Hyperspectral imaging can be used in medical applications to distinguish tissue types based on differences in their spectral signature.^{1 – 5} Currently, many researchers use a variety of methods to develop classification algorithms for hyperspectral imaging, such as machine learning, statistical analysis, or fitting algorithms, which identify tissue types based on spectral differences between tissue types. The quality of the data used by the classification algorithm influences the accuracy and robustness of the developed algorithm. Ideally, the spectra used to distinguish tissue types are only influenced by the tissue composition, but unfortunately these spectra are also influenced by other factors, such as surface reflections and sample thickness variations. Removing influences not related to tissue composition is expected to improve the accuracy of the developed algorithm itself, and it would also make the developed algorithm more generalizable, because it does not depend on the amount of surface reflections and sample thickness variations present in the dataset used to develop the algorithm. Thus, reducing these influences is a valuable step in the development of classification algorithms for hyperspectral imaging.

Recent advancements in multispectral (MS) and hyperspectral (HS) microscopy have focused on sensor and system improvements, yet sample processing remains overlooked. We conducted an analysis of the literature, revealing that 40% of studies do not report sample thickness. Among those that did report it, the vast majority, 98%, used 2–10 µm samples. This study investigates the impact of unstained sample thickness on MS/HS image quality through light transport simulations. Monte Carlo simulations were conducted on various tissue types (i.e., breast, colorectal, liver, and lung) using optical property parameters extracted from the literature. The simulations revealed that thin samples reduce tissue differentiation, while higher thicknesses (approximately 500 µm) improve discrimination, though at the cost of reduced light intensity. Although the results are based on idealized conditions and exclude certain real-world factors such as sample variability and instrument-specific effects, they highlight the need to study and optimize sample thickness for enhanced tissue characterization and diagnostic accuracy in MS/HS microscopy.

Introduction: Optical property retrieval in diffuse reflectance imaging, like diffuse reflectance spectroscopy (DRS) and hyperspectral imaging (HSI), often involves fitting measured spectra to analytical solutions using approximations such as Diffusion Theory (DT). This method, while accurate, is not always generalizable due to the assumptions inherent in DT and results in non-unique solutions for optical properties and physiological parameters. In addition, it is computationally intensive. Physics-inspired deep learning offers generalizable data descriptions guided by physical principles but requires extensive labelled data, which is hard to obtain, especially in medical contexts.Methods: We propose a deep learning approach to retrieve physiological parameters from DRS and HSI spectra using DT-simulated training data. The DT-simulated data is synthesised using a range for the optical properties: Blood Volume Fraction (BVF), Saturation, water-fat ratio (WFR), average blood vessel radius (R), scattering amplitude (SA), and scattering slope (SL). The range for these parameters we have extracted from literature.Results: Our feed-forward neural network achieved median relative errors of 4% and 2% for DRS and HSI, respectively.Discussion: Results suggest that the proposed method is robust and that retrieval of optical properties is possible with similar results to DT but also reducing operation time.

SignificanceMinimally invasive surgery (MIS) offers substantial benefits to patients, including reduced trauma and faster recovery. However, it limits visual and tactile feedback, which can affect surgical decision-making. Hyperspectral and multispectral imaging (HSI/MSI) are imaging technologies with the potential to provide detailed, real-time tissue characterization, enhancing minimally invasive intraoperative guidance.AimThis perspective aims to summarize the current state-of-the-art in the use of HSI/MSI in laparoscopy and endoscopy. It focuses on both technological development and clinical implementation, providing an overview of performance characteristics and translational state.ApproachA structured literature search was conducted to identify relevant studies. These were analyzed for key system specifications (spectral range, resolution, and acquisition speed) and clinical applications. The studies were further categorized using a “bench-to-bedside” framework to evaluate their level of translational progress.ResultsThe majority of studies fall within early translational stages (T1 to T2), with no reports of large-scale clinical trials. Most systems operate in the visible spectrum (450 to 650 nm), optimized for blood-related imaging. Applications include perfusion assessment, nerve visualization using short-wave infrared wavelengths, and tumor detection.ConclusionsHSI/MSI for MIS is a rapidly developing field with demonstrated potential across a range of applications. Continued research and validation are essential to transition these technologies from experimental use to routine surgical practice.

Positive surgical margins (PSMs) after radical prostatectomy increase biochemical recurrence and secondary treatment, highlighting the need for reliable margin assessment. This study aimed to evaluate how tissue heterogeneity, glandular origin, and malignancy grade affect the performance of diffuse reflectance spectroscopy (DRS) in distinguishing prostate tissue types. In an ex vivo study of 104 patients, spectra were acquired and matched to histology for classification. Performance differed by anatomical region, with tumor–healthy discrimination varying across zones. Grade effects were inconclusive due to limited data. Heterogeneous tissues yielded about 72% sensitivity and specificity.

Optical phantoms are widely used to characterize diffuse optical setups and data analysis methods for in-vivo/ex-vivo measurements. Coconut oil is an interesting compound to use in phantoms, because it could be used to model lipidic tissues, such as the one in breast tissue. In this paper, we measure the absorption and scattering spectra of coconut oil from 600 to 1600 nm, encompassing the so-called “therapeutic window”. To cover the entire range, we exploit a supercontinuum pulsed laser and a superconducting nanowire single photon detector operating in the time domain. Finally, we demonstrate the use of a homogeneous coconut oil phantom to characterize a hyperspectral continuous-wave (CW) setup.