**13. A device**

environment improved greatly by 2010; her EKG was not measured in 2011, and she had a

Another subject was caring for her aging mother at home (Figure 30). The caregiver did not sleep from 28–29 September because her mother did not sleep. Consequently, the subject's SI was very low (3 PM on 29 September). However, she slept well the following two nights because her mother slept well. This subject visited my exhibition booth at Innovation Japan

**Figure 29.** EKG data from a woman aged 27 years collected in September 2006 at 3 PM and five-year follow-up results.

**Figure 30.** Sleep deprivation decreases SI. These data were collected from a woman in her 50s who was taking care of

good SI in 2012 (data not shown).

378 Advances in Bioengineering

2011.

her aging mother.

I am currently collaborating with a company to construct a device equipped with a stablebaseline (meaning, the EKG trace is NOT lost from the recorder chart or screen) EKG amplifier and data logger. In the future, an mDFA program will be incorporated into the device. The device will instantaneously compute SI, and EKG data, mDFA results, and SI values will be stored on an SD memory card. All data will be transmitted to a PC through a web-based system (Figure 31).

**Figure 31.** A device that records stable-baseline EKG (Symphodia Phil Confidential, Japan).

## **14. Discussion**

Unique characteristics of mDFA are shown in Figure 18 and 20. In addition, the unique mDFA computation is specified in equation (2); DFA does not use this procedure.

mDFA calculates SI using a method that includes preparation of time series data, removing trends from the data, determining statistical variations, constructing a graph of F(n) vs. (n), where F(n) is variance and (n) is box size, and determining the slope, i.e., SI. In order to construct a gadget that rapidly calculates SI, I determined an appropriate scaling zone where a straight line and a regression line were drawn using the least-squares method; this zone is [30: 270] in beat per min (BPM).

Until now, a DFA-user must determine the box range before drawing a regression line to obtain the SI. I determined the best range to compute SI automatically is [30: 270]. Biological systems do not operate optimally because ranges that are too small or too large are avoided. As a biomedical tool, however, mDFA uses a special box-range [30: 270]. Currently, people can construct their own device to analyze their own cyclic phenomena. The device would auto‐ matically return SI values when people include time series data, and SI could be computed for any time series interval.

The time length of 270 heartbeats in the [30: 270] box range corresponds to 270 sec if the heart beats at a steady ratio of 60 BPM, which is equivalent to 3–4 min. This time length has a particular significance with respect to neuroscience, because internal biochemical reactions can retain a steady state for this length of time. I can wait for three minutes before instant noodles are cooked. A boxer fights with full power for three minutes. The neuron network functions under the three-minute law. Short-term memory and acute stress, for example, can last for three minutes but they can also decline after three min. The length [30: 270] (three min) could be a biologically important number.
