Infrared Light Sees Deeper into Tissue Samples Now

Scientists are continually developing new methods to image biological samples, particularly when dealing with the challenges posed by light scattering within dense tissues. Emilia Wdowiak, Piotr Arcab, and Mikołaj Rogalski, working at the Institute of Micromechanics and Photonics, Warsaw University of Technology, in collaboration with Anna Chwastowicz and Paweł Matryba from the Laboratory of Neurobiology and Department of Immunology at the Nencki Institute of Experimental Biology of Polish Academy of Sciences and Medical University of Warsaw, and Małgorzata Lenarcik from the Department of Pathology at the Maria Sklodowska-Curie National Research Institute of Oncology and the Department of Gastroenterology, Hepatology and Clinical Oncology at the Centre of Postgraduate Medical Education, alongside Julianna Winnik, Piotr Zdańkowski, and Maciej Trusiak from Warsaw University of Technology, present a novel approach to lensless digital holographic microscopy. Their research introduces near-infrared lensless holographic microscopy (NIR-LDHM) utilising a standard visible-light CMOS sensor, enabling high-throughput, label-free imaging even in strongly scattering media. This advancement significantly extends imaging depth and resolution compared to conventional visible-light techniques, offering potential for improved diagnostics and fundamental biological research within complex tissue samples.

Scientists have devised a new imaging technique allowing them to see deeper into biological tissues than previously possible. The method bypasses limitations caused by scattering, offering clearer views without the need for staining or labels, with potential benefits for medical diagnostics and biological research. Scientists have extended the reach of lensless digital holographic microscopy into strongly scattering materials by employing near-infrared light and a standard visible-light camera sensor.

This work addresses a long-standing limitation of holographic techniques, their inability to clearly image through turbid samples like biological tissues. The study leverages the properties of near-infrared light, which penetrates deeper into tissues due to reduced scattering and absorption compared to visible wavelengths. Remarkably, this was achieved using a conventional CMOS sensor designed for visible imaging, operating near its silicon cutoff at approximately 1100nm, successfully reconstructing images under limited photon availability.

Increasing the distance between the sample and the sensor, from 3mm to 12mm, further enhanced lateral resolution by twofold under strong scattering conditions, an advantageous mechanism previously unrecognised. Researchers applied this near-infrared lensless digital holographic microscopy (NIR-LDHM) to visualize uncleared mouse tissue slices, resolving internal structures in brain slices up to 250μm deep and liver slices up to 60μm, offering label-free amplitude-phase imaging without sample preparation.

A near-infrared lensless digital holographic microscopy (NIR-LDHM) platform underpinned this work, utilising a conventional, board-level CMOS sensor originally designed for visible (VIS) imaging, operating up to the silicon cutoff of approximately 1100nm. Six illumination wavelengths, 480nm, 532nm, 632nm, 850nm, 940nm, and 1100nm, were employed using a supercontinuum laser source and acousto-optic tunable filter, providing spectral filtering with a full width at half maximum bandwidth of 1.8-8.5nm.

Tissue-mimicking milk scattering layers and calibrated resolution targets were used to quantify reconstruction performance, allowing controlled variation of scattering strength and sample distance for systematic evaluation. The experimental arrangement followed a standard in-line LDHM configuration, but increasing the sample-sensor propagation distance from approximately 3mm to 12mm improved lateral resolution twofold under strong scattering conditions, reframing distance as a tunable parameter.

Despite the CMOS sensor exhibiting only 0.19% quantum efficiency at 1100nm, the system achieved robust reconstructions even under low-photon-budget conditions. Wide-field, label-free amplitude-phase imaging of uncleared mouse brain and liver tissue slices resolved internal structures up to 250μm and 60μm respectively. Investigating NIR-LDHM phase and amplitude imaging under high- and low-photon-budget regimes revealed robust reconstructions despite a detector efficiency of only 0.19% at 1100nm, even with average power density as low as 4.9 nW/mm2.

Establishing high-photon-budget reference measurements in the visible range allowed for consistent comparison across wavelengths, necessitating moderate increases in exposure time for near-infrared wavelengths to achieve comparable signal levels. NIR-LDHM maintains resolvable features through scattering layers up to approximately 1.4mm, a substantial improvement over the visible regime, which resolves features below 350μm.

The system employed a multi-plane Gerchberg-Saxton reconstruction algorithm, acquiring five images at different distances, each positioned approximately 0.6mm farther from the sample. Scientists have long sought to see deeper into living tissue without causing damage, a challenge that has limited medical diagnostics and biological research. Conventional microscopy struggles with thick samples because light scatters, creating blurry images.

This new work bypasses lenses, employing near-infrared light to penetrate further into turbid materials like biological tissues. Achieving clear images through scattering isn’t simply about wavelength; the team’s use of near-infrared light, specifically extending beyond the visible spectrum to around 1100 nanometres, is a key advance, allowing them to resolve features within millimetre-thick samples.

Surprisingly, increasing the distance between the imaging system and the sample improved resolution under strong scattering conditions, opening new avenues for optimising image acquisition. Adapting existing, inexpensive CMOS sensors, typically used for visible light, to operate in the near-infrared has lowered the cost and complexity of building powerful microscopes.

However, the low efficiency of the sensor at these wavelengths remains a limitation, demanding brighter light sources or longer exposure times. Future work will likely focus on improving sensor sensitivity and exploring the potential of this technique for in vivo imaging, potentially offering real-time diagnostics without invasive procedures. This approach could ultimately extend beyond medical applications, finding use in environmental monitoring or industrial quality control where seeing through opaque materials is essential.