1. Background and Questions Expected to be Solved by This Study

Understanding the internal structure of entire organs at cellular resolution is crucial in modern biology and medicine. Tissue clearing techniques, which make tissues transparent, are central to this effort because they reduce light scattering caused by the natural complexity of biological materials allowing for the assessment of the three-dimensional complexity of organs without the need to section them into smaller pieces (Fig. 1). Despite major advancements, the field still heavily relies on qualitative observations to assess clearing effectiveness, meaning evaluations are often subjective and inconsistent. Existing methods either require expensive equipment, complex protocols, or only allow measurement at fixed time points rather than dynamically over time.

Figure 1. Principles of tissue optical clearing

Reducing light scattering and light absorption are the main physical principles of tissue optical clearing. RI matching can reduce scattering, and removal of pigments can reduce light absorption. Various chemical-biomolecule interactions contribute to these physical effects, resulting in tissue optical clearing. Reproduced from Yu, T., Zhu, J., Li, D. & Zhu, D. Physical and chemical mechanisms of tissue optical clearing. iScience 24, 102178 (2021).

The major question addressed by this study is: Can we create a simple, real-time, quantitative method to monitor tissue clearing processes dynamically, allowing better control and optimization without the need for complex equipment or expert knowledge?

The study published titled " ClearAIM: dynamic quantitative label-free monitoring of tissue optical clearing" in Advanced Imaging seeks to introduce a method that objectively and continuously measures both level of tissue transparency and how much its shape changes during the clearing process.

2. Theories and Experimental Focus

The research is built on a combination of computational optical imaging, deep learning, and tissue engineering principles. The authors developed a method called ClearAIM (Clearing AI Monitoring), which uses a straightforward optical setup and a specially tailored deep-learning segmentation algorithm (based on the "Segment Anything Model", or SAM).

The experimental focus included:

• Using a straightforward demagnified brightfield imaging system to record time-lapse images of mouse brain slices undergoing clearing with the CUBIC-R1 solution (one of the most commonly used and non-toxic, clearing reagents).
• Automatically analyzing these images to extract two key parameters over time:
  1. Transparency, calculated in-frame by comparing the brightness of the tissue to the background.
  2. Area changes, tracking how the tissue swells or shrinks during clearing.

The deep learning algorithm was especially adapted to cope with the dynamic changes in tissue appearance by automatically updating - via robust k-means clustering - how it identifies the tissue in each frame, making the method reliable even as transparency and size change (Fig. 2).

Figure 2. ClearAIM data processing algorithm for metrics calculation.

The described approach enables straightforward and reproducible monitoring of changes in sample transparency and size in near real-time, limited only by the preset frequency of image acquisition. Adapted from Kalinowski, K. et al. ClearAIM: dynamic quantitative label-free monitoring of tissue optical clearing. ai 2, 021003 (2025).

To validate ClearAIM, the researchers applied it to brain slices of two thicknesses (500 µm and 1 mm) and monitored their clearing over 3 days, analyzing how transparency and area evolved. Finally, tissue fluorescence labelling and imaging was performed to quantitatively compare optical clearing via its direct real-life application.

3. Innovation, Significance, and Potential Application of the Study

Innovation:

• ClearAIM introduces real-time, automated, and quantitative monitoring of tissue clearing, something previous methods lacked.
• The method uses a novel adaptive frame-to-frame guidance technique with deep learning, ensuring that segmentation remains accurate even as tissues change dramatically.
• The setup is low-cost and made from off-the-shelf components, meaning almost any lab can adopt it easily without specialized expertise. ClearAIM app (and all codes) is freely available.

Significance:

• Researchers can now objectively measure, compare and optimize tissue clearing procedures, improving reproducibility across labs.
• Continuous real-time monitoring helps scientists understand when the optimal balance between transparency and minimal structural distortion is reached, which is critical for high-quality imaging later.
• By correlating transparency levels with structural changes, scientists can tailor protocols more precisely to their needs, instead of relying on trial and error.

