Near-infrared holography breaks the visible-light scattering barrier in lensless microscopy

Light scattering has haunted microscopy for as long as scientists have tried to peer inside thick biological tissue. In a perfectly transparent specimen, photons travel along paths that can be traced back to the structures that shaped them. In dense, heterogeneous samples, however, photons are deflected again and again, scrambling the information they carry. The image-forming signal loses power, fine detail vanishes, and even the most sophisticated reconstruction algorithms are left struggling with little more than haze. Conventional microscopes can fight back with specialized optics, scanning routines, fluorescent labels, or tissue-clearing protocols, but each of these interventions costs time, field of view, or instrument complexity.

Lensless digital holographic microscopy has long approached this problem from a radically different direction. Instead of magnifying the sample through an objective lens, the specimen is placed directly above a bare image sensor. A compact light source illuminates it, and the sensor records an interference pattern rather than an ordinary photograph. Numerical propagation then reconstructs two complementary maps: amplitude, which describes how strongly the specimen attenuates light, and phase, which describes how much it delays the wave. Because no objective narrows the scene, the entire sensor surface becomes the field of view. In the new study, that amounted to roughly 116 square millimeters captured in a single exposure, without raster scanning or image stitching.

Yet this elegant simplicity has always concealed a fundamental weakness. The in-line holography scheme pioneered by Dennis Gabor assumes that a strong, unscattered reference wave interferes with a much weaker object wave. That assumption holds beautifully for isolated cells and very thin, transparent specimens. Once multiple scattering becomes strong, the reference wave is depleted, holographic fringes drown in diffuse background, and reliable reconstruction breaks down. For decades, the method has therefore remained confined to the weak-scattering side of microscopy, limited to visible wavelengths and thin samples.

A team at the Quantitative Computational Imaging Lab at Warsaw University of Technology, led by Professor Maciej Trusiak, asked whether that boundary was as fixed as it appeared. Their reasoning rested on a well-known physical fact: near-infrared light is scattered less strongly by many tissue structures than visible light, so more useful information survives propagation through a turbid sample. In a conventional microscope, exploiting that advantage would require infrared-optimized detectors, lenses, and coatings. In a lensless system, there are no imaging lenses to replace, and even the faint residual response of an ordinary silicon sensor near 1100 nanometers might be enough for holographic reconstruction. The idea was disarmingly simple: change the wavelength, preserve the architecture, and test whether image information lost to visible light could be recovered through the near-infrared wave.

The researchers tested this concept across six illumination wavelengths spanning 480 to 1100 nanometers, all on the same core imaging platform. Tissue-mimicking milk layers provided controlled and repeatable levels of scattering, while calibrated resolution targets measured the finest features that remained recoverable under each condition. The findings were published online in the journal Opto-Electronic Advances on September 20, 2026, in a paper led by first author Dr. Emilia Wdowiak.

The result was a decisive spectral shift in performance. With visible-light illumination, the system became fully unresolved once the scattering layer reached thicknesses between 180 and 360 micrometers, depending on the exact wavelength. At 1100 nanometers, resolvable features survived through approximately 1.4 millimeters of the tissue-like medium, nearly four times the thickness achievable at 632 nanometers. Perhaps most strikingly, the standard visible-range CMOS sensor delivered robust reconstructions even though its quantum efficiency at 1100 nanometers was just 0.19 percent. The experiment demonstrated that residual sensor sensitivity, usually dismissed as negligible, can become genuinely useful when paired with a holographic method that operates on a low photon budget.

The study also overturned one of the field’s most deeply held practical instincts. Lensless systems normally position the sample as close to the sensor as possible to maximize resolution, since diffraction-based reconstruction improves with proximity. Under strong scattering, however, the team found that increasing the sample-sensor distance from about 3 to 12 millimeters actually improved lateral resolution from roughly 11 to 5.5 micrometers. The proposed explanation is that the useful interference-carrying light remains more spatially concentrated than the broadly diffused background; with additional propagation distance, the noise spreads faster than the signal, making the holographic fringes easier to detect. The effect still requires deeper mechanistic study, but it immediately reframes sample-sensor distance as a parameter to optimize rather than simply minimize.

Most importantly, the controlled experiments translated directly to biology. At 1100 nanometers, the system produced ultra-wide-field, label-free amplitude and phase images of uncleared mouse liver sections up to 60 micrometers thick and brain sections up to 250 micrometers thick, revealing internal organization that visible illumination largely obscured. In the liver sections, near-infrared illumination exposed internal trabecular patterns and vascular pathways that remained invisible at 532 nanometers, even though the visible light preserved the overall tissue outline. Rather than claiming non-invasive imaging deep inside a living organ, the advance establishes a new operating regime for compact in vitro holography: rapid, wide-area recovery of intrinsic tissue structure in thick, strongly scattering sections, without stains, clearing agents, or sectioning to vanishing thinness.

The work was a genuinely interdisciplinary effort. Optical engineering from Warsaw University of Technology was combined with biological and clinical expertise from the Laboratory of Neurobiology at the Nencki Institute of Experimental Biology of the Polish Academy of Sciences, the Department of Immunology at the Medical University of Warsaw, the Department of Pathology at the Maria Sklodowska-Curie National Research Institute of Oncology, and the Centre of Postgraduate Medical Education. According to Dr. Wdowiak, this collaboration enabled the team to move beyond resolution targets and test the method on biologically meaningful tissue architecture, a step that separates a laboratory demonstration from a potentially practical imaging tool.

Looking ahead, Professor Trusiak notes that future work can combine distance-aware acquisition with scattering-aware or physics-informed reconstruction algorithms and more sensitive detectors, pointing toward faster tissue surveying, computational imaging through complex media, and compact instruments for laboratories where conventional microscopy platforms are difficult to scale. The QCI Lab, which develops optical methods in which computation becomes part of the microscope, has previously contributed lensless techniques for suppressing holographic twin images, reconstructing data recorded with very limited photon budgets, and creating gigavoxel-scale three-dimensional maps of biological specimens. The research was supported by Poland’s National Centre for Research and Development through the LIDER XIV programme, with equipment co-funded by the Warsaw University of Technology Excellence Initiative: Research University programme. The authors declare no competing financial interests. If the approach continues to mature, the humble silicon sensor sitting in countless laboratories may soon see far deeper into scattering tissue than anyone assumed its visible-range design would ever allow.

Subject of Research: Near-infrared lensless holographic microscopy for label-free imaging through multiple scattering in thick biological tissue

Article Title: Breakthrough near-infrared holography pushes lensless microscopy beyond its traditional visible-range scattering limit

Article References: Breakthrough near-infrared holography pushes lensless microscopy beyond its traditional visible-range scattering limit. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: lensless microscopy, digital holography, near-infrared imaging, light scattering, CMOS sensor, quantitative phase imaging, label-free imaging, mouse liver sections, mouse brain sections, computational imaging, tissue imaging, Opto-Electronic Advances