Archer's breathing and heart rate: Lo Presti 2019

A group of Italian engineers built two elastic bands with an optical fibre sensor inside them and checked whether they are enough to read an archer's breathing and heartbeat while shooting. The bands go around the chest and the abdomen and stretch very slightly as the rib cage moves: the sensor, a Bragg grating encapsulated in a silicone rubber, turns that deformation into a shift in the wavelength of the light.

Calibration was done on nine healthy volunteers (four men and five women, aged 22 to 30), comparing the bands against a spirometer for breathing and a fingertip sensor for the heartbeat. Across 135 breaths, respiratory frequency matches the reference instrument almost exactly, with a mean percentage error of 1.92% for the chest band and 1.97% for the abdominal one. Across 149 beats, measured during breath holding, the chest band does better (mean error 3.92%, correlation 0.91) than the abdominal one (5.74%, correlation 0.76), and the limits of agreement stay wide: from minus 11.2 to plus 12.5 beats per minute for the chest, and from minus 22.5 to plus 22.7 for the abdomen.

They then went to the field with two archers only, a man and a woman, with two and three years of practice. Each shot two ends of six arrows at 70 metres, one for practice and one scored, and the bands picked out all six shooting actions. Respiratory frequency before each arrow ran from 20.0 to 26.5 acts per minute in the first archer and from 7.1 to 15.2 in the second; heart rate during the aiming phase sat between 94.9 and 120.5 in the first and between 82.1 and 97.3 in the second. The scores for the ends are not reported, and none of these measures is related to the outcome of the shot.

Each band is a polyamide strip 600 by 40 millimetres, fastened with Velcro, carrying a 10-millimetre Bragg grating inside a silicone rubber package of 90 by 24 by 1 millimetre. The two bands sit on anatomical landmarks, the xiphoid process for the chest and the navel for the abdomen, so that the same measurement point is found every time. The fibres run to an optical interrogator sampling at 250 hertz.

On the nine volunteers the protocol was a short breath hold to synchronise the instruments, sixteen quiet breaths and a final breath hold for as long as each person could manage. The reference for breathing was a flow sensor connected to a differential pressure transducer, and for the heartbeat a photoplethysmography sensor on the left index fingertip. The respiratory signal was filtered between 0.05 and 0.5 hertz and the cardiac signal with a Butterworth band pass between 0.6 and 20 hertz; respiratory frequency comes from the spacing between maximum peaks and heart rate from the spacing between minima. The comparison with the reference instruments uses Bland-Altman analysis and linear regression.

At the shooting range the bands were in the same positions, with the optical interrogator 3 metres from the archer. Since calibration had already been done in the laboratory, no reference instrument was used while shooting, so as not to impede the movement. The field analysis uses only the chest band, because the abdominal one gave a more irregular signal: for each end the six shots are isolated, the portion matching the aiming phase is filtered to give heart rate, and the breathing that precedes the shot is filtered to give respiratory frequency.

Calibration and the field test are two separate things, and only the first has a reference. The nine volunteers are not archers and were measured while standing still, breathing quietly or holding their breath: the accuracy reported applies to that condition, not to a body at full draw. While shooting, in order not to impede the movement, no comparison instrument was used, so the heart rate and breathing values reported for the two archers are not checked against anything.

The archery sample is two people, with two and three years of practice: there are no experienced archers, and the study does not claim any. There are no scores, no link between measurement and outcome, and no statistics. The shots number twelve per archer, six for practice and six scored, on a single day.

Questions remain about the measurement too. Heart rate was calibrated during breath holding, that is, in the condition where breathing does not disturb the signal, whereas at full draw an archer holds or slows the breath without stopping it: the authors themselves write that how to extract heart rate in the presence of breathing remains to be worked out. The limits of agreement are wide, especially for the abdominal band, where a single beat can be off by more than twenty beats per minute. A Bragg grating also responds to temperature, and the sensor characterisation the authors cite gives 0.012 nanometres per degree, a factor that matters little indoors in a laboratory but was not controlled outdoors.

This is an instrumentation paper, and as a piece of evidence about shooting it counts for what it makes measurable, not for what it claims. It shows that an archer's breathing and heartbeat can be followed arrow by arrow with two light bands, connected to a fibre that does not get in the way of the movement, and that the chest band picks out both the phases of the shot and the heart rate during aiming. It is the kind of instrument needed to study the question of when the release falls within the cardiac cycle, not an answer to that question.

Nothing changes about shooting itself. The two archers measured are two, with two and three years of practice, and their numbers are individual curiosities: in the first, heart rate during aiming rose by some fifteen beats in the scored end compared with the practice end, and in the second, respiratory frequency fell, but with no scores and no statistics these differences cannot be read. The breathing advice reported in the introduction, to breathe in during the set-up and either breathe out or hold during aiming, comes from other papers cited there: nobody tested it here.

For anyone measuring themselves, the useful point is one of scale. A wrist or chest heart rate monitor gives an average figure over a few seconds; a measurement fine enough to say where in the cardiac cycle an arrow left calls for instruments like these, whose error on a single beat stays around 4% even in the better of the two positions tested.