Respiratory Profiling: A Window into the Nervous System

A Simple, but Powerful, Way to Measure Internal States

In previous articles, we have considered how the breath is a portal into our Nervous System, as recently covered in depth and in detail by James Nestor in his book:

Book Review: "Breath; The New Science of a Lost Art"

The breath can be either autonomic (involuntary), or under our somatic (voluntary) control if we focus on it. When we a breathing unconsciously, our mode of breathing, including the rate and depth, is a reflection of our internal Nervous System state. Conversely, when we take control of our breathing, we can rapidly change our internal Nervous System states through different breathing styles. For example, we can use our breath to rapidly calm ourselves down. Breathing regulation and dysregulation has profound impact on those of who are suffering from chronic illness, trauma, or stress.

My friend and colleague, Samar Singh [email available on request], has developed a novel way to extract deep insights into our internal, and in-the-moment, Nervous System states via a technique he calls “Respiratory Profiling” (RP). The concept is quite simple, easy to use, totally non-invasive, and virtually cost-free.

Samar has created a program where a mobile device or phone with a video camera can capture a recording of the belly movements due to breathing of a person being observed. As the person breaths in and out, Samar’s program detects the movement of the belly, from which the program generates a wave pattern or trace, a bit like an ECG, of the belly position versus time. Importantly for privacy, the program does not actually record or transmit any video, it just collects the waveform information, and sends back numerical information about the state of the person being observed, together with a unique “ANS [Autonomic Nervous System] Score”.

While this is easy to implement, we have been finding, using modern wave pattern analysis tools, that the information contained in the trace is profound, deep and rich, with far reaching implications for the insights it gives into the Nervous System states. The information the program automatically captures includes: the number of breaths per minute; the ratio of exhalation to inhalation times; unconscious breath holding at the end of the inhale or exhale; the breath to breath variation; the shallowness or depth of the breath; the diaphragmatic movement, and so forth.

In fact, Samar’s method even reflects Heart Rate Variability (HRV) measurements, which are often taken with pulsometers or wearable wrist devices, via an app on a phone. However, Samar’s technique doesn’t require any devices beyond the phone with video capabilities. Samar has found that RP not only reproduces the information from HRV measurements, but that the information from RP is higher quality/resolution, and is captured well via a relatively short recording (Samar recommends a two minute recording is sufficient).

I personally believe the reason that the RP method picks up heart rate information, as well as breathing rate, is because of the aorta which runs down into the belly region. If you lie on your back, and look down at your belly, while holding your breath at the end of exhale, you can sometimes see the pulse of the aorta. With modern, high resolution phone cameras, this aortic pulse movement is also captured by Samar’s method, and hence heart rate information is also contained in the RP data.

However, using machine learning (more on this below), Samar has been able to reduce the complexity of the HRV readings, and the large degree of interpretability of HRV graphs, into his unique ANS Score. Indeed, the ability to quantify internal states from RP measurements is one of the strengths of this method.

Samar originally developed this idea, because he was helping families with kids who are on the autistic spectrum to regulate their Nervous Systems using the Safe & Sound Protocol, and he needed a way to assess what was helping or not, and optimize the benefits of the procedure.

The Ears as a Portal into the Nervous System

As he was working in poor parts of India, over a large region so he could not be everywhere at once, he needed a way for parents to take measurements, including while a child was asleep, in a way which didn’t require expensive devices or equipment, beyond a standard mobile phone. So he created his RP program, and indeed found that he could use it to fast track results, and improve outcomes and success rates. For example, he used his method to discover a unique pieces of music which could be played quietly with small speakers on the mastoid bone to provide nervous system calming effects.

Samar has since developed the idea towards detecting, diagnosing, and monitoring, other conditions, including alerts of early warning signs of symptoms. This included being able to predict the onset of seizures in a client 24 hours before the actual seizure occurred, and thereafter putting interventions in place for the client to ameliorate the seizure.

By comparing and contrasting the readings of people diagnosed with covid, to matched healthy people, he has demonstrated that the program together with machine learning, is also capable of early warning of symptoms of covid.

The software also permits measuring of immune system states, presumably because the breath contains information about the immune function and baroreceptor relationship, often discussed in the HRV literature. He has also been able to detect tell-tale signs of trauma.

In other preliminary work, the machine learning model based can track blood sugar values after eating a meal from initial increase to return to baseline in about 30 minutes with readings taken about 5 minutes apart. The comparison with observations processed using the Glucose oxidase Trinder method and the model seem to indicate a difference of about 10%.

Samar has since been trying to get healthcare providers and doctors to work with him, especially to help him access large cohorts of patients with specific conditions to collect and analyse readings, and prove the utility of the method. Unfortunately, he has had very little interest.

This is why I suggested to Samar that we should put the information about RP out there, via this article, since its potential applications are very much wider than just healthcare, and perhaps we should seek partners outside of the medical establishment per se.

For example, I am in particular thinking of the advantages this could provide for sports and athletes, and human performance more generally. The window into internal state provided by RP could definitely help athletes optimize individual training regimes and performance, such as optimizing recovery times. As small margins can make a huge difference in modern sports, the big advantages RP could bring, could provide a massive competitive edge.

I am also thinking of emergency services, in particular where equipment is limited and conditions are difficult, such as natural disasters and war zones, to triage and monitor the injured in real time.

It also has application for therapist of all sorts, via assessing or proving the worth of, and fine tuning or improving, interventions, therapies, and devices, through taking “before and after” readings. For example this could be used to assess the relative benefits of various breathing protocols, various diets, various supplements, and so forth.

It can also be used in neurofeedback and biofeedback, because the person who is being observed can watch the trace being recorded in real time, so they can learn how to change the wave pattern by adjusting their breath, e.g. for helping to learn how to overcome anxiety through self-calming techniques.

It could also support neurophysiological basic research [does anyone know anyone who knows Prof. Andrew Huberman? Perhaps we can get this message to him and alert him of Samar’s developments?]

While Samar has used machine learning to extract important information from time series traces that he has already collected, this is a good project here for AI developer’s to collect and an analyse very large data sets, to create all sorts of diagnostic tools.

Update: I have used the new Substack Notes facility to include some slidedecks presentations we had previously created about the RP method: