Patent Application
20200061377
This application generally relates to devices that can be held against a person's body or attached to something connected to the body to treat depression in the person.
Stress is a specific global impairing problem with no globally effective solution. Approximately a third of the US population describes having more stress than they would like to without effective coping strategies. Chronic unmanaged stress results in time missed from work and impaired productivity, and can lead to debilitating mental and physical illnesses from depression and anxiety to heart disease and dementia. Interventions for stress primarily consist of psychological therapies/techniques and drug/prophylactic therapies. Psychological techniques (e.g., as taught in psychotherapy or meditation) depend on the user attending to their mental events and emotions, and applying learned techniques in the moment. Literature suggests such interventions do not work for many people. For example, individuals who tend to ruminate do most poorly in even the best psychological therapies. Drug/prophylactic therapies depend on an agent taken at one time point to have effects much later when they are needed. These treatments have many side effects and are often addictive. A common observation is that most psychiatric medications, in tolerable doses, are not strong enough to prevent or even improve in-the-moment reactivity to stressors. What is needed is an intervention that becomes effective at times of stress, intervenes without requiring strong presence of mind on the part of the user, and then stops being active when the stress dissipates. It would be desirable to provide a technology that is applicable to solving problems throughout multiple fields and markets broadly spanning from the medical field (e.g., psychiatry, psychology), to enterprise applications (e.g., HR and training), military applications, consumer models (e.g., stress management), and the entertainment business (e.g., emotionally augmented adaptive virtual reality).
Depression is a condition described in the Diagnostic and Statistical Manual of Mental Disorders Fifth Edition, involving sadness or inability to experience pleasure and a host of other symptoms. One line of treatment of depression includes administration of anti-depression medication such as selective serotonin reuptake inhibitors (SSRIs), e.g., fluoxetine. Research in this area is ongoing. However, an estimated 10 to 30 percent of depressed patients who are taking an anti-depressant are partially or totally resistant to the treatment.
As such, alternative treatment options are being explored for treatment of depression in conjunction with and/or without the use of anti-depression medication, some of which involve external stimulation of the nervous system. For example, electrical stimulation of the brain with implanted electrodes has been approved for use in the treatment of various conditions, including pain and mental health. Many such therapeutic applications of electrical stimulation involve the surgical implantation of electrodes within a patient. However, such methods are invasive. Alternatives thus would be desirable.
The invention regards regulation of physiological states using cutaneous transducers (vibration or electricity) at times of stress or deviation from a target emotion state (e.g., “calm”). Thus, included in the invention are 1) regulatory devices that provide specific types of physiologically reactive stimulation useful for regulating arousal and stress, 2) invocation via manual switching or sensors and algorithms to determine its need, 3) individualized calibrations and learning algorithms to determine optimal parameters for sensing and regulating a given person's arousal, stress, or distance from a target physiological state such as calm, and 4) treatment of depression in a person. A more general utility of the same technologies regards techniques that associate physiological reactivity profiles with stored idiosyncratic emotion states (e.g., “calm”), yield dynamic quantification of distance from these target states, and provide physiological stimulation that returns users to these target emotion states via closed loop stimulation.
According to an aspect of the present intervention, a regulatory device touching the skin of a user, or something that is touching the skin or body of the user, provides stimulation at frequencies demonstrated in the published literature to change sympathetic and parasympathetic tone. The device can be worn, e.g., on a wrist, the neck, or the sternum, all of which have been demonstrated to yield increases in vagal/parasympathetic tone or sympathetic tone with certain frequencies of cutaneous oscillation. The transducer for cutaneous oscillation is strong enough to produce non-ignorable signals. Electricity and vibration can both serve that function. The invention thus includes software for generating a tactile vibratory signal with specific oscillatory characteristics, a transceiver (e.g., Bluetooth transceiver) for tactile vibratory signals generated, by the device wirelessly, an amplifier, and any tactile transducer capable of feeding the amplified tactile vibratory signal to the body in a way that is not ignorable.
