$400 You Can Build a Tissue Processor

 

This post draws directly on the original paper: Wang M, Pelletier S, Ellis A, et al. "Design and evaluation of low-cost, DIY programmable tissue processor for solvent exchange in biological sample preparation." PLoS One. 2026;21(3):e0341033. The article is published under a Creative Commons Attribution License, so we quote its language closely where it's the clearest way to describe the system.

Introduction

Imaging techniques are fundamental tools in biology for examining cell growth and responses to the environment, but many tissues require fixing, staining, and/or clearing before they can be visualized under a microscope. A classic example is propidium iodide (PI), a fluorescent cationic stain that is membrane-impermeant and selectively enters cells with compromised membranes — i.e., dead or damaged cells. In animal cells it binds to DNA and RNA in the nucleus, while in plant cells it binds to negatively charged pectin in the cell wall, emitting red fluorescence in both cases.

The catch is that such stains greatly enhance image contrast, but their application typically requires elaborate sample preparation. Many staining protocols involve sequential dehydration and rehydration steps to incrementally replace the tissue's fluid environment, ensuring thorough infiltration of dyes into the tissue and enabling subsequent tissue clearing. A typical PI staining protocol for plant non-transparent tissues moves samples through a graded series of ethanol solutions — from water up to 100% ethanol and back down — with each step lasting on the order of 20–30 minutes. In practice, implementing such a protocol requires manually transferring solvent over several hours or days, a process that is labor-intensive and tedious, and introduces risk of variability, such as tissues drying out if a step is delayed or the solvent concentration is miscalculated.

Automated tissue processors exist and could, in principle, execute these staining sequences, but existing commercial tissue processors are often expensive and engineered for far more complex tasks than needed for routine biology workflows. High-end histology processors — which automate dehydration, clearing, and paraffin infiltration — can cost on the order of tens of thousands of US dollars, and typically include added features like vacuum and heating for embedding tissues in wax. Such systems are financially out of reach for many research groups and are overbuilt for protocols that only require solvent exchanges.

A Cornell University team (Wang, Pelletier, et al.) set out to close that gap. In a 2026 PLOS One paper, they developed a low-cost, open-source tissue processor that automates sequential solvent exchanges for biological tissue. The processor can be constructed for approximately $400 in parts and materials, using off-the-shelf components commonly found in makerspaces and a few custom 3D-printed pieces.

1. How the System Works

The processor consists of two main glass containers: a premix container and a tissue container. The tissue sample is placed in the tissue container, where it remains for the duration of processing. The premix container serves as a reservoir where the next solvent solution is prepared at the desired concentration before being transferred into the tissue container — a two-container design that allows precise control of concentration changes while keeping the tissue mostly stationary and submerged in liquid to avoid drying.

The authors use peristaltic pumps to draw from stock solvent reservoirs — typically distilled water, 70% ethanol, and 100% ethanol — and mix into the premix container to achieve the target concentration. Once the premix container has the correct solution, a motorized ball valve opens to fill the tissue container by gravity-driven flow. After the programmed soaking time, a second valve at the base of the tissue container opens to drain the solution, and the next step begins. This cycle ensures that the tissue remains submerged except during the few seconds required for the drain/refill transition, during which it is exposed to air.

Peristaltic pumps were chosen specifically because they are self-priming and prevent cross-contamination, and are reliable for repetitive liquid handling, with routine maintenance limited to periodic tubing replacement. To transfer solvent from the premix to the tissue container, the team used gravity-driven flow controlled by motorized valves instead of additional pumps, since this step does not require precise volume delivery.

The processor is controlled by an Arduino Mega microcontroller running custom software, with a text-based menu system accessible via a keypad and an LCD screen. This interface allows users to configure, modify, and execute solvent exchange protocols directly on the device. The main menu offers several functions:

• Run Program — execute a predefined solvent exchange routine, with the display showing the current step number, target concentration, elapsed time, and estimated time remaining.
• Modify Programs — create or edit protocols stored on the SD card, organized in a JSON file that can hold multiple named routines; advanced users can also remove the SD card and edit the JSON file directly on a PC for quicker editing of many steps.
• Flush System — opens all valves and runs all pumps simultaneously to rinse the tubing and containers after a protocol is finished.
• Sanity Check — a diagnostic mode that sequentially activates each pump, toggles each valve, and reads the keypad, providing on-screen feedback for each test, which is especially useful the first time the system is built or if any component is replaced.
• Calibrate — guides the user through calibration of pump flow rates or volumes, with the software allowing adjustment of a scaling factor for delivered volume if, for example, a different pump model is used.

Most structural components were fabricated with an FDM printer and a DLP printer, with a combined print time of about fifteen to twenty hours on a hobby-grade machine. The CAD models were developed in SolidWorks and are provided in STEP and SLDPRT formats, and the Arduino code — along with all data — is provided in the paper's Supporting Information.

