Patent Application
20220386954
The present invention generally relates to biosensors. More specifically, the present invention relates to cumulative biosensor systems for detecting alcohol.
The usage of alcohol has many negative health and economic effects on individuals as well as society. In the United States (US) alone, excessive alcohol use led to approximately 90,000 deaths in the last year and is responsible for 10% of deaths in working-age adults, making alcohol-related deaths the fourth leading preventable cause of death (4). In addition, the economic burden of excessive alcohol consumption is enormous. In 2006, this cost was estimated at $223.5 billion in the US, or approximately $700 per individual (5).
The current market for alcohol monitoring technologies is primarily catered to alcohol enforcement at the government and business level. And while there are consumer grade options on the market, these available devices have major practical limitations for widespread use. They are bulky, electronics-dependent, and expensive. Even so called personal breathalyzers are only small enough to fit on a keychain. Moreover, these devices require active and “correct” use by the customer; i.e. customers must blow into the device in a specific way, for a set amount of time, or they may get an incorrect reading (6). This active use may also carry a strong stigma preventing true widespread use. The ability to conceal or personalize current breathalyzers and transdermal BAC devices does not exist today, as their electronics dictate a certain arrangement and size. Finally, since breathalyzers are not continuously sensing devices, they cannot offer a continuous cumulative readout of one's alcohol consumption.
Alcohol may further be linked to certain crimes, including date rapes which inflict a profound emotional and physiological impact on victims. Recent studies, conducted by the University of Iowa, (7) which include victims of attacks at US college campuses, highlight both the emotional and staggering victimization costs (up to $151,423 in one instance).
There is, therefore, a need in the art for improved systems and methods for cumulative biosensor-based detection of alcohol.
The present invention includes a wearable biosensor that detects cumulative alcohol consumption, as well as an associated computing device application. In particular, the wearable biosensor is capable to detecting cumulative consumption of alcohol over time (e.g., since the biosensor was applied). As such, the biosensor readings differs from presently known methods of measuring alcohol consumption (e.g., Breathalyzer), which reflect only a current level of alcohol content. The biosensor described herein, however, reflects a cumulative amount of alcohol that has been consumed over the course of time since the biosensor was first applied (e.g., throughout the course of an evening). The detected accumulation of alcohol may be reflected in a variety of ways, including colors (and gradients of the same), as well as other visual and sensory means.
In one embodiment, an alcohol monitoring solution is offered that addresses many problems discussed above. Furthermore, suchg solution may appeal to a significantly larger market and user demographic, as learned from surveying 250 subjects across age, sex, education, drinking habits, and geography in the US. For example, a statistically significant indicator from these surveys was that women who drink casually are the most likely user segment to wear an electronics-free temporary tattoo that indicates alcohol levels via colorimetry. The form factor may serve as an adequate safety net, for example, during a date.
The ability to monitor alcohol consumption in a cumulative fashion by simple means has the potential to raise awareness of alcohol usage and could promote a healthier lifestyle across a wide user demographic, including young adults. For example, in cases where an individual's habitual drinking is gradually becoming more common, the device described herein could help identify and prevent a nascent negative trend developing into a severe health condition. In general, by addressing alcohol-related problems at their onset and throughout their development, the present invention may contribute more effectively to the mitigation of the health and economic burdens that such alcohol-related problems eventually devolve into.
Some embodiments may include a wearable instantaneous alcohol biosensor, as well as a related biosensor that detects cumulative alcohol levels. To allow for the option of tracking values electronically over longer periods of time, computer vision algorithms may be developed and trained to permanently record and interpret cumulative alcohol levels for the individual. Electronic tracking may allow for further guidance and awareness, and enabling anonymized data collection for research purposes.
The present invention includes a wearable biosensor that detects cumulative alcohol consumption, as well as an associated computing device application. In particular, the wearable biosensor is capable to detecting cumulative consumption of alcohol over time (e.g., since the biosensor was applied). As such, the biosensor readings differs from presently known methods of measuring alcohol consumption (e.g., Breathalyzer), which reflect only a current level of alcohol content. The biosensor described herein, however, reflects a cumulative amount of alcohol that has been consumed over the course of time since the biosensor was first applied (e.g., throughout the course of an evening). The detected accumulation of alcohol may be reflected in a variety of ways, including colors (and gradients of the same), as well as other visual and sensory means.
