and then traverses the entire solution interval. We reconstruct the target distribution in this movable region and the results are used as the basis for its next movement.Main Results. The performance of the proposed framework is evaluated with a series of simulations andin vivoexperiment, including accuracy robustness of different target distributions, noise immunity, depth robustness, and spatial resolution. The results collectively demonstrate that the framework can reconstruct targets with a high accuracy, stability and versatility.Significance. The proposed framework has high accuracy and robustness, as well as good generalizability. Compared with traditional regularization-based reconstruction methods, it eliminates the need to manually delineate feasible regions and adjust regularization parameters. Compared with emerging deep learning assisted methods, it does not require any training dataset, thus saving a lot of time and resources and solving the problem of poor generalization and robustness of deep learning methods. Thus, the framework opens up a new perspective for the reconstruction of three-dimension optical imaging.

Logarithmic total variation regularization via preconditioned conjugate gradient method for sparse reconstruction of bioluminescence tomography

BACKGROUND AND OBJECTIVE

Bioluminescence Tomography (BLT) is a powerful optical molecular imaging technique that enables the noninvasive investigation of dynamic biological phenomena. It aims to reconstruct the three-dimensional spatial distribution of bioluminescent sources from optical measurements collected on the surface of the imaged object. However, BLT reconstruction is a challenging ill-posed problem due to the scattering effect of light and the limitations in detecting surface photons, which makes it difficult for existing methods to achieve satisfactory reconstruction results. In this study, we propose a novel method for sparse reconstruction of BLT based on a preconditioned conjugate gradient with logarithmic total variation regularization (PCG-logTV).

METHOD

This PCG-logTV method incorporates the sparsity of overlapping groups and enhances the sparse structure of these groups using logarithmic functions, which can preserve edge features and achieve more stable reconstruction results in BLT. To accelerate the convergence of the algorithm solution, we use the preconditioned conjugate gradient iteration method on the objective function and obtain the reconstruction results. We demonstrate the performance of our proposed method through numerical simulations and in vivo experiment.

RESULTS AND CONCLUSIONS

The results show that the PCG-logTV method obtains the most accurate reconstruction results, and the minimum position error (LE) is 0.254mm, which is 26%, 31% and 34% of the FISTA (0.961), IVTCG (0.81) and L1-TV (0.739) methods, and the root mean square error (RMSE) and relative intensity error (RIE) are the smallest, indicating that it is closest to the real light source. In addition, compared with the other three methods, the PCG-logTV method also has the highest DICE similarity coefficient, which is 0.928, which means that this method can effectively reconstruct the three-dimensional spatial distribution of bioluminescent light sources, has higher resolution and robustness, and is beneficial to the preclinical and clinical studies of BLT.

NIR-II fluorescence-guided liver cancer surgery by a small molecular HDAC6 targeting probe

BACKGROUND

Hepatocellular carcinoma (HCC) is the sixth most common malignancy globally and ranks third in terms of both mortality and incidence rates. Surgical resection holds potential as a curative approach for HCC. However, the residual disease contributes to a high 5-year recurrence rate of 70%. Due to their excellent specificity and optical properties, fluorescence-targeted probes are deemed effective auxiliary tools for addressing residual lesions, enabling precise surgical diagnosis and treatment. Research indicates histone deacetylase 6 (HDAC6) overexpression in HCC cells, making it a potential imaging biomarker. This study designed a targeted small-molecule fluorescent probe, SeCF3-IRDye800cw (SeCF3-IRD800), operating within the Second near-infrared window (NIR-II, 1000-1700 nm). The study confirms the biocompatibility of SeCF3-IRD800 and proceeds to demonstrate its applications in imaging in vivo, fluorescence-guided surgery (FGS) for liver cancer, liver fibrosis imaging, and clinical samples incubation, thereby preliminarily validating its utility in liver cancer.