2. Cost Breakdown

As summarized in the paper's bill of materials, the overall system can be assembled for approximately $400 (the authors' itemized total comes to $422), with the main cost contributions coming from the glassware (~$100), the peristaltic pumps (~$75), and the microcontroller (~$50). The remainder is spread across the motorized ball valve, relay board, silicone tubing, PVC pipe, wood and metric screws, and other off-the-shelf hardware.

3. Performance: Does It Actually Work as Well as Manual Processing?

The team characterized pump performance by pumping water for fixed time intervals and measuring the volume dispensed with a graduated cylinder, finding a consistent flow rate of 1.10 ± 0.02 mL/s (mean ± SD). For a typical exchange volume of 250 mL, the corresponding delivery time was approximately 4 minutes — well before each typical 20–30 minute incubation step — which enabled predictable and repeatable solvent exchange without the need for flow sensors or closed-loop control.

To evaluate the processor's performance, the team applied it to a standard PI staining protocol on tomato (Solanum lycopersicum) stem tissues, with identical samples processed either manually or with the automated system, then imaged on a Zeiss LSM 880 confocal microscope. They compared signal-to-noise ratios (SNR), defined as the mean intensity of the stained cell wall divided by background intensity. Both manual staining and staining with the tissue processor yielded similarly high SNR values (~80), with no significant difference between methods (t-test, P > 0.05; n = 3 tissue samples processed in three independent runs) — indicating comparable staining quality between the two approaches.

On the efficiency side: for the PI staining workflow, the processor executed more than 20 sequential solvent exchange steps, each with a programmed duration of 30 minutes, resulting in a total automated processing time of approximately 10 hours. User hands-on time was limited to protocol programming and solvent loading, requiring less than ~10 minutes per run — compared to several hours of intermittent manual intervention for hand-processed samples.

To demonstrate versatility, the team also applied the processor to PI-stain a fig leaf (Ficus carica), a tomato leaf, an Euphorbia peplus leaf, and a cross-section of an Arabidopsis thaliana flower, all subsequently cleared with methyl salicylate and imaged by confocal microscopy. Throughout, the processor executed more than 20 solvent exchange steps — including dehydration, rehydration, and post-stain washes — and by keeping tissues submerged and automating each step, the device minimized drying artifacts and user-induced variability. The data were collected across different users and independent processing batches, which the authors say highlights the processor's versatility and reproducibility.

4. Limitations (In the Authors' Own Words)

The paper is candid about what the system doesn't do:

• No direct comparison to commercial processors. The authors note they did not perform a direct comparison with commercial tissue processors due to their high cost and limited accessibility in many laboratories, including their own — so manual staining, as the most commonly used approach, served as the practical benchmark instead.
• No closed-loop sensing. The current implementation does not include closed-loop sensing (for example, flow or liquid-level sensors) to autonomously detect errors such as pump failure or valve blockage. That said, the tissue processor has been used routinely in their laboratory for over one year without observed hardware or software failures. Incorporating feedback mechanisms would further improve robustness but would also increase system complexity and cost — a tradeoff the team says was deliberately deprioritized in favor of low cost and ease of replication.
• Tubing material constrains solvent choice. The authors used silicone tubing, which is compatible with ethanol and a wide range of polar solvents, but note that users should avoid non-polar solvents like toluene or xylene, as silicone degrades under prolonged exposure to such liquids; chemically resistant alternatives such as PTFE tubing should be used in those cases.

5. Closing Thoughts

The paper's own framing is useful here: unlike commercial tissue processors — often expensive and overengineered for standard lab needs — this design is affordable (~$400), compact, and easy to replicate using parts commonly found in makerspaces. And it's explicitly not limited to hydration or dehydration steps; the authors note it can be readily adapted for protocols requiring stepwise solution changes, including staining, clearing, fixation, or washing — with the caveat that solvent compatibility depends on the tubing material and should be considered case by case.

What stands out from reading the paper directly isn't just "$400 is cheap" — it's how the authors deliberately scoped the design: rather than trying to replicate the full feature set of a commercial processor, they built specifically for the routine solvent-exchange workflows that many basic biology labs actually run day to day, and validated that the tradeoff didn't cost them staining quality. The full design files (CAD models in STEP/SLDPRT, Arduino source code, and a step-by-step operation video) are openly available via the paper's Supporting Information and a linked Zenodo repository, so any lab with basic 3D-printing and Arduino capability can, in principle, replicate one.

---

Reference: Wang M, Pelletier S, Ellis A, Shepherd RF, Frank MH, Stroock AD, Mishra AK, Bacheva V. Design and evaluation of low-cost, DIY programmable tissue processor for solvent exchange in biological sample preparation. PLoS One. 2026;21(3):e0341033. https://doi.org/10.1371/journal.pone.0341033. Open access, distributed under the Creative Commons Attribution License.