In bioimaging, the quest to see beyond the diffraction limit has produced a mature toolkit. Super-resolution microscopy techniques, including stimulated emission depletion, structured illumination microscopy, and single-molecule localization microscopy, now routinely achieve sub-50 nanometer resolution in living cells. Expansion microscopy offers a complementary route by physically enlarging specimens before observation on conventional microscopes, delivering nanoscale performance without exotic optics. Meanwhile, aggregation-induced emission probes have improved photostability for long-term imaging sessions, and the integration of adaptive optics with computational reconstruction has pushed super-resolution methods into thicker tissues and whole organisms. Lattice light-sheet microscopy combined with adaptive optics now enables low-phototoxic volumetric imaging of subcellular dynamics in developing embryos, a capability that was largely theoretical a decade ago.

Phase-based and vibrational imaging have expanded the optical repertoire beyond fluorescence. Quantitative phase imaging enables non-invasive measurements of cellular mass, morphology, and dynamics, with improvements in holographic and interferometric methods supporting millisecond-scale temporal resolution for real-time cell monitoring. Vibrational techniques, including Raman spectroscopy, coherent anti-Stokes Raman scattering, and stimulated Raman scattering, add chemical specificity by probing molecular vibrations, allowing researchers to visualize lipids, proteins, and metabolites without labels. Hyperspectral stimulated Raman scattering further accelerates chemical mapping for diagnostic applications. Light-sheet fluorescence microscopy, which illuminates samples with a thin orthogonal sheet of light, provides rapid and gentle volumetric imaging, while nonlinear approaches such as two- and three-photon excitation and second- and third-harmonic generation enable deep-tissue imaging with intrinsic optical sectioning and reduced photodamage.

Despite this progress, the authors identify phototoxicity as a central limitation for long-term live-cell imaging, particularly in high-intensity super-resolution modalities. Light-sheet methods help mitigate these constraints but introduce their own challenges related to optical aberrations and instrument complexity. The Perspective also stresses that standardized imaging protocols and quantitative metrics are increasingly necessary for reproducibility across laboratories, a concern echoed by community initiatives such as QUAREP-LiMi, which seeks to advance quality assessment and reproducibility in light microscopy. Looking ahead, the authors anticipate multimodal platforms that integrate fluorescence, super-resolution, quantitative phase, Raman, light-sheet, and nonlinear signals, together with brighter fluorophores, improved adaptive optics, novel lasers, and machine-learning-guided acquisition.

Optical biosensors have evolved from laboratory curiosities into clinical diagnostic tools, driven by advances in nanofabrication and surface chemistry. Surface plasmon resonance and localized surface plasmon resonance sensors exploit the sensitivity of plasmonic nanostructures to local refractive index changes, enabling label-free detection of biomolecular binding events. Recent innovations in nanostructured substrates and metamaterials have pushed detection limits toward single-molecule sensitivity. Photonic crystal sensors and whispering gallery mode resonators offer alternative transduction mechanisms with high quality factors and compact footprints, and these platforms have demonstrated femtomolar detection limits for disease biomarkers while becoming increasingly integrated into microfluidic systems for point-of-care diagnostics.

At the opposite end of the cost spectrum, paper-based biosensors represent what the authors call a paradigm shift toward low-cost, disposable diagnostics for resource-limited settings. By integrating colorimetric or fluorescent detection with microfluidic paper devices, these sensors achieve rapid, equipment-free analysis of complex samples, and recent work has demonstrated multiplexed detection of infectious disease markers with sensitivity approaching laboratory-based assays. Fiber-optic biosensors combine optical detection sensitivity with remote sensing capability, and functionalized fiber tips enable minimally invasive in vivo measurements ranging from intracellular pH sensing to real-time metabolite monitoring in tissue. The next generation of biosensors will likely emphasize multiplexing, miniaturization, and integration with digital health platforms, with wearable and implantable sensors for continuous biomarker monitoring seen as a particularly promising frontier. Key challenges include improving selectivity in complex biological matrices, extending sensor lifetime and stability, and developing robust calibration methods.

