Journal of Biomedical Optics最新文献

Significance: Histopathological examination of surgical biopsies, such as in glioma and glioblastoma resection, is hindered in current clinical practice by the long time required for the laboratory analysis and pathological screening, typically taking several days or even weeks to be completed.

Aim: We propose here a transportable, high-density, spectral scanning-based hyperspectral imaging (HSI) setup, named HyperProbe1, that can provide in situ, fast biochemical analysis, and mapping of fresh surgical tissue samples, right after excision, and without the need for fixing, staining nor compromising the integrity of the tissue properties.

Approach: HyperProbe1 is based on spectral scanning via supercontinuum laser illumination filtered with acousto-optic tunable filters. Such methodology allows the user to select any number and type of wavelength bands in the visible and near-infrared range between 510 and 900 nm (up to a maximum of 79) and to reconstruct 3D hypercubes composed of high-resolution (4 to $5\mu m$ ), widefield images ( $0.9\times 0.9{\mathrm{mm}}^2$ ) of the surgical samples, where each pixel is associated with a complete spectrum.

Results: The HyperProbe1 setup is here presented and characterized. The system is applied to 11 fresh surgical biopsies of glioma from routine patients, including different grades of tumor classification. Quantitative analysis of the composition of the tissue is performed via fast spectral unmixing to reconstruct the mapping of major biomarkers, such as oxy-( ${\mathrm{HbO}}_2$ ) and deoxyhemoglobin (HHb), as well as cytochrome-c-oxidase (CCO). We also provided a preliminary attempt to infer tumor classification based on differences in composition in the samples, suggesting the possibility of using lipid content and differential CCO concentrations to distinguish between lower and higher-grade gliomas.

Conclusions: A proof of concept of the performances of HyperProbe1 for quantitative, biochemical mapping of surgical biopsies is demonstrated, paving the way for improving current post-surgical, histopathological practice via non-destructive, in situ streamlined screening of fresh tissue samples in a matter of minutes after excision.

Significance: Hyperspectral imaging (HSI) of murine tumor models grown in dorsal skinfold window chambers (DSWCs) offers invaluable insight into the tumor microenvironment. However, light loss in a glass coverslip is often overlooked, and particular tissue characteristics are improperly modeled, leading to errors in tissue properties extracted from hyperspectral images.

Aim: We highlight the significance of spectral renormalization in HSI of DSWC models and demonstrate the benefit of incorporating enhanced green fluorescent protein (EGFP) excitation and emission in the skin tissue model for tumors expressing genes to produce EGFP.

Approach: We employed an HSI system for intravital imaging of mice with 4T1 mammary carcinoma in a DSWC over 14 days. We performed spectral renormalization of hyperspectral images based on the measured reflectance spectra of glass coverslips and utilized an inverse adding-doubling (IAD) algorithm with a two-layer murine skin model, to extract tissue parameters, such as total hemoglobin concentration and tissue oxygenation ( ${\mathrm{StO}}_2$ ). The model was upgraded to consider EGFP fluorescence excitation and emission. Moreover, we conducted additional experiments involving tissue phantoms, human forearm skin imaging, and numerical simulations.

Results: Hyperspectral image renormalization and the addition of EGFP fluorescence in the murine skin model reduced the mean absolute percentage errors (MAPEs) of fitted and measured spectra by up to 10% in tissue phantoms, 0.55% to 1.5% in the human forearm experiment and numerical simulations, and up to 0.7% in 4T1 tumors. Similarly, the MAPEs for tissue parameters extracted by IAD were reduced by up to 3% in human forearms and numerical simulations. For some parameters, statistically significant differences ( $p<0.05$ ) were observed in 4T1 tumors. Ultimately, we have shown that fluorescence emission could be helpful for 4T1 tumor segmentation.

Conclusions: The results contribute to improving intravital monitoring of DWSC models using HSI and pave the way for more accurate and precise quantitative imaging.

Significance: Hyperspectral imaging (HSI) is an emerging imaging modality for oncological applications and can improve cancer detection with digital pathology.

Aim: The study aims to highlight the increased accuracy and sensitivity of detecting the margin of thyroid carcinoma in hematoxylin and eosin (H&E)-stained histological slides using HSI and data augmentation methods.

Approach: Using an automated microscopic imaging system, we captured 2599 hyperspectral images from 65 H&E-stained human thyroid slides. Images were then preprocessed into 153,906 image patches of dimension $250\times 250\times 84\text{\hspace{0.17em}\hspace{0.17em}pixels}$ . We modified the TimeSformer network architecture, which used alternating spectral attention and spatial attention layers. We implemented several data augmentation methods for HSI based on the RandAugment algorithm. We compared the performances of TimeSformer on HSI against the performances of pretrained ConvNext and pretrained vision transformers (ViT) networks on red, green, and blue (RGB) images. Finally, we applied attention unrolling techniques on the trained TimeSformer network to identify the biological features to which the network paid attention.