The mechanisms by which vibration and electrical stimulation can induce changes in stress and parasympathetic tone are well described in the literature. The skin has other receptors that process touch (mechanoreceptors), pain (nociceptors), pleasure (c-tactile fibers) and temperature (thermoreceptors). Information gathered by skin receptors travels to the brain's somatosensory cortex and in parallel, to areas of the brain involved in cognitive processes, motor processes, social processes, and networks more involved in emotional functions, including recognizing and generating emotion (the amygdala, insula, and striatum, for example), and regulating emotion (the prefrontal cortex, for instance). (Davidson, 2000, American Psychologist, 55 (1196-1214). Emotions can be generated not just from present inputs but from our interpretation of bodily cues (Damasio, 1999, xii, 386 p.) based on previous experiences as well. There are numerous other pathways by which tactile stimulation could affect brain function. For example, deep tactile stimuli can affect specific nerves directly yielding predictable neural effects. Slow moving touch around the carotid sinus (e.g., via “carotid massage”) can potentiate vagus nerve function (Ha, et al., 2015, Am J Emerg Med, 33(7), 963-5, McDonald, et al., 2014, J Am Geriatr SOC, 62(10), 1988-9, Laine Green and Weaver, 2014, J Clin Neurosci, 21(1), 179-80, Cronin and Blake, 2011, J Cardiovasc Electrophysiol, 22(5), 600) possibly via direct nerve stimulation or other routes such as stimulation of underlying muscles (Scali, et al., 2013, Spine J, 13(5), 558-63). Manual therapies like massage have implicated varying mechanisms such as release of endocannabinoids (Lindgren, et al., 2015, BMC Res Notes, 8(504). Different styles of massage to the same area have been found to activate different brain networks (Sliz, et al., 2012, Brain Imaging Behav, 6(1), 77-87) suggesting a more complex picture of the touch-brain relationship. Deep chest pressure (e.g., as used in “hug machines”) directly appears to affect baroreceptors (Edelson, et al., 1999, Am J Occup Ther, 53(2), 145-52) yielding decreased blood pressure and feelings of calm. Accupressure, which involves deep touch has been conceptualized from a variety of perspectives and also appears to affect peripheral physiology such as heart rate variability (McFadden, et al., 2012, Complement Ther Med, 20(4), 175-82, Huang, et al., 2005, Am J Chin Med, 33(1), 157-64) suggesting potential vagus nerve involvement. Electrical stimulation, in particular, activates lamina 1 fibers (McMahon and Wall, 1984, Pain, 19(3), 235-47) which convey pain, itch, sensual touch, and temperature through the spinal cord to the primitive brain areas like the amygdala, yielding emotional reactions and, through the thalamus, to the insula to yield awareness of our body's reactions (Craig, 2002, Nat Rev Neurosci, 3(8), 655-66). The insula passes this info on to cortical regions that govern our attention and emotion regulation such as the orbitofrontal cortex (Cavada, et al., 2000, Cereb Cortex, 10(3), 220-42). Electricity likely produces opioids (Chen, et al., 1996, J Pharmacol Exp Ther, 277(2), 654-60) along with beta endorphins (Bossut, et al., 1986, Am J Vet Res, 47(3), 669-76), which affect emotion and pain thresholds (Lundeberg, et al., 1989, Am J Chin Med, 17(3-4), 99-110). Electrical stimulation also activates the locus coeruleus (Snow, et al., 1999, Arch Ital Biol, 137(1), 1-28) which is responsible for norepinephrine production which could moderate stress responses.
Slow whole-body vibration, in the 0.01 to 0.3 Hz range, is associated with increased ratings of pleasantness and increased parasympathetic tone at lower frequencies with increasing predominance of sympathetic tone approaching 0.6 Hz (Uchikune, 2004, Journal of Low Frequency Noise Vibration and Active Control, 23(2), 133-138, Uchikune, 2002, Journal of Low Frequency Noise Vibration and Active Control, 21(1), 29-36). That said, whole-body stimulation at 1 Hz has been observed to produce increased parasympathetic tone (Takahashi, et al., 2011, J Hum Ergol (Tokyo), 40(1-2), 119-28), with the specific observation of being able to produce vibration-synchronized heart beats. The 0.1 Hz frequency may be particularly useful stimulate at to promote parasympathetic tone, based on biofeedback literature. The highest amplitude of RSA is achieved at 0.1 Hz (6 breaths/minute), because this is the optimal respiratory rate for oxygenation and removal of CO2 from the body. As such, 0.1 Hz is known as the frequency of cardiac resonance (high-amplitude synchronous oscillations at a single frequency). This cardiac resonance is achieved with generally lower frequency stimulation in tall people/men and generally higher frequency stimulation in smaller people/women (Vaschillo, et al., 2006, Appl Psychophysiol Biofeedback, 31(2), 129-42). Thus, it has been suggested that “any source of rhythmic stimulation that affects the cardiovascular system should produce the same effect (achieving cardiac resonance of 0.1 Hz).” (Lehrer and Gevirtz, 2014, Front Psychol, 5(756). Thus, stimulation via rhythmic muscle tension at 0.1 Hz has been found to increase high frequency heart rate variability (Lehrer, et al., 2009, Biol Psychol, 81(1), 24-30, Vaschillo, et al., 2011, Psychophysiology, 48(7), 927-36) as has rhythmic visual stimulation at 0.1 Hz (Vaschillo, et al., 2008, Psychophysiology, 45(5), 847-58). Wrist-worn vibratory cues in this range have been shown to affect hemodynamics following stressful tasks (Nogawa, et al., 2007, Conf Proc IEEE Eng Med Biol Soc, 2007(5323-5). Increased heart rates have been found for whole-body vibration at lower frequencies (3-6 Hz) (Maikala, et al., 2006, Int Arch Occup Environ Health, 79(2), 103-14) though individual differences in cardiac effects act these frequencies have been observed (Ullsperger and Seidel, 1980, Eur J Appl Physiol Occup Physiol, 43(3), 183-92). Vibration in the 6-10 Hz range appears to decrease parasympathetic tone, as measured by high frequency heart rate variability (Jiao, et al., 2004, Int Arch Occup Environ Health, 77(3), 205-12, Bjor, et al., 2007, Int Arch Occup Environ Health, 81(2), 193-9) and is specifically associated in increased low-to-high frequency heart rate variability rations, potentially suggestive of increased sympathetic tone (Watanabe and Ujike, 2012, Health, 4(11), 1029-1035). Increasingly peripheral stimulation at 10 hz is being used to modulate activity in the vagus nerve, which governs heart-rate variability, and thus parasympathetic tone(Ma, et al., 2016, J Neurol Sci, 369(27-35, Bauer, et al., 2016, Brain Stimul, 9(3), 356-63, Jiang, et al., 2016, Neurochem Int, 97(73-82, He, et al., 2016, Cardiovasc Ther, 34(3), 167-71).
This invention also relies on base frequencies in the 20-300 Hz range which are also well-documented to be physiologically reactive. A large literature devoted to whole-body vibration has primarily examined the 20-36 Hz range, which is generally associated with increased heart rates (e.g., Gojanovic, et al., 2014, Physiol Res, 63(6), 779-92, Cochrane, et al., 2008, Arch Phys Med Rehabil, 89(5), 815-21). 25 Hz whole-body vibration is associated with decreased heart rate and baroreceptor activity consistent with parasympathetic response, particularly in obese people (Dipla, et al., 2016, Exp Physiol, 101(6), 717-30), though another study found no change in HRV for 20-36 Hz following exercise (Cheng, et al., 2010, J Sports Med Phys Fitness, 50(4), 407-15). That said, electrical stimulation in the 10 Hz range has been found to produce increased HF HRV (Stein, et al., 2011, Auton Neurosci, 165(2), 205-8). Typical personal massage devices operate at about 100 Hz (Prause, et al., 2012, Sexual and Relationship Therapy, .27(1), pp), with the goal of producing subjective and muscle relaxation. Stimulation at this frequency activates the posterior insula (Coghill, et al., 1994, J Neurosci, 14(7), 4095-108) which is associated with increased attention to interoception, as promoted in many meditative traditions. Vibrator massage is associated with both increased heart rate but also increased theta and alpha EEG associated with relaxation (Diego, et al., 2004, Int J Neurosci, 114(1), 31-44). Electroacupuncture produces the sensation of rapid vibration; stimulation at 100 Hz is associated with increased subsequent high frequency heart rate variability (Hideaki, et al., 2015, Acupunct Med, 33(6), 451-6) though TENS unit stimulation at 100 Hz has been observed to produce decreased high frequency heart rate variability (Stein, et al., 2011, Auton Neurosci, 165(2), 205-8). Facial vibration at 89 Hz has specifically been associated with increased parasympathetic tone (Hiraba, et al., 2014, Biomed Res Int, 2014(910812).
According to another aspect of the present intervention, the regulatory device combines low and high frequencies described as physiologically reactive in the literature to yield unique superadditive sensations. For example, complex musical and complex vibrational stimuli have been shown to modulate a variety of neural responses associated with vagal tone (Bergstrom-Isacsson, et al., 2014, Res Dev Disabil, 35(6), 1281-91). Vibrational stimulation has been shown to affect electroencephalographic recordings as well as positive mood (Sandler, et al., 2016, Brain Topogr, 29(4), 524-38).
According to another aspect of the present intervention, use of intervention is invoked by detecting a predetermined condition, such as a manual switching on the regulatory device or the detection by the regulatory device of a predetermined condition in the form of physiological changes associated with an increase or decrease in arousal or stress. Manual switching is done by invoking software, e.g., on a smartphone, or via a switch on the device. Many physiological parameters can be used to detect emotion and arousal states such as stress or fatigue, including but not limited to changes in vocal tone, galvanic skin response, and heart rate variability, pupil dilation, and gamma band EEG. These parameters are interpretable in the presence of a baseline measurement, e.g., of stressed, relaxed, and fatigued states to which similarities on assessed parameters can be inferred. The invention thus includes software for establishing an individual's resting, stressed, fatigued, or user-defined target profiles on available physiological measurements, and for comparing, in real time, measurements of derived parameters to these states, and inferring the extent to which a current state appears similar to those stored states (e.g., stress or fatigue) with respect to individually titrated parameters.
According to another aspect of the invention, individuals differ in the extent to which different frequencies affect physiological reactions. Thus the invention includes calibration procedures to assess subjective and, when possible, physiological reactivity to different combinations of frequencies to allow best-suited regulatory stimulation.
According to another aspect of the invention, an individual's physiological features associated with potentially user-defined emotion and arousal states along with stimulation parameters that optimally restore those states can be stored in a library for later recall by that or another individual. For example, if an individual is having a particularly good day, physiology associated with that state can be stored, and stimulation parameters that optimally maintain that state or move from stressed or fatigued states towards that state can be stored in the library for later recall.
According to another aspect of the invention, an improved method of treating depression in a person can be generally said to include applying to a body area of the person a therapeutic stimulation device having a tactile transducer that is positioned proximate the body area, selecting a therapeutic stimulation pattern that comprises a first oscillation at a first frequency that is in the range of 20-300 Hz and a second oscillation at a second frequency that differs from the first frequency by 0.01-10 Hz, the first oscillation and the second oscillation together forming a beat output at the modulation frequency, and emitting with the tactile transducer the therapeutic stimulation pattern as a vibrational output toward the body area. The method can optionally include administering an anti-depressant drug to the person; stimulation in the aforementioned manner may improve the action of the anti-depressant drug by activating brain regions which facilitate drug response.
According to another aspect of the invention, an improved therapeutic stimulation device structured to be applied to a body area of a person proximate a body area to provide stimulation therapy to the person. The therapeutic stimulation device can be generally said to include a processor, a tactile transducer, and a storage, the storage having stored therein one or more routines which, when executed on the processor, cause the therapeutic stimulation device to perform operations that can be generally stated as including selecting a therapeutic stimulation pattern that comprises a first oscillation at a first frequency that is in the range of 20-300 Hz and a second oscillation at a second frequency that differs from the first frequency by 0.01-10 Hz, the first oscillation and the second oscillation together forming a beat output at the modulation frequency, and emitting with the tactile transducer the therapeutic stimulation pattern as a vibrational output toward the body area. The operations may optionally be in conjunction with the administering of an anti-depressant drug to the person, and the operations may further optionally include improving the action of the anti-depressant drug by employing the emitting of the vibrational output to stimulate activity in areas of the brain which facilitate drug response.
According to another aspect of the invention, an improved non-transitory machine-readable storage medium has stored therein instructions which, when executed on a processor of a therapeutic stimulation device having a tactile transducer that is positioned proximate a body area of a person, cause the therapeutic stimulation device to perform operations that can be generally stated as including treating the person for depression, selecting a therapeutic stimulation pattern that comprises a first oscillation at a first frequency that is in the range of 20-300 Hz and a second oscillation at a second frequency that differs from the first frequency by 0.01-10 Hz, the first oscillation and the second oscillation together forming a beat output at the modulation frequency, and emitting with the tactile transducer the therapeutic stimulation pattern as a vibrational output toward the body area.
In the figures to be discussed, the associated boxes and arrows represent functions of the process according to the present invention, which may be implemented as electrical circuits and associated wires or via wireless protocols such as Bluetooth, which transport vibrational signals. Alternatively, one or more associated arrows may represent communication (e.g., data flow) between software routines, particularly when the present method or apparatus of the present invention is a digital process.
The regulatory device 4 depicted generally in
The regulatory software routines 56 emit a combination of sine wave oscillations of different frequencies to result in a beat frequency that is output to the user. The combination of oscillations comprise a main oscillation in the range of about 20-300 Hz and at least one modulation oscillation in the range of about 0.01-10 Hz yielding a beat output that provides to the user a feeling of slow waves of stimulation at a frequency determined to be arousing or calming via the calibration software. Any of multiple base waveform types described as being physiologically active in the literature (the implemented device in
The Bluetooth receiver 32 pairs with whatever device generates the chosen waveforms via external software that runs on a computer or smartphone.
The battery 8 is rechargeable and is sufficient to power the amplifier 12 and transducer 16, e.g., 6000 mAh Li-Ion battery pack.
The amplifier 12 boosts the oscillation signals to a level that is useable by the transducer 16. For vibratory stimulation, which is applied as an output to the body of the user, the amplifier 12 converts the oscillation signals to a level that a 20 W 8 ohm tactile transducer can faithfully reproduce. For electrical stimulation, which is applied as an output to the user's skin, the amplifier converts audio signals to pulse-width-modulated versions (250 μs pulses separated by 5 μs gaps) to prevent skin heating, using standard algorithms, and amplifies them to a physiologically detectable threshold (approx. 2 mAmps) and includes optical isolation and voltage limitation for safety.
The vibratory tactile transducer 16 is designed to be used over any area of the body, which might respond to oscillations produced by the software. The tactile transducer 16 may also deliver whole-body vibrations by being attached to a chair or bed. By way of description and not limitation,
Another regulatory device 104 depicted generally in
Initial data supports the use of a vibratory device for regulating stress and its physiological correlates. In an experiment with N=38 individuals of whom 9 were associates and 29 were community participants who were compensated for participation, vibratory stimulation at a main frequency of 100 Hz modulated by a modulation frequency of 0.1 Hz, delivered to the wrist, improved performance reliably (p<0.05) and, to the chest, marginally (p<=0.1) during a stressful (paced auditory serial attention) task above and beyond a no-stimulation condition for those whose performance was at least moderate (above 1-standard deviation below the mean) during the no-stimulation condition. Vibration at this frequency also moderated changes in heart-rate variability (which is a proxy for parasympathetic tone), with statistically significant (p<0.05) increases in heart rate variability delivered to the wrist in the full sample and delivered to the sternum in the compensated sample (p=0.09 in the full sample). Vibration at this frequency delivered to the sternum also decreased self-reported stress in those whose stress was at least moderate (above 1 standard deviation below the mean) compared to the no stimulation condition (p<0.05).
The physiological detection suite involves using custom and commercial software and hardware to acquire physiological parameters and analyzing them in real time to detect the onset of individualized signatures of stress, fatigue, or other (e.g., user specified) emotion or arousal states.
In one implementation, software for detection of physiological states takes in pulse plethysmograph and galvanic skin response (GSR) inputs, sampled at 1000 Hz, from existing hardware (e.g., Bitalino, Arduino) implemented as generic serial streams. The pulse plethysmograph detects a heartbeat signal that is representative of the heartbeat of the user.
GSR data are preprocessed via spike removal and smoothing (4 second kernel) to yield a smooth running estimate of GSR which is associated with sympathetic nervous system reactivity and stress.
Plethysmograph data are preprocessed via spike removal and peak detection to yield heartbeats which are converted to an inter-beat interval series. The inter-beat series includes a time duration between each successive beat in the detected heartbeat signal.
The inter-beat interval sere is subjected to calculation of heart a e (#beats per second).
The inter-beat interval series for 30 seconds is subjected to continuous Morlet waveform transform to yield a running estimate of power in the high frequency heart rate variability (HF-HRV) band (0.18-0.4 Hz), which is associated with parasympathetic nervous system activity and emotion regulation capability, and which can be referred to as an emotional regulation parameter or value. It is noted that other spectral analysis techniques such as Fourier transformation and the like can be employed without departing from the spirit of the disclosed and claimed concept.
One aspect of the algorithm for detection of physiological stress includes quantifying change or slope over a period of time, i.e., 100 ms to 30 seconds, in physiological parameters to detect state onset.
The algorithm for detecting physiological stress is initially seeded for stress detection as reflecting increasing detected current physical parameters such as GSR Of heart rate without a corresponding or subsequent change in HF-HRV, i.e.: estimated-stress=β0+β1ΔGSR5 seconds+β2ΔHeart Rate5 seconds−β3|ΔHF-HRV5 seconds| where the coefficients are initially β0=β1=0.5, β2=0.5, β3=1 and GSR, HR, and HF-HRV are normalized based on their mean and variability during an initial resting calibration period of 30-seconds. Other detected current physical parameters could include a number of audio parameters that are representative of vocal stress, and other such parameters.
Another aspect of the algorithms includes a calibration operation using software that guides the user to experience resting, stressed, fatigued or other user-specified states, and which includes brief exposure to a stress induction known to provoke increases in sympathetic tone and decreases in parasympathetic tone (e.g., paced serial attention task), to yield an individually calibrated profile for these states (e.g., “stress profile”). For instance, a number of calibration physical parameters of the user such as heart rate signal, GSR signal, and other such parameters, may be detected and stored in, the storage 52. The software that performs the guiding can be executed on the regulatory device or can be deployed on a smart phone or other computerized device remote from the regulatory device.
Another aspect of the algorithms includes machine learning to derive individualized best-fit profiles for what stress-onset, fatigue onset, or other user-determined states look like for the individual. In the implemented software, as an example, a machine learning algorithm such as a three-layer pattern recognition neural network with 8 input nodes, 4 hidden nodes, and 1 output node is used to estimate how GSR, Heart Rate, HF-HRV, and estimated change in each of these in the previous 5 seconds combine to predict based on the calibration task described in the preceding paragraph. Effectively this algorithm allows a quantized (sigmoid) ridge-regression estimation of parameters for main effects of each of these parameters, and their potential n-way interactions:
estimated-stress=β0+β1ΔGSR5 seconds+β2ΔHeart Rate5 seconds+β3|ΔHF-HRV5 seconds|+β4GSR5 seconds+β5Heart Rate5 seconds+β6HF-HRV5 seconds+β7ΔGSR5 seconds*ΔHeart Rate5 seconds+β8ΔGSR5 seconds*|ΔHF-HRV5 seconds| . . . +βNΔGSR5 seconds*ΔHeart Rate5 seconds*|ΔHF-HRV5 seconds|*GSR5 seconds* Heart Rate5 seconds*HF-HRV5 seconds
To derive beta weights for the preceding equation, stress values are set to zero during rest and one (1) during the target state, e.g., stress. Thus, “estimated-stress” represents the extent to which a current state is more like the stress vs the resting state. The same type of analysis can be performed for a fatigue or user-specified period. The various p coefficients that are derived through the use of the pattern recognition neural network form a part of the individually calibrated profile that can be used to detect the onset of a period of stress or fatigue. It is understood that additional elements can be added to the above equation in order to derive additional coefficients for use with calibration physical parameters and current physical parameters that are indicative of vocal stress in the user.