METHODS

SeCF3-IRD800 was synthesized by combining the near-infrared fluorescent dye IRDye800cw-NHS with an improved HDAC6 inhibitor. Initially, a HepG2-Luc subcutaneous tumor model (n = 12) was constructed to investigate the metabolic differences between SeCF3-IRD800 and ICG in vivo. Subsequently, HepG2-Luc (n = 12) and HCCLM3-Luc (n = 6) subcutaneous xenograft mouse models were used to assess in vivo targeting by SeCF3-IRD800. The HepG2-Luc orthotopic liver cancer model (n = 6) was employed to showcase the application of SeCF3-IRD800 in FGS. Liver fibrosis (n = 6) and HepG2-Luc orthotopic (n = 6) model imaging results were used to evaluate the impact of different pathological backgrounds on SeCF3-IRD800 imaging. Three groups of fresh HCC and normal liver samples from patients with liver cancer were utilized for SeCF3-IRD800 incubation ex vivo, while preclinical experiments illustrated its potential for clinical application.

FINDINGS

The HDAC6 inhibitor 6 (SeCF3) modified with trifluoromethyl was labeled with IRDy800CW-NHS to synthesize the small-molecule targeted probe SeCF3-IRD800, with NIR-II fluorescence signals. SeCF3-IRD800 was rapidly metabolized by the kidneys and exhibited excellent biocompatibility. In vivo validation demonstrated that SeCF3-IRD800 achieved optimal imaging within 8 h, displaying high tumor fluorescence intensity (7658.41 ± 933.34) and high tumor-to-background ratio (5.20 ± 1.04). Imaging experiments with various expression levels revealed its capacity for HDAC6-specific targeting across multiple HCC tumor models, suitable for NIR-II intraoperative imaging. Fluorescence-guided surgery experiments were found feasible and capable of detecting sub-visible 2 mm tumor lesions under white light, aiding surgical decision-making. Further imaging of liver fibrosis mice showed that SeCF3-IRD800's imaging efficacy remained unaffected by liver pathological conditions. Correlations were observed between HDAC6 expression levels and corresponding fluorescence intensity (R2 = 0.8124) among normal liver, liver fibrosis, and HCC tissues. SeCF3-IRD800 identified HDAC6-positive samples from patients with HCC, holding advantages for perspective intraoperative identification in liver cancer. Thus, the rapidly metabolized HDAC6-targeted small-molecule NIR-II fluorescence probe SeCF3-IRD800 holds significant clinical translational value.

INTERPRETATION

The successful application of NIR-II fluorescence-guided surgery in liver cancer indicates that SeCF3-IRD800 has great potential to improve the clinical diagnosis and treatment of liver cancer, and could be used as an auxiliary tool for surgical treatment of liver cancer without being affected by liver pathology.

FUNDING

This paper is supported by the National Natural Science Foundation of China (NSFC) (92,059,207, 62,027,901, 81,930,053, 81,227,901, 82,272,105, U21A20386 and 81,971,773), CAS Youth Interdisciplinary Team (JCTD-2021-08), the Zhuhai High-level Health Personnel Team Project (Zhuhai HLHPTP201703), and Guangdong Basic and Applied Basic Research Foundation under Grant No. 2022A1515011244.

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.

X-ray Cherenkov-luminescence tomography reconstruction with a three-component deep learning algorithm: Swin transformer, convolutional neural network, and locality module

SIGNIFICANCE

X-ray Cherenkov-luminescence tomography (XCLT) produces fast emission data from megavoltage (MV) x-ray scanning, in which the excitation location of molecules within tissue is reconstructed. However standard filtered backprojection (FBP) algorithms for XCLT sinogram reconstruction can suffer from insufficient data due to dose limitations, so there are limits in the reconstruction quality with some artifacts. We report a deep learning algorithm for XCLT with high image quality and improved quantitative accuracy.

AIM

To directly reconstruct the distribution of emission quantum yield for x-ray Cherenkov-luminescence tomography, we proposed a three-component deep learning algorithm that includes a Swin transformer, convolution neural network, and locality module model.