Potential Applications:

• Optimized application of newer, faster clearing protocols, some of which can clear tissues in under an hour.
• Quality control in large-scale imaging studies, where consistency between samples is crucial (e.g., brain mapping projects, cancer tissue studies).
• Broader imaging fields where tracking changes in sample transparency and morphology over time is important.

By making ClearAIM open-access, the authors hope to democratize this technology and encourage its widespread use, potentially accelerating research in biology, neuroscience, oncology, and other fields relying on tissue imaging.

In the follow-up work, the Group is conducting comparative studies of different tissue clearing methods across various organs to find the most suitable ones for each tissue type.

Main authors

Kamil Kalinowski is a first year PhD student in Quantitative Computational Imaging lab at Warsaw University of Technology, Institute of Micromechanics and Photonics. He is interested in image analysis, segmentation methods, and numerical techniques for biomedical applications and optical imaging.

Anna Chwastowicz is a MD-PhD student in Medical University of Warsaw, Department of Immunology, and Nencki Institute of Experimental Biology of Polish Academy of Sciences, Laboratory of Neurobiology, (BRAINCITY). She is interested in optical tissue clearing and imaging.

Piotr Arcab is a third PhD student in Quantitative Computational Imaging lab at Warsaw University of Technology, Institute of Micromechanics and Photonics, working in ERC – NaNoLens project. His research interests include lensless digital in-line holographic microscopy, Gabor hologram processing and reconstruction, designing and assembling holographic microscopy setups. His work has been published in renowned scientific journals like ACS Photonics and Optics Express.

Dr Mikołaj Rogalski recently defended (cum laude) his PhD and is now a postdoctoral researcher in Quantitative Computational Imaging lab at Warsaw University of Technology, Institute of Micromechanics and Photonics. His research interests include Fourier ptychographic microscopy (FPM), lensless digital in-line holographic microscopy, deep learning aided optical measurements, and novel methods for efficient analysis of interferograms and holograms. His work has been published in renowned scientific journals like Laser & Photonics Reviews, Bioinformatics, and ACS Photonics.

Dr Paweł Matryba (MD-PhD) graduated in medicine and biotechnology, with his research focusing on advanced imaging techniques, including optical tissue clearing and expansion microscopy.
He is a recipient of prestigious scholarships such as the START program by the Foundation for Polish Science and the Minister of Education's Scholarship for outstanding young researchers.
Matryba has led research projects funded by the National Science Centre, and his work has been published in renowned scientific journals like Laser & Photonics Reviews and View.

Dr Maciej Trusiak is an associate professor at the Institute of Micromechanics and Photonics, Faculty of Mechatronics, Warsaw University of Technology WUT. Following his doctoral studies (cum laude 2017), he completed a one-year postdoctoral fellowship in the Optoelectronic Image Processing Group led by Prof. Javier García and Prof. Vicente Micó at the University of Valencia, Spain. In 2022, he obtained his habilitation degree and launched the Quantitative Computational Imaging Lab (qcilab.mchtr.pw.edu.pl), focusing on computational imaging, lensless microscopy, optical metrology, interferometry and holography, quantitative phase imaging, and fringe pattern analysis. In 2023, he was awarded the ERC Starting Grant for research on lensless, label-free nanoscopy. He is a Senior Member of SPIE and Optica, and served on the SPIE Award Committee. He has held various organizational roles, including Co-Chair and Committee Member at the SPIE Optics + Photonics 2022 and 2025, the Warsaw Summer School for Advanced Optical Imaging 2024 and Computational Optical Sensing and Imaging (COSI) at the Optica Imaging Congress 2024 and 2025. He currently serves as Associate Editor for Applied Optics (Optica Publishing Group) and Optics and Lasers in Engineering (Elsevier), Executive Editorial Board Member for Journal of Physics: Photonics (IOP), and Editorial Board Member for Advanced Devices & Instrumentation (AAAS Science Partner Journal).