Another aspect of the algorithms includes real-time comparison of incoming physiological data in the form of current physical parameters of the user to the individualized best-fit profiles to determine when an individual is beginning to look stressed, fatigued, or a critical distance from a user defined state, so as to trigger the delivery of therapeutic stimulation. Stimulation is signaled whenever the stress index is outside 1.5 standard deviations from its mean, which refers to the aforementioned threshold of “3” in
One aspect of the vocal stress detection algorithm is that speech is recorded in 5-second segments. These segments are processed to extract common vocal parameters such as speech rate, pitch, mean frequency, frequency of the first fundamental, variance of the first fundamental, etc. using publicly available code. The speech data from any 5-second segment is not saved after parameters are extracted, and thus no lasting voice recordings are made.
Another aspect is that a 4-layer pattern-network classifier was trained to recognize the emotion associated with short vocalizations (neutral, calm, happy, sad, fearful, angry, disgusted, surprised) using the “RAVDESS” speech corpus (http://smartlaboratory.org/ravdess/designfeatures/) preprocessed to be z-scores normalized by the mean of vocalizations and divided by the standard deviation. Outliers were Windsorized to the next good value outside the Tukey Hinges. The network had 15 inputs for vocal parameters, 2 hidden layers with 15 and 10 units respectively, and 8 outputs—one per classified emotion and was trained with a standard back-propagation algorithm. Classification was 30-80% accurate for specific valences depending on the valence.
Another aspect of he algorithm is that extracted speech parameters are normalized by subtracting the mean of a set of six five-second “neutral” calibration vocalizations and dividing by the standard deviation these vocalizations.
Another aspect of the vocal stress detection algorithm is that it begins with a calibration consisting of recording 30 seconds of silence in a specific room. The variance of incoming 5-second vocalizations are, at each iteration, compared to the variance of the “silence” recording. Audio waveforms with variability outside 2 standard deviations (SD) from the silence recording are considered to be vocalizations; otherwise they are considered silence and not categorized.
Another aspect of the vocal stress detection algorithm is that a second calibration records an individual person, who is the subject of measurement, speaking in a neutral tone for 30 seconds.
Another aspect of the algorithm is that when more than a user-selected number of the vocal parameters (the user can select from 2-8 parameters) are outside 2 SD from the mean of neutral vocalizations, and when the person is deemed, via classification based on the RAVDESS corpus classifier, to have a negative tone (fear, sadness, disgust), the software provides user-selected stimulation waveforms to the stimulation generator. This is depicted in the user interface capture from the software that is depicted in
The stimulation calibration algorithm performs a customization operation that involves presenting individuals with a plurality of customization stimulations in the form of a range of stimulation parameters and allowing them to rate the emotionality and arousal associated with these types of stimulation. In response to each customization stimulation, the user inputs to the software a number of responses using a single selection on a grid. The number of responses are representative of how the user perceived the customization stimulation on an arousal scale between very calming and very arousing, and are further representative of how the user tolerated the customization stimulation on a valence scale between very negatively and very positively. To best tune stimulation to an individuals' preferences, the software selects stimulation patterns based on these ratings. The pattern which is rated as maximally positive and maximally calming (sort of the squared distance on each axis from neutral) is used as the calming stimulation pattern for that individual in the event of detecting that the individual is experiencing a stress condition. The pattern which is maximally arousing, regardless of its valence, is used as the arousing stimulation pattern for that individual in the event of detecting that the individual is experiencing a fatigue condition.
Physiological parameters in the form of baseline physical parameters associated with named target emotional states (e.g., “stress” or “positive affect calm”) can be stored for later, recall as targets (triggers stimulation that decreases distance to the state when it is determined that a number of current physical parameters are more than a predetermined distance of those of the target emotional state) or alarms (triggers stimulation that increases distance from the state when it is determined that a number of current physical parameters are within a predetermined distance of those of the target emotional state).
Another aspect of the invention includes software that allows subjective and physiologically based storage of stimulation parameters that optimally yield approach or departure from target or alarm states.
Another aspect of the present invention includes software that allows users to specify potentially new or idiosyncratic target or alarm emotion states for storage in a library which includes associated physiological profiles and stimulation parameters.
Another aspect of the present invention is that physiological profiles can be used to gauge distance from normed and idiosyncratically named categories by the cosine of current physiological parameters with those for calibrated states, e.g., yielding a “Closeness” value C for each parameter, e,g., CGSR=(GSRcurrent”GSRsession_mean)*(GSR_calibration state−GSR_calibration session mean) and deriving the “Closeness” to a state as BGSR*CGSR+BHRV*CHRV+BVocal Pitch*CVocal Pitch where associated B weights are derived via neural network classifiers as described herein before.
This algorithm provides distance from target states. Before calibration, a priori rules are used to specify output transduction to optimally restore a state via minimizing distance (e.g., Euclidean) of current data from template vectors as described herein before.