APPROACH

A data-to-image model x-ray Cherenkov-luminescence tomography is developed based on a Swin transformer, which is used to extract pixel-level prior information from the sinogram domain. Meanwhile, a convolutional neural network structure is deployed to transform the extracted pixel information from the sinogram domain to the image domain. Finally, a locality module is designed between the encoder and decoder connection structures for delivering features. Its performance was validated with simulation, physical phantom, and in vivo experiments.

RESULTS

This approach can better deal with the limits to data than conventional FBP methods. The method was validated with numerical and physical phantom experiments, with results showing that it improved the reconstruction performance mean square error ( > 94.1 % ), peak signal-to-noise ratio ( > 41.7 % ), and Pearson correlation ( > 19 % ) compared with the FBP algorithm. The Swin-CNN also achieved a 32.1% improvement in PSNR over the deep learning method AUTOMAP.

CONCLUSIONS

This study shows that the three-component deep learning algorithm provides an effective reconstruction method for x-ray Cherenkov-luminescence tomography.

Excitation-based fully connected network for precise NIR-II fluorescence molecular tomography

Fluorescence molecular tomography (FMT) is a novel imaging modality to obtain fluorescence biomarkers' three-dimensional (3D) distribution. However, the simplified mathematical model and complicated inverse problem limit it to achieving precise results. In this study, the second near-infrared (NIR-II) fluorescence imaging was adopted to mitigate tissue scattering and reduce noise interference. An excitation-based fully connected network was proposed to model the inverse process of NIR-II photon propagation and directly obtain the 3D distribution of the light source. An excitation block was embedded in the network allowing it to autonomously pay more attention to neurons related to the light source. The barycenter error was added to the loss function to improve the localization accuracy of the light source. Both numerical simulation and in vivo experiments showed the superiority of the novel NIR-II FMT reconstruction strategy over the baseline methods. This strategy was expected to facilitate the application of machine learning in biomedical research.

A novel weighted auxiliary set matching pursuit method for glioma in Cerenkov luminescence tomography reconstruction

Cerenkov luminescence tomography (CLT) is a promising three-dimensional imaging technology that has been actively investigated in preclinical studies. However, because of the ill-posedness in the inverse problem of CLT reconstruction, the reconstruction performance is still not satisfactory for broad biomedical applications. In this study, a novel weighted auxiliary set matching pursuit (WASMP) method was explored to enhance the accuracy of CLT reconstruction. The numerical simulations and in vivo imaging studies using tumor-bearing mice models were conducted to evaluate the performance of the WASMP method. The results of the above experiments proved that the WASMP method achieved superior reconstruction performance than other approaches in terms of positional accuracy and shape recovery. It further demonstrates that the atom selection strategy proposed in this study has a positive effect on improving the accuracy of atoms. The proposed WASMP improves the accuracy for CLT reconstruction for biomedical applications.

Elastic net-based non-negative iterative three-operator splitting strategy for Cerenkov luminescence tomography

Cerenkov luminescence tomography (CLT) provides a powerful optical molecular imaging technique for non-invasive detection and visualization of radiopharmaceuticals in living objects. However, the severe photon scattering effect causes ill-posedness of the inverse problem, and the location accuracy and shape recovery of CLT reconstruction results are unsatisfactory for clinical application. Here, to improve the reconstruction spatial location accuracy and shape recovery ability, a non-negative iterative three operator splitting (NNITOS) strategy based on elastic net (EN) regularization was proposed. NNITOS formalizes the CLT reconstruction as a non-convex optimization problem and splits it into three operators, the least square, L_1/2-norm regularization, and adaptive grouping manifold learning, then iteratively solved them. After stepwise iterations, the result of NNITOS converged progressively. Meanwhile, to speed up the convergence and ensure the sparsity of the solution, shrinking the region of interest was utilized in this strategy. To verify the effectiveness of the method, numerical simulations and in vivo experiments were performed. The result of these experiments demonstrated that, compared to several methods, NNITOS can achieve superior performance in terms of location accuracy, shape recovery capability, and robustness. We hope this work can accelerate the clinical application of CLT in the future.

