VentMon 0.1 - Inline Ventilator Test Fixture and Monitor

Since March 16th, I've been working full time on this. Along with volunteers at EndCornaVirus.org and Public Invention, we have been tracking and analyzing all open-source ventilator projects we can find. It has become apparent that the biggest piece they are missing and the place Public Invention can most help is not in designing a new ventilator, but in building a test/monitoring fixture to assist other teams.

Because monitoring is critical to keeping patients alive and so similar to the need to test, we are attempting to make a combine tester/monitor.

At the moment, this repo is for that work, and also our overall strategic plan for a process that will give clinicians enough confidence to actually deploy an open-source ventilator. Our goal is to certify, in a certain limited, last-resort sense, ventilators in the case of a shortfall of commercial and professional equipment. That Strategy is discussed lower in this document.

This includes rapidly doing high-quality tests, but also includes burn-in and multi-day tests, of ventilators. Perhaps we can produce a benchmark for the multiple efforts going on right now to try to develop a Free-libre Open Source ventilator design.

Volunteering

This project is co-led by Robert L. Read read.robert@gmail.com and Geoff Mulligan.

Righ now we have several needs. People need to be experienced, have read the essential reading list, be able to work at least 20 hours a week, and able to work with limited supervision. They also must be able to tolerate the chaos produced by the pandemic itself, and the chaos produced by my own mistakes. Finally, they must accept and hope for the possibility that industry will increase production and the disease will be controlled so that this solution is not required.

Righ now, we could use:

1. An experienced Arduino engineer to build and improve the design below
2. Someone who knows how to reliably process a time series to tell when a breath begins and respiration rate
3. A website build who can make nice graphs of pressure/volume curves show up
4. Someone to build a software test-bench to analyze 48 hours worth of data and mark anomalies.
5. Someone to compare our Javascript analys algorithms to BreathMetrics in MatLab.
6. Someone to design the User Interface of the "monitor and alarm" feature. Would have to be or become familiar with existing ventilators to make a similar interface. This is likely to be a small LCD or OLED text screen, 4 physical buttons, and a small speaker or buzzer.

Licensing

All code in this repo is licensed under the GNU 3 public license. All documents and diagrams and licensed under CC0.

Initial Design Idea

The idea is to make a standalone inline device plugged into the airway. It serves a dual purpose as a monitor/alarm when used on an actual patient, and a test devices for testing prototype ventilators. It also allows for burnin.

Status

This is an emergeny project in response to the COVID-19 pandemic. We are in a learning mode.

Got the SensirionFM3200 flow meter working. Documentation on this hard to find---had to guess a bit.

Saturday Night (March 28th): Can record a breath trace and accurately identify transitions. Now must try integration to compute volume!

Sunday morning: The first two VentMon units are loosely promised to the teams lead by Ethan Chaleff and John Bennett.

Diagrams of the current system architechure and electrical design are below.

Parts

We are moving fast and Agile; expect this to change.

Adafruit BME680 Note: Adrafuit is currently closed due to pandemic, but this part is in stock at Mouser.

Sensiron Flow Meter Note: This requires an unusual connector (JST ZH 1.5mm 4 pin).

JST ZH 1.5MM 4 Pin Female connectors

Obviously, Arduinos, with control shield, and probably speakers, but I haven't selected them yet.

Progress

Motivation of Strategy

As of March 18, 2020, the COVID-19 pandemic is growing quickly. The Maker, Humanitarian Enginering, and Free Sofware communities are attempting to answer the call. Public Invention is attempting to help the potential shortfall of ventilators, which is uncertain but could mean hundreds of thousands of needless deaths in a worst-case scenario.

Working with NECSI and EndCoronavirus.org, a self-organized project instigated by Bruce Fenton has arisen. It has become clear, however, that there are and should be many technical approaches.

In order to unify this community, some of us at the #response-ventilators-meta channel at the necsi-edu.slack.com, include Nariman Poushin, have decided any practical response will have to focus on clinical effectiveness and reliability testing, because ventilators are life-critical pieces of machinery. In effect, we need an informal certification process similar in spirit only to FDA approval, which can convince a doctor or medical decision maker to deploy unapproved ventilators as an absolute last resort if there is a shortage of ventilators.

The Free Software and Maker communities know how to form teams; our plan is to form teams for design, testing, training, building, and even fundraising. The basic process we propose is captured in the diagram below.

Necessity of Extensive Testing

To deploy a ventilator, extensive testing is absolutely required. This repo is an attempt to build a "test fixture" capable of running a 48-hour test on any ventilator design and collecting data on many important parameters. We hope to create a "gold standard" test that all DIY teams can work to; but this project will proceed in parallel with that.

As part of extensive testings, in combination with BreathOfLife (thanks to Jonathan Orban) and with help from Juan E. Villacres Perez, biomedical engineering student at UT, we have create a Google Sheet downloaded as an OpenDocument Calc and Microsoft Excel spreadsheet of user tests which attempt to match the RMVS standard for Rapidly Manufactured Ventilator Systems. (There are checkboxes in the Google Sheet that don't look nice in the other two formats.) The spreadsheet is licenced under CC0, so you can make a copy of it and use it for testing your own ventilators.

Physical Test Lungs

I believe that a test lung is meant to provide a resistance to inflation which models a human lung. An example of a purchasble device is:

Allied Healthcare AHP300 Mechanical Test Lung

5cm H20/L/sec. Resistance Test Lung

Note the units of resistance here: Pressure/volume/sec. Presumable a test that used open air would not accurate model a ventilator.

Certifiers

An example of a certifier, including an explanation of its functionality is here: Alnor-4070 Certifier Flow Analyzer

However, we also need to do multi-iteration reliabillity tests. An initial design is here:

Older diagram

Request for Donations to Public Invention

Public Invention is a small, all-volunteer 501(c)3 US public charity. Your small donation is a stamp of approval for our work that will really help us.