Another aspect of the present invention includes software that allows users to share emotion state names, associated physiological profiles, and stimulation parameters for approaching or avoiding them.
A source for generating physiologically reactive oscillation patterns (e.g., smartphone) wherein the oscillation patterns are frequencies in the range of 20-300 Hz modulated by frequencies from 0.01-10 Hz.
The source generates oscillations and transmits them via Bluetooth.
Software is used to calibrate and store what vibration patterns maximally yield specific emotional states, including those specified by a user, such as arousal, or positive-affect calm for a given individual, or that individual's state on a given day, to which they would like to return in the future.
This software can store what vibration patterns individuals are using and dynamically update its calibration to “learn” if users choose to use patterns not suggested by the previous calibration.
A Bluetooth receiver for oscillation patterns generated by the source.
A switch, software control, or physiological/vocal measurement device to determine when the oscillation patterns are delivered.
The software computes the magnitude and slope of physiological or vocal measures over a local window (0.5-30 seconds) to determine onsets of emotion state deviations from a neutral state (e.g., stress, fatigue).
The software compares incoming physiological measurements to an individually calibrated profile to determine likelihood of onset of an emotion state and the appropriate reaction.
The software has a calibration routine that requires the individual to attain relaxed/neutral, stressed (e.g., via stressful cognitive task), or fatigued states, and record data for profile derivation.
The software uses machine learning algorithms (e.g., neural networks) to derive individually calibrated emotion state (e.g., stress) indices from the calibration data used as a training set.
The software stores learned physiological patterns in a library that can be recalled in combination with associated stimulation patterns as described to allow “restoration” of “saved” states.
A battery to power the amplifier,
An amplifier that raises the received oscillation patterns to a non-ignorable level.
Transducers that provide one of vibratory and electrical stimulation.
A sleeve for the vibratory transducer that can be removed and washed.
The sleeve may allow the device to be attached to the individual, e.g., via a band or other means for securing the device to the body part without disrupting the transducer's functionality.
Referring to
In the depicted exemplary embodiment, the regulatory device 4 or 104 includes the processor apparatus 46, which includes the processor 48 and the storage 52, with the storage 52 having stored therein the number of routines 56 which, when executed on the processor 48, cause the regulatory device 4 or 104 to perform operations such as are set forth herein. The routines 56 stored in the storage 52 serve as a non-transitory machine-readable storage medium in accordance with the disclosed and claimed concept.
The regulatory device 4 or 104 can additionally be said to include an input apparatus 7 that is structured to provide input signals to the processor 48 and an output apparatus 17 that is structured to receive output signals from the processor 48. The input apparatus 7 can be said to include, by way of example, the buttons 28, a receiver component of the Bluetooth transceiver 32, the switch 40, the dial 44, and other input devices that can control regulatory device 4 or 104 or can provide input to the processor 48, such as might include a tough-sensitive overlay of a touchscreen and other input devices without limitation. The output apparatus 17 can include structures such as the tactile transducer 16, the amplifier board 12, the LED 36, a transmitter component of the Bluetooth transceiver 32, and other appropriate output devices.
In various embodiments, the regulatory device 4 or 104 may be configured to apply the stimulation to one or more areas of the body of the person 110 that might respond to oscillations produced by the system. In various embodiments, a vibratory transducer of the device may be used on neck or sternum of the person or elsewhere on the person.
Evidence to date suggests that the stimulation 1) increases parasympathetic tone (e.g., high frequency heart rate variability) which is associated with increased capability to regulate emotions, and 2) activates cortical regions including the anterior cingulate cortex which is positively associated with increased response to antidepressant drugs. Associations of cingulate activity with SSRI response are summarized in DeRubeis, R. J., Siegle, G. J., Hollon, S., (2008) Cognitive therapy versus medications for depression: treatment outcomes and neural mechanisms. Nature Neuroscience: Reviews, 9, 788-796, PMID: 18784657. We have observed that the kinds of vibrational stimulation described elsewhere herein are associated with pupil dilation, which is further associated with activity in the anterior cingulate cortex of the brain (e.g., Breeden et al (2017) European Journal of Neuroscience, 45(2), 260-266). Such triggered activity in the anterior cingulate cortex could thus advantageously make a patent who had been resistant to an antidepressant drug less resistant to the drug and thus more able to respond to the antidepressant drug. The concept is intended to advantageously make the antidepressant drug effective in patients for whom antidepressant drugs had previously been ineffective.
An improved method in accordance with an aspect of the disclosed and claimed concept is depicted generally with a flowchart in
Additionally, such vibrational output can enhance the efficacy of an antidepressant drug in the person. For instance, the operations can additionally and optionally include, as at 325, administering an antidepressant drug to the person 1110. The operations can additionally optionally include improving, as at 335, the action of the antidepressant drug by employing the emitting of the vibrational output to stimulate activity in the anterior cingulate cortex of the brain of the person 1110. Other variations will be apparent.