Application of Cherenkov radiation in tumor imaging and treatment

Cherenkov radiation (CR) is the characteristic blue glow that is generated during radiotherapy or radioisotope decay. Its distribution and intensity naturally reflect the actual dose and field of radiotherapy and the location of radioisotope imaging agents in vivo. Therefore, CR can represent a potential in situ light source for radiotherapy monitoring and radioisotope-based tumor imaging. When used in combination with new imaging techniques, molecular probes or nanomedicine, CR imaging exhibits unique advantages (accuracy, low cost, convenience and fast) in tumor radiotherapy monitoring and imaging. Furthermore, photosensitive nanomaterials can be used for CR photodynamic therapy, providing new approaches for integrating tumor imaging and treatment. Here the authors review the latest developments in the use of CR in tumor research and discuss current challenges and new directions for future studies.

Cherenkov Luminescence in Tumor Diagnosis and Treatment: A Review

Malignant tumors rank as a leading cause of death worldwide. Accurate diagnosis and advanced treatment options are crucial to win battle against tumors. In recent years, Cherenkov luminescence (CL) has shown its technical advantages and clinical transformation potential in many important fields, particularly in tumor diagnosis and treatment, such as tumor detection in vivo, surgical navigation, radiotherapy, photodynamic therapy, and the evaluation of therapeutic effect. In this review, we summarize the advances in CL for tumor diagnosis and treatment. We first describe the physical principles of CL and discuss the imaging techniques used in tumor diagnosis, including CL imaging, CL endoscope, and CL tomography. Then we present a broad overview of the current status of surgical resection, radiotherapy, photodynamic therapy, and tumor microenvironment monitoring using CL. Finally, we shed light on the challenges and possible solutions for tumor diagnosis and therapy using CL.

VoxDMRN: a voxelwise deep max-pooling residual network for bioluminescence tomography reconstruction

Bioluminescence tomography (BLT) has extensive applications in preclinical studies for cancer research and drug development. However, the spatial resolution of BLT is inadequate because the numerical methods are limited for solving the physical models of photon propagation and the restriction of using tetrahedral meshes for reconstruction. We conducted a series of theoretical derivations and divided the BLT reconstruction process into two steps: feature extraction and nonlinear mapping. Inspired by deep learning, a voxelwise deep max-pooling residual network (VoxDMRN) is proposed to establish the nonlinear relationship between the internal bioluminescent source and surface boundary density to improve the spatial resolution in BLT reconstruction. The numerical simulation and in vivo experiments both demonstrated that VoxDMRN greatly improves the reconstruction performance regarding location accuracy, shape recovery capability, dual-source resolution, robustness, and in vivo practicability.

Total Variation Constrained Graph Manifold Learning Strategy for Cerenkov Luminescence Tomography

Harnessing the power and flexibility of radiolabeled molecules, Cerenkov luminescence tomography (CLT) provides a novel technique for non-invasive visualisation and quantification of viable tumour cells in a living organism. However, owing to the photon scattering effect and the ill-posed inverse problem, CLT still suffers from insufficient spatial resolution and shape recovery in various preclinical applications. In this study, we proposed a total variation constrained graph manifold learning (TV-GML) strategy for achieving accurate spatial location, dual-source resolution, and tumour morphology. TV-GML integrates the isotropic total variation term and dynamic graph Laplacian constraint to make a trade-off between edge preservation and piecewise smooth region reconstruction. Meanwhile, the tetrahedral mesh-Cartesian grid pair method based on the k-nearest neighbour, and the adaptive and composite Barzilai-Borwein method, were proposed to ensure global super linear convergence of the solution of TV-GML. The comparison results of both simulation experiments and in vivo experiments further indicated that TV-GML achieved superior reconstruction performance in terms of location accuracy, dual-source resolution, shape recovery capability, robustness, and in vivo practicability. Significance: We believe that this novel method will be beneficial to the application of CLT for quantitative analysis and morphological observation of various preclinical applications and facilitate the development of the theory of solving inverse problem.