Results: In the testing dataset, TimeSformer achieved an accuracy of 90.87%, a weighted $F_1$ score of 89.79%, a sensitivity of 91.50%, and an area under the receiving operator characteristic curve (AU-ROC) score of 97.04%. Additionally, TimeSformer produced thyroid carcinoma tumor margins with an average Jaccard score of 0.76 mm. Without data augmentation, TimeSformer achieved an accuracy of 88.23%, a weighted $F_1$ score of 86.46%, a sensitivity of 85.53%, and an AU-ROC score of 94.94%. In comparison, the ViT network achieved an 89.98% accuracy, an 88.14% weighted $F_1$ score, an 84.77% sensitivity, and a 96.17% AU-ROC. Our visualization results showed that the network paid attention to biological features.

Conclusions: The TimeSformer model trained with hyperspectral histological data consistently outperformed conventional RGB-based models, highlighting the superiority of HSI in this context. Our proposed augmentation methods improved the accuracy, the $F_1$ score, and the sensitivity score.

Significance: Hyperspectral cameras capture spectral information at each pixel in an image. Acquired spectra can be analyzed to estimate quantities of absorbing and scattering components, but the use of traditional fitting algorithms over megapixel images can be computationally intensive. Deep learning algorithms can be trained to rapidly analyze spectral data and can potentially process hyperspectral camera data in real time.

Aim: A hyperspectral camera was used to capture $1216\times 1936\text{\hspace{0.17em}\hspace{0.17em}pixel}$ wide-field reflectance images of in vivo human tissue at 205 wavelength bands from 420 to 830 nm.

Approach: The optical properties of oxyhemoglobin, deoxyhemoglobin, melanin, and scattering were used with multi-layer Monte Carlo models to generate simulated diffuse reflectance spectra for 24,000 random combinations of physiologically relevant tissue components. These spectra were then used to train an artificial neural network (ANN) to predict tissue component concentrations from an input reflectance spectrum.

Results: The ANN achieved low root mean square errors in a test set of 6000 independent simulated diffuse reflectance spectra while calculating concentration values more than 4000× faster than a conventional iterative least squares approach.

Conclusions: In vivo finger occlusion and gingival abrasion studies demonstrate the ability of this approach to rapidly generate high-resolution images of tissue component concentrations from a hyperspectral dataset acquired from human subjects.

Significance: Optical properties of biological tissues, such as refractive index (RI), are fundamental properties, intrinsically linked to the tissue's composition and structure. We hypothesize that, as the RI and the functional properties of articular cartilage (AC) are dependent on the tissue's structure and composition, the RI of AC is related to its biomechanical properties.

Aim: This study aims to investigate the relationship between RI of human AC and its biomechanical properties.

Approach: Human cartilage samples ( $n=22$ ) were extracted from the right knee joint of three cadaver donors (one female, aged 47 years, and two males, aged 64 and 68 years) obtained from a commercial biobank (Science Care, Phoenix, Arizona, United States). The samples were initially subjected to mechanical indentation testing to determine elastic [equilibrium modulus (EM) and instantaneous modulus (IM)] and dynamic [dynamic modulus (DM)] viscoelastic properties. An Abbemat 3200 automatic one-wavelength refractometer operating at 600 nm was used to measure the RI of the extracted sections. Similarly, Spearman's and Pearson's correlation coefficients were employed for non-normal and normal datasets, respectively, to determine the correlation between the depth-wise RI and biomechanical properties of the cartilage samples as a function of the collagen fibril orientation.

Results: A positive correlation with statistically significant relations ( $p-\text{values}<0.05$ ) was observed between the RI and the biomechanical properties (EM, IM, and DM) along the tissue depth for each zone, e.g., superficial, middle, and deep zones. Likewise, a lower positive correlation with statistically significant relations ( $p-\text{values}<0.05$ ) was also observed for collagen fibril orientation of all zones with the biomechanical properties.

Conclusions: The results indicate that, although the RI exhibits different levels of correlation with different biomechanical properties, the relationship varies as a function of the tissue depth. This knowledge paves the way for optically monitoring changes in AC biomechanical properties nondestructively via changes in the RI. Thus, the RI could be a potential biomarker for assessing the mechanical competency of AC, particularly in degenerative diseases, such as osteoarthritis.