The regulatory device 4 or 104 used in this application may, for example, include that described in International Patent Application No. PCT/US2017/025702 filed on Apr. 3, 2017, the disclosure of which is incorporated by reference as noted above. The features and functions described above, as well as alternatives, may be combined into many other different systems or applications. Various alternatives, modifications, variations or improvements may be made by those skilled in the art, each of which is also intended to be encompassed by the disclosed embodiments.
As used in this document, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art. As used in this document, the term “comprising” (or “comprises”) means “including (or includes), but not limited to.” When used in this document, the term “exemplary” is intended to mean “by way of example” and is not intended to indicate that a particular exemplary item is preferred or required. In this document, when terms such “first” and “second” are used to modify a noun, such use is simply intended to distinguish one item from another, and is not intended to require a sequential order unless specifically stated. The term “approximately,” when used in connection with a numeric value, is intended to include values that are close to, but not exactly, the number. For example, in some embodiments, the term “approximately” may include values that are within +/−10 percent of the value.
When used in this document, terms such as “top” and “bottom,” “upper” and “lower”, or “front” and “rear,” are not intended to have absolute orientations but are instead intended to describe relative positions of various components with respect to each other. For example, a first component may be an “upper” component and a second component may be a “lower” component when a device of which the components are a part is oriented in a first direction. The relative orientations of the components may be reversed, or the components may be on the same plane, if the orientation of the structure that contains the components is changed. The claims are intended to include all orientations of a device containing such components.
An “electronic device” or a “computing device” refers to a device or system that includes a processor and memory. Each device may have its own processor and/or memory, or the processor and/or memory may be shared with other devices as in a virtual machine or container arrangement. The memory will contain or receive programming instructions that, when executed by the processor, cause the electronic device to perform one or more operations according to the programming instructions. Examples of electronic devices include personal computers, servers, mainframes, virtual machines, containers, gaming systems, televisions, digital home assistants and mobile electronic devices such as smartphones, fitness tracking devices, wearable virtual reality devices, Internet-connected wearables such as smart watches and smart eyewear, personal digital assistants, cameras, tablet computers, laptop computers, media players and the like. In a client-server arrangement, the client device and the server are electronic devices, in which the server contains instructions and/or data that the client device accesses via one or more communications links in one or more communications networks. In a virtual machine arrangement, a server may be an electronic device, and each virtual machine or container also may be considered an electronic device. In the discussion below, a client device, server device, virtual machine or container may be referred to simply as a “device” for brevity.
The terms “processor” and “processing device” refer to a hardware component of an electronic device that is configured to execute programming instructions. Except where specifically stated otherwise, the singular terms “processor” and “processing device” are intended to include both single-processing device embodiments and embodiments in which multiple processing devices together or collectively perform a process.
The terms “memory,” “memory device,” “data store,” “data storage facility” and the like each refer to a non-transitory device on which computer-readable data, programming instructions or both are stored. Except where specifically stated otherwise, the terms “memory,” “memory device,” “data store,” “data storage facility” and the like are intended to include single device embodiments, embodiments in which multiple memory devices together or collectively store a set data or instructions, as well as individual sectors within such devices.
As used herein, the term “treat”, “treating” or “stimulating” refers to enhancing a person's positive outlook or suppressing a person's negative outlook. This may refer to a person's psychological well-being, including but not limited to their emotional, cognitive, and motivational states.
In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word “comprising” or “including” does not exclude the presence of elements or steps other than those listed in a claim. In a device claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The word “a” or “an” preceding an element does not exclude the presence of a plurality of such elements. In any device claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The mere fact that certain elements are recited in mutually different dependent claims does not indicate that these elements cannot be used in combination
Although the invention has been described in detail for the purpose of illustration based on what is currently considered to be the most practical and preferred embodiments, it is to be understood that such detail is solely for that purpose and that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover modifications and equivalent arrangements that are within the spirit arid scope of the appended claims. For example, it is to be understood that the present invention contemplates that, to the extent possible, one or more features of any embodiment can be combined with one or more features of any other embodiment.
This application is a Continuation-in-Part Application of U.S. patent application Ser. No. 16/088,608, filed on Sep. 26, 2018, which is a U.S. National Stage Application of PCT/US2017/025702, filed on Apr. 3, 2017, which claims priority from U.S. Provisional Patent Application No. 62/316,695, filed on Apr. 1, 2016, and also claims priority from U.S. Provisional Patent Application No. 62/742,861, filed on Oct. 8, 2018, all of whose contents are incorporated herein by reference.
This invention was made with government support under grant #s MH082998 and MH096334 awarded by the National Institute of Health and under grant #IIP-1449702 awarded by the National Science Foundation. The government has certain rights in the invention.
| Number | Date | Country | |
|---|---|---|---|
| 62316695 | Apr 2016 | US | |
| 62742861 | Oct 2018 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 16088608 | Sep 2018 | US |
| Child | 16596351 | US |