Optical tweezers, which use focused laser beams to trap and manipulate microscopic objects, have revolutionized single-molecule biophysics by enabling precise force measurements at the piconewton scale. Researchers can now directly observe molecular motors, DNA-protein interactions, and protein folding dynamics. Recent instrumentation advances, including high-speed detection, active feedback control, and multi-trap configurations, have expanded the range of accessible biological phenomena. Holographic optical tweezers enable simultaneous manipulation of multiple particles in three dimensions, facilitating studies of collective cellular behaviors, and integration with fluorescence microscopy provides simultaneous mechanical and biochemical information. Beyond single molecules, optical tweezers have become essential for probing cellular mechanics and the viscoelastic properties of biological fluids: active microrheology using optically trapped probe particles reveals the frequency-dependent mechanical response of cytoplasm, extracellular matrix, and mucus, while measurements of forces during cell division, migration, and adhesion have connected mechanical cues to cellular decision-making and revealed mechanosensitive signaling pathways.

Particle tracking techniques round out the experimental toolkit. Single-particle tracking has become indispensable for studying molecular dynamics in living cells, revealing heterogeneous diffusion behaviors, transient binding events, and spatial organization that ensemble measurements obscure. Recent advances in localization algorithms and high-speed imaging have pushed temporal resolution to microseconds and spatial precision to nanometers, and sophisticated analysis methods now extract quantitative information about diffusion modes, confinement, and molecular interactions from trajectories. Particle image velocimetry and particle tracking velocimetry map velocity fields in biological fluids and tissues, with applications from cardiovascular flow analysis to ciliary beating, and optical coherence tomography-based particle tracking extends these capabilities to optically scattering samples for non-invasive flow measurements in tissues. The authors expect future methods to emphasize label-free detection, increased throughput, machine-learning-assisted analysis, and super-resolution tracking capable of resolving nanoscale dynamics in crowded cellular environments.

Artificial intelligence has transformed the analysis side of the field. Convolutional neural networks now routinely perform cell segmentation, classification, and tracking with accuracy exceeding human experts, and notable applications include AI-enhanced super-resolution microscopy, where neural networks predict high-resolution images from diffraction-limited inputs, dramatically reducing acquisition time and phototoxicity. Generative models have enabled virtual staining, predicting fluorescence images from label-free brightfield or phase contrast data. Yet the authors caution that many AI models require large annotated datasets that are expensive to generate and may not generalize across experimental conditions or laboratories, making standardization of training data, validation protocols, and performance metrics essential for clinical translation. They highlight physics-informed machine learning, which incorporates physical constraints and domain knowledge into neural network architectures, as a promising route toward more reliable and interpretable analysis, alongside causal inference frameworks and explainable AI techniques that reveal which image features drive model predictions.

The Perspective closes with a unifying vision. The greatest opportunities, the authors argue, lie at the intersections of the five domains: multimodal platforms that combine complementary optical techniques, the transition from qualitative observation to quantitative measurement through calibration and reference materials, and the democratization of advanced optics through open-source hardware, smartphone-based imaging, and cloud-based analysis. They also flag ethical considerations, noting that data privacy, algorithmic bias, transparency, and accountability must be addressed proactively as AI becomes embedded in biological research and clinical diagnostics. Among the grand challenges for the next decade, they list achieving molecular-resolution imaging in living organisms over extended timescales, developing multiplexed wearable biosensors for continuous health monitoring, creating autonomous microscopy systems that design and execute experiments, and building interpretable AI models that not only analyze data but generate testable biological hypotheses. Realizing these ambitions, the authors conclude, will require sustained interdisciplinary collaboration among physicists, engineers, biologists, and computer scientists, with attention to reproducibility, standardization, and accessibility so that advanced optical technologies benefit the broadest possible community.

Subject of Research: Advances in optical methods for biological sample characterization

Article Title: A perspective of advances in optical methods for biological sample characterization

Article References: Avila, R., Morales-Narváez, E., & Loza-Alvarez, P. (2026). A perspective of advances in optical methods for biological sample characterization. Frontiers in Photonics, 7, Article 1773615. https://doi.org/10.3389/fphot.2026.1773615

Image Credits: AI Generated

DOI: 10.3389/fphot.2026.1773615

Keywords: biophotonics, super-resolution microscopy, optical biosensors, optical tweezers, particle tracking, artificial intelligence, quantitative phase imaging, light-sheet microscopy, machine learning, biosensing, quantitative imaging, interdisciplinary research