Significance: Hyperspectral imaging sensors have rapidly advanced, aiding in tumor diagnostics for in vivo brain tumors. Linescan cameras effectively distinguish between pathological and healthy tissue, whereas snapshot cameras offer a potential alternative to reduce acquisition time.

Aim: Our research compares linescan and snapshot hyperspectral cameras for in vivo brain tissues and chromophore identification.

Approach: We compared a linescan pushbroom camera and a snapshot camera using images from 10 patients with various pathologies. Objective comparisons were made using unnormalized and normalized data for healthy and pathological tissues. We utilized the interquartile range (IQR) for the spectral angle mapping (SAM), the goodness-of-fit coefficient (GFC), and the root mean square error (RMSE) within the 659.95 to 951.42 nm range. In addition, we assessed the ability of both cameras to capture tissue chromophores by analyzing absorbance from reflectance information.

Results: The SAM metric indicates reduced dispersion and high similarity between cameras for pathological samples, with a 9.68% IQR for normalized data compared with 2.38% for unnormalized data. This pattern is consistent across GFC and RMSE metrics, regardless of tissue type. Moreover, both cameras could identify absorption peaks of certain chromophores. For instance, using the absorbance measurements of the linescan camera, we obtained SAM values below 0.235 for four peaks, regardless of the tissue and type of data under inspection. These peaks are one for cytochrome b in its oxidized form at $\lambda =422\mathrm{nm}$ , two for ${\mathrm{HbO}}_2$ at $\lambda =542\mathrm{nm}$ and $\lambda =576\mathrm{nm}$ , and one for water at $\lambda =976\mathrm{nm}$ .

Conclusion: The spectral signatures of the cameras show more similarity with unnormalized data, likely due to snapshot sensor noise, resulting in noisier signatures post-normalization. Comparisons in this study suggest that snapshot cameras might be viable alternatives to linescan cameras for real-time brain tissue identification.

Significance: Label-free multimodal imaging methods that can provide complementary structural and chemical information from the same sample are critical for comprehensive tissue analyses. These methods are specifically needed to study the complex tumor-microenvironment where fibrillar collagen's architectural changes are associated with cancer progression. To address this need, we present a multimodal computational imaging method where mid-infrared spectral imaging (MIRSI) is employed with second harmonic generation (SHG) microscopy to identify fibrillar collagen in biological tissues.

Aim: To demonstrate a multimodal approach where a morphology-specific contrast mechanism guides an MIRSI method to detect fibrillar collagen based on its chemical signatures.

Approach: We trained a supervised machine learning (ML) model using SHG images as ground truth collagen labels to classify fibrillar collagen in biological tissues based on their mid-infrared hyperspectral images. Five human pancreatic tissue samples (sizes are in the order of millimeters) were imaged by both MIRSI and SHG microscopes. In total, 2.8 million MIRSI spectra were used to train a random forest (RF) model. The other 68 million spectra were used to validate the collagen images generated by the RF-MIRSI model in terms of collagen segmentation, orientation, and alignment.

Results: Compared with the SHG ground truth, the generated RF-MIRSI collagen images achieved a high average boundary $F$ -score (0.8 at 4-pixel thresholds) in the collagen distribution, high correlation (Pearson's $R$ 0.82) in the collagen orientation, and similarly high correlation (Pearson's $R$ 0.66) in the collagen alignment.

Conclusions: We showed the potential of ML-aided label-free mid-infrared hyperspectral imaging for collagen fiber and tumor microenvironment analysis in tumor pathology samples.

Significance: Esophageal cancer is becoming increasingly prevalent in Western countries. Early detection is crucial for effective treatment. Multimodal imaging combining optical coherence tomography (OCT) with complementary optical imaging techniques may provide enhanced diagnostic capabilities by simultaneously assessing tissue morphology and biochemical content.

Aim: We aim to develop a tethered capsule endoscope (TCE) that can accommodate a variety of point-scanning techniques in addition to OCT without requiring design iterations on the optical or mechanical design.

Approach: We propose a TCE utilizing exclusively reflective optics to focus and steer light from and to a double-clad fiber. Specifically, we use an ellipsoidal mirror to achieve finite conjugation between the fiber tip and the imaging plane.

Results: We demonstrate a functional all-reflective TCE. We first detail the design, fabrication, and assembly steps required to obtain such a device. We then characterize its performance and demonstrate combined OCT at 1300 nm and visible spectroscopic imaging in the 500- to 700-nm range. Finally, we discuss the advantages and limitations of the proposed design.

Conclusions: An all-reflective TCE is feasible and allows for achromatic high-quality imaging. Such a device could be utilized as a platform for testing various combinations of modalities to identify the optimal candidates without requiring design iterations.