Sensitivity improved Cerenkov luminescence endoscopy using optimal system parameters

Background

The challenges of clinical translation of optical imaging, including the limited availability of clinically used imaging probes and the restricted penetration depth of light propagation in tissues can be avoided using Cerenkov luminescence endoscopy (CLE). However, the clinical applications of CLE are limited due to the low signal level of Cerenkov luminescence and the large transmission loss caused by the endoscope, which results in a relatively low detection sensitivity of current CLE. The aim of this study was to enhance the detection sensitivity of the CLE system and thus improve the system for clinical application in the detection of gastrointestinal diseases.

Methods

Four optical fiber endoscopes were customized with different system parameters, including monofilament (MF) diameter of imaging fiber bundles, fiber material, probe coating, etc. The endoscopes were connected to the detector via a specifically designed straight connection device to form the CLE system. The β-2-[¹⁸F]-Fluoro-2-deoxy-D-glucose (¹⁸F-FDG) solution and the radionuclide of Gallium-68 (⁶⁸Ga) were used to evaluate the performance of the CLE system. The images of the ¹⁸F-FDG solution acquired by the CLE were used to optimize imaging parameters of the system. By using the endoscope with optimized parameters, including the MF diameter of imaging fiber bundles, fiber materials, etc., the resolution and sensitivity of the assembled CLE system were measured by imaging the radionuclide of ⁶⁸Ga.

Results

The results of ¹⁸F-FDG experiments showed that larger MF diameter led to higher collection efficiency. The fiber material and probe coating with high transmission ratios in the range of 400-900 nm also increased signal collection and transmission efficiency. The results of ⁶⁸Ga evaluations showed that a minimum radioactive activity of radionuclides as low as 0.03 µCi was detected in vitro within 5 minutes, while that of 0.68 µCi can be detected within 1 minute. In vivo experiments also demonstrated that the developed CLE system achieved a high sensitivity at a submicrocurie level; that is, 0.44 µCi within 5 minutes, and 0.83 µCi within 1 minute. The weaker in vivo sensitivity was due to the attenuation of the signal by the mouse tissue skin and the autofluorescence interference produced by biological tissues.

Conclusions

By optimizing the structural parameters of fiber endoscope and imaging parameters for data acquisition, we developed a CLE system with a sensitivity at submicrocurie level. These results support the possibility that this technology can clinically detect early tumors within 1 minute.

Attention mechanism-based locally connected network for accurate and stable reconstruction in Cerenkov luminescence tomography

Cerenkov luminescence tomography (CLT) is a novel and highly sensitive imaging technique, which could obtain the three-dimensional distribution of radioactive probes to achieve accurate tumor detection. However, the simplified radiative transfer equation and ill-conditioned inverse problem cause a reconstruction error. In this study, a novel attention mechanism based locally connected (AMLC) network was proposed to reduce barycenter error and improve morphological restorability. The proposed AMLC network consisted of two main parts: a fully connected sub-network for providing a coarse reconstruction result, and a locally connected sub-network based on an attention matrix for refinement. Both numerical simulations and in vivo experiments were conducted to show the superiority of the AMLC network in accuracy and stability over existing methods (MFCNN, KNN-LC network). This method improved CLT reconstruction performance and promoted the application of machine learning in optical imaging research.

