Bova - Imperial

Inspiration

Completely untapped market, yet huge!
We use animals all the time in research, why not as the clinical target?
Companies like Loyal or Dognosis and the animal->human medical startup model
The potential to collect large datasets by being present in many animals, which could hugely help neuroscience, and human clinical applications

A vagus nerve stimulator for cattle to improve:

immune response and inflammation during disease, especially bovine respiratory disease
reducing stress before slaughtering, a welfare goal, that also produces better quality meat
gut mobility, for improved feed efficiency and yields
some control over methane release, via the burp reflex, to reduce greenhouse gas emissions

Simulation Pipeline:

built a bovine vagus nerve model based on a pig dataset (bovine data is extremely sparse, or completely unavailable)
developed a Finite Element Model (FEM) pipeline coupled with NEURON simulations using the opensource NRV framework
this allowed us to derive activation thresholds for different unmyelinated and myelinated nerve populations
then re-ran the simulations with different electrode configurations to optimise the electrode cuff design (electrode site count)
produced selectivity index figures of merit for different configurations, ensuring minimal off-target stimulation
viability proof of concept in the bovine vagus

Huge simulation pipeline challenges, extremely RAM heavy, had to parallelise work across personal home server, local, and an AWS instance
Additionally, nasty bugs that set us back hours due to the fairly difficult to use NRV package
Huge gap in literature coverage, barely could find cow vagus nerve images, and quantitative reporting
Real operator cost data is 3-4x higher than published literature, hard to access over a weekend
Commercial feedlot operators are not on forums, reaching the right people proved nearly impossible in 48 hours
Had to pivot away from ear tag (ABVN) to direct cervical VNS mid-hackathon after honest assessment of the bovine anatomy evidence gap

To the best of our knowledge, produced the first (approximated) cow vagus stimulation simulation
Managed to get in touch with very relevant people despite locking our idea on Friday evening
Got Dr. Ronald Tessman DVM PhD (Elanco, 17 years) on the phone and validated the core thesis in a single call
Our team collaboration quality, never panicked, very fluid, despite working together for the first time
Independent operator validation across CT Forums and Reddit confirming $80-200 all-in BRD cost vs $23.60 published figure
Identified total white space, zero competitors, zero published research, zero funded programs globally
Built a scientifically honest dossier including an honest risk assessment, we know what we know and what we don't

Cow-calf, backgrounder, and feedlot operators are completely different customers, we were talking to the wrong people early
The feedlot veterinarian is the real decision-maker, not the owner
Drug resistance is making the antibiotic alternative framing more urgent than we initially understood
The data moat is worth more than the device margin
How to run FEM+NEURON simulations (first time experience)
How to go to primary sources when the literature is lacking

Week 1: finalise university partner (KSU or UC Davis), convert Tessman network into pilot feedlot contact
Month 1: bovine cadaver anatomy characterisation, IJV and carotid diameter, vein-to-nerve distance, tissue impedance, nerve anatomy analysis
Months 1-24: Phase 1 nerve cuff pen trial, 40 animals, $50-90k
Year 2: Phase 2 stent validation and CVM pre-submission
Year 3-4: Phase 3 commercial stent at 100,000 units
At 1% market penetration: world's largest chronic vagal dataset, licensable to SetPoint, LivaNova, Inspire, possibly applicable to human appetite control, and gut disorders long term

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.