In addition, an associated computing device application may be optionally used with the wearable biosensor to track alcohol consumption trends over longer periods of time. Such an application may receive input regarding the reaction exhibited by the wearable biosensor indicating a cumulative amount of alcohol that has been detected. Such input may occur in the form of camera image, user input (e.g., selecting from a menu), or a wireless communication (e.g., in embodiments where the wearable biosensor may be coupled to a radio transmitter). Such input may be evaluated to determine what level of cumulative alcohol consumption is indicated and stored. The input may further be analyzed in context with previous input associated with the same user to identify trends, anomalies, and various other health or physical conditions of interest. In some embodiments, the application may further be trained and calibrated for more accurate or more detailed, granular results.
Some embodiments may include a wearable, electronics-free, cumulative alcohol biosensor, and an optional mobile companion app.
Passive, continuous, and cumulative. Effortlessly raises awareness about a user's overall alcohol consumption patterns and how they may be changing over time.
Easy to wear, learn, and use. Worn much like a regular temporary tattoo or sticker, a wearable biosensor relies on the observation of color changes with the unaided eye, without strictly requiring the use of a separate device or application to understand the sensor; i.e., the tattoo is the user interface. A smartphone application may be offered to users as an option, not a requirement, thus maintaining a low threshold for user adoption while allowing smartphone users to track values over longer periods of time, and eventually providing other services and up sale opportunities.
Low cost. Whereas currently available solutions are electronics-dependent, the biosensors described herein are not.
Greater privacy and control of information encourages more people to wear an alcohol biosensor. Wearable alcohol biosensors can have a semi-arbitrary or user-specific design, potentially concealing information such that only the user understands its meaning. It can also be placed on different parts of the body, e.g. in the underarm or on the chest. When noticed, it bypasses the perception of being a medical device, thereby reducing the associated risk of stigmatization.
Design and characterization of a cumulative alcohol biosensor.
One embodiment may include an instantaneous electronics-free, transdermal blood alcohol content (BAC) biosensor in the form of a temporary tattoo. A wide variety of ethanol-sensing formulations may be provided based on the particular design of the biosensor. Such formulations may not only sense ethanol at a point in time, but also sense and represent its accumulation over time. The design of the temporary tattoo may therefore be configured so as to be present a visual indication system of such accumulation.
A toolkit of the temporary tattoo form factor and novel active ink formulations may be used to create a biosensor that is driven by the amount of alcohol consumed since the sensor was applied. At the core of the sensor is the active ink formulation. The formulation is based on the reaction of alcohol dehydrogenase (ADH), which alongside color changing tetrazolium salts form the active ink. Upon ethanol reaching the active ink containing ADH, the enzyme metabolizes ethanol to acetaldehyde and reduces NAD+ to NADH. NADH then transfers electrons through the electron acceptor phenazine methosulfate (PMS) to a tetrazolium salt, yielding a colored product (8).
Alongside the active ink, a series of graded reference colors may be printed. These colors may allow the users to immediately develop a qualitative sense of their cumulative alcohol consumption with the unaided eye.
An active ink formulation may include a tetrazolium salt such as 3-(4,5-Dimethylthiazol2-yl)-2,5-Diphenyltetrazolium Bromide (MTT), as well as commercially available ADH and PMS. The ink may be applied to the tattoo paper via silk screening. Because the product of MTT reduction, formazan, is insoluble, its formation is an irreversible reaction, resulting in increased color change as more ethanol is absorbed from the sweat (9). Thus, as the user consumes more ethanol, more ethanol is excreted in the sweat over a longer period of time and collected by the sensor, resulting in increased color intensity that can be used as a readout of cumulative consumption. Ethanol excretion in sweat is mainly independent of sweat flow rate (10), thus producing a robust cumulative metric across different conditions.
To characterize the formulations of an colorimetric reaction mixture, solution mixtures may be assayed in 24 and 48 well plates using spectrophotometry in a Tecan M200 plate reader. Standardized 0 to 0.3% ethanol solutions may be used to confirm colorimetric readout changes by absorbance at 540-570 nm with the use of the insoluble MTT. Reproducibility may be tested across different types of sweat by measuring the same 0 to 0.3% ethanol solutions in 10-100 μL solutions of 0-100 mM Na+ and Cl− to mimic the known ranges in sweat concentrations (11,12). Solutions of manually collected sweat spiked with 0 to 0.3% ethanol may subsequently be applied to the wells. Reversible ethanol sensing reactions may be characterized as presented in
Having formulated the chemical reaction mixture, different patterns may be screen printed onto paper substrates to test the kinetics of the colorimetric reaction mixture. Screen printing may require high viscosity and thixotropy of the printed fluid, previously developed and the use of food-grade thickeners with established safety profiles including xanthan gum, gum Arabic, corn starch, tapioca starch and carboxymethylcellulose (CMC). Rigorous characterization of the screen printing resolution properties may be done using microscopy and other approaches, determining that a 4% w/v carboxymethylcellulose (CMC) solution results in appropriate rheological properties with a uniform suspension of the reaction mixture. Using this mixture as a starting point for the cumulative alcohol biosensor, the kinetics and color changes of the ethanol sensing reaction may be assayed on transfer paper using time-lapse photography, and then analyze the color changes over time using ImageJ (13). The same procedure may be followed as with the previous mixtures, assaying solutions with varied ethanol and salt contents as well as sweat with predefined ethanol content.
Data analysis and statistics: The data generated in this section may resemble the data presented in
Expected outcomes and alternatives: Developing and characterizing the cumulative alcohol-level sensor may be completed within six months. Reformulation for the cumulative ethanol readout with the irreversible dye MTT may follow a similar path, but may require adjustment of the chemical formulations and kinetics for creating the appropriate dynamic range, including grading the reference colors. Determining such range is a key factor in determining the final active product life since it limits the number of drinks before the wearable needs to be renewed. A limit may be defined high enough for at least two days of heavy drinking. Also, if the kinetics of the MTT reaction are not desirable, other salts, like nitro blue tetrazolium (NBT) (15), may be used. Because it is inert, the use of CMC as a resuspension medium for screen-printing is not expected to affect the chemical reactions in the mixture, but if there are any undesired effects, other thickeners like those listed above may be used. If screen printing produces solutions that are too viscous and inhibit the reaction kinetics, dilution may occur before screen printing, or an alternative bio-printing inkjet method for deposition may be used. Lastly, if the reaction with ADH does not function as expected, other enzymes such as alcohol oxidase (AOX), which is known to be compatible with such salts, may be used.
If it is determined that the graded reference colors are not an ideal readout system, a different gradation approach may be used. In such an approach, the sensor may consist of multiple printed sections that reflect a gradient of values, like an ‘energy bar” interface (
Another embodiment may include an optimized sweat collection layer that operates under diverse conditions, and that prevents the active ink from being in direct contact with the skin. Such a product may be tested in controlled and observational studies involving a representative sample of human subjects. Several sweat collecting—and inducing—methods (16-20) exist that are compatible with the methods described herein. More active ink colors using different salts may be used, thus enabling further customization opportunities. Further embodiments may use a gradient tattoo UI when designing a biosensing temporary tattoo-like device. Thus, a user may customize their tattoo design to include (1) one more discrete and concealed (i.e. single tattoo design element changing in color) or (2) more explicit and potentially easier to understand (i.e. the gradient-based ‘energy bars’ approach).
Unaided-eye interpretation of color changes in sensor representing the cumulative levels of simulated transdermal alcohol concentrations over a 48-hour period with 90% accuracy.
Development of computer vision algorithms to track cumulative alcohol levels.
Wearable alcohol biosensors do not require electronic devices to read cumulative alcohol concentration values, thus contributing to targeting of a wider user demographic when compared to the current state of electronics-dependent devices. However, some users may prefer to use a smartphone app to track cumulative values over longer periods of time.
By giving the option to these users to capture and store values electronically, a future application may also provide other services such as trends analysis, and provide guidance or coaching
Recognizing tattoos and interpreting them is an active area of research (21,22,23,24), faced with challenges that include their arbitrary design, and arbitrary placement—and deformation—when applied on different skin surfaces. An important part of the existing research focuses on features outside of the scope of alcohol biosensing (e.g. in criminology). A computer vision pipeline may be provided that is tailored to reading biomarker values and that tracks cumulative concentration values from a semi-arbitrarily designed temporary tattoo-like cumulative alcohol biosensor that can be placed on different parts of the body. Such a pipeline is composed of five modules: capture, detection, identification, interpretation, and storage.
The first module (
Training of the CNN may happen initially on temporary tattoos printed with regular inks. Training may always include a representative sample of skin types along the Fitzpatrick phototyping scale. A third module (
At this point, the reference shape may be populated with normalized color values from the extracted shape across its surface area. To assess the cumulative alcohol concentration levels, calibration may be done by running a more granular sliding window of n×n pixel size to traverse the color-populated reference shape. In each new region, color value may be matched against the color versus concentration curves calibrated as discussed