A Multilevel Probabilistic Cerenkov Luminescence Tomography Reconstruction Framework Based on Energy Distribution Density Region Scaling

Cerenkov luminescence tomography (CLT) is a promising non-invasive optical imaging method with three-dimensional semiquantitative in vivo imaging capability. However, CLT itself relies on Cerenkov radiation, a low-intensity radiation, making CLT reconstruction more challenging than other imaging modalities. In order to solve the ill-posed inverse problem of CLT imaging, some numerical optimization or regularization methods need to be applied. However, in commonly used methods for solving inverse problems, parameter selection significantly influences the results. Therefore, this paper proposed a probabilistic energy distribution density region scaling (P-EDDRS) framework. In this framework, multiple reconstruction iterations are performed, and the Cerenkov source distribution of each reconstruction is treated as random variables. According to the spatial energy distribution density, the new region of interest (ROI) is solved. The size of the region required for the next operation was determined dynamically by combining the intensity characteristics. In addition, each reconstruction source distribution is given a probability weight value, and the prior probability in the subsequent reconstruction is refreshed. Last, all the reconstruction source distributions are weighted with the corresponding probability weights to get the final Cerenkov source distribution. To evaluate the performance of the P-EDDRS framework in CLT, this article performed numerical simulation, in vivo pseudotumor model mouse experiment, and breast cancer mouse experiment. Experimental results show that this reconstruction framework has better positioning accuracy and shape recovery ability and can optimize the reconstruction effect of multiple algorithms on CLT.

Prior Compensation Algorithm for Cerenkov Luminescence Tomography From Single-View Measurements

Cerenkov luminescence tomography (CLT) has attracted much attention because of the wide clinically-used probes and three-dimensional (3D) quantification ability. However, due to the serious morbidity of 3D optical imaging, the reconstructed images of CLT are not appreciable, especially when single-view measurements are used. Single-view CLT improves the efficiency of data acquisition. It is much consistent with the actual imaging environment of using commercial imaging system, but bringing the problem that the reconstructed results will be closer to the animal surface on the side where the single-view image is collected. To avoid this problem to the greatest extent possible, we proposed a prior compensation algorithm for CLT reconstruction based on depth calibration strategy. This method takes full account of the fact that the attenuation of light in the tissue will depend heavily on the depth of the light source as well as the distance between the light source and the detection plane. Based on this consideration, a depth calibration matrix was designed to calibrate the attenuation between the surface light flux and the density of the internal light source. The feature of the algorithm was that the depth calibration matrix directly acts on the system matrix of CLT reconstruction, rather than modifying the regularization penalty items. The validity and effectiveness of the proposed algorithm were evaluated with a numerical simulation and a mouse-based experiment, whose results illustrated that it located the radiation sources accurately by using single-view measurements.

Noninvasive multimodal fluorescence and magnetic resonance imaging of whole-organ intervertebral discs

Low back pain (LBP) is a commonly experienced symptom posing a tremendous healthcare burden to individuals and society at large. The LBP pathology is strongly linked to degeneration of the intervertebral disc (IVD), calling for development of early-stage diagnostic tools for visualizing biomolecular changes in IVD. Multimodal measurements of fluorescence molecular tomography (FMT) and magnetic resonance imaging (MRI) were performed on IVD whole organ culture model using an in-house built FMT system and a high-field MRI scanner. The resulted multimodal images were systematically validated through epifluorescence imaging of the IVD sections at a microscopic level. Multiple image contrasts were exploited, including fluorescence distribution, anatomical map associated with T1-weighted MRI contrast, and water content related with T2 relaxation time. The developed multimodality imaging approach may thus serve as a new assessment tool for early diagnosis of IVD degeneration and longitudinal monitoring of IVD organ culture status using fluorescence markers.

Development of tomographic reconstruction for three-dimensional optical imaging: From the inversion of light propagation to artificial intelligence

Optical molecular tomography (OMT) is an imaging modality which uses an optical signal, especially near-infrared light, to reconstruct the three-dimensional information of the light source in biological tissue. With the advantages of being low-cost, noninvasive and having high sensitivity, OMT has been applied in preclinical and clinical research. However, due to its serious ill-posedness and ill-condition, the solution of OMT requires heavy data analysis and the reconstruction quality is limited. Recently, the artificial intelligence (commonly known as AI)-based methods have been proposed to provide a different tool to solve the OMT problem. In this paper, we review the progress on OMT algorithms, from conventional methods to AI-based methods, and we also give a prospective towards future developments in this domain.