Accendo

Accendo is a collapsible, motorized spiral attachment designed to help clinicians access the deep small intestine more safely and efficiently. The device pairs the existing spiral “pleating” method with a motor and an added safety mechanism: a soft, collapsible spiral that engages the intestinal wall, rotates to advance by pleating tissue, and can collapse on demand mid-procedure to rapidly disengage if needed. In our proof-of-concept work for a pneumatic design, we evaluated advancement performance in ex vivo swine tissue, characterized inflation-deflation behavior, and validated that the device does not significantly restrict commercial endoscope articulation. I presented this work at the 2025 MIT URTC conference as an IEEE conference paper.

As the project matured, we also explored a second "buckling" collapsing strategy to reduce system complexity. The mechanically “buckling” spiral had a spiral-pattern that buckles under compression and returns when the load is removed.

Accendo Conference Paper

Accendo Presentation

To play, press and hold the enter key. To stop, release the enter key.

Accendo gave me experience in taking a medical device from early concept to validated proof-of-concept, then into a second design direction based on what we learned from testing. We built protocols to evaluate the device in ways that translate to clinical constraints, such as how well the spiral advances by pleating in tissue, how quickly it can disengage, and whether it preserves endoscope flexibility.

On the technical side, I gained experience in soft robotic mechanism design and validation. I was especially involved in the ex vivo swine small-intestine tests, using a custom suspension rig to quantify advancement per turn and compare against a positive control. This is an especially rigorous process in the medical device space.

I gained experience on how to pivot and fully lead a design. After building confidence in the pneumatic collapsibility concept, I led the exploration of a mechanically simpler “buckling” spiral approach that engaged under axial compression and retracted when a load is removed. I learned about material selection and fabrication methods with different TPUs, resin prints/coatings, and tube laser cutting.

Finally, this project strengthened my scientific communication. I presented our work at MIT URTC conference and prepared an IEEE conference paper for publication through the conference proceedings.

Takeaways

Team leadership
Medical device design
Tissue testing
Fabrication techniques
Paper writing and presenting

Process

Motivation

Reaching the small intestine is necessary to diagnose and treat many gastrointestinal conditions, but it remains difficult with standard endoscopy. Current deep bowel enteroscopy options include balloon-based methods that are slow and operator-intensive or a non-motorized spiral systems. Olympus had a motorized spiral that was incredibly efficient at advancing deep into the small intestine, but following a patient death due to the device being unable to be pulled out of the patient mid-operation, the device was recalled for safety risks.

There was still a huge opportunity in a more efficient and safe way to reach the small intestine.

Goals

The goal of Accendo was to develop a collapsible, motor-driven spiral endoscope attachment that improves small-intestine access safely.

Key design requirements included:

On-demand engagement for pleating/advancement and rapid disengagement to safely withdraw if needed
Demonstrate proof-of-concept performance in tissue by quantifying movement rate in an ex vivo swine model and compare against controls
Ensure the device does not significantly limit bending/articulation of existing commercial endoscopes
Stay within safe physiological and geometric constraints such as maximum outer diameter < 40 mm, retracted OD < 20 mm, and mucosal contact pressure < 45 mmHg

Progress

Concept Generation and Downselection

After brainstorming sessions with the team, we settled on three ways to approach creating a retractable spiral. We did some first order calculations and modeled how the physics of the prototypes would work out, then downselected using a pugh chart to a pneumatic spiral. We prototyped two ways of pneumatic spirals in parallel: an entirely thermoformed TPU spiral and a pneumatic thread spiral. The goal was to find a geometry and material combination that could engage tissue, pleat effectively, and collapse on command.

Prototype Fabrication

We heavily iterated through prototypes to find optimal spiral dimensions. We first created completely solid spiral geometries and tested them for each geometry's pleat efficiency and ability. Then, with the solid prototypes that performed well, we made a pneumatic spiral with those same dimensions.

For the thermoformed spiral, we used an additive/thermoforming workflow of 3D-printing a mold, vacuum forming around it, then performing a final seal. For the pneumatic thread spiral, a TPU sheet was folded and heat-sealed into an inflatable thread, then bonded into a helical structure on a flexible core to create a spiral that expands with pressure and collapses when depressurized.

Early ex vivo trials comparing multiple geometries and materials showed the pneumatic-thread spiral produced more regular pleating, so we selected it for further iterations and characterization.

Testing

We formalized a prototype test plan with an existing manual spiral device, Spirus, as a positive control, and having no device as a negative control. We defined measurable metrics for performance and safety-relevant behavior.

Key tests we used to validate the pneumatic device included:

Collapsibility time: time to engage and fully retract, targeting <10 s
Endoscope compatibility: ensuring the attachment doesn't significantly affect endoscope bending or minimum bend radius using a bench comparison with and without the device
Inflation-deflation characterization: measuring spiral outer diameter across pressures using a pump, gauge, and calipers for engagement safety and retraction pressures
Small intestine advancement: quantifying how far the system advances per number of spiral turns in an ex vivo swine model

Ultimately, the pneumatic-thread spiral achieved 13cm advancement over 10 spiral turns (n=3) and was able to disengage as a safety feature. This moved faster than existing solutions and on par with the previous motorized spiral device, with the added safety. We also validated that adding the pneumatic spiral did not significantly affect bending behavior or minimum bend radius compared to the standalone scope. A big takeaway from this stage was that tissue is not a clean, analytical system. There is lots of variability in friction, compliance, hydration state, size, and handling, which can largely affect outcomes.

Conference Writing and Presenting

After our proof-of-concept validation of the device, I led the paper and presentation effort. I presented at the 20205 MIT URTC to an IEEE panel, and the paper published in the IEEE Xplore.

Buckling Spiral Design

After our pneumatic proof of concept, we pivoted to explore a buckling-based collapsible spiral to simplify the fully-integrated motorized system. This was because combining rotation with pneumatic routing introduces challenges like sealing and rotary union constraints.

The buckling concept uses columns cut into the sleeve wall that buckle when the device is axially compressed to create the spiral and retract when the load is removed. We tested early buckling spiral prototypes and found compression was easy by hand, bending was relatively unaffected, and rotation propogation from outside the body seemed feasible.

Next Steps

The major next steps would be to design a complete system with the collapsable spiral and motorized rotation mechanism. While we had concept of the full system, we need to build and test the whole system with ex vivo models and eventually in vivo. At the same time, we need to think about scalable fabrication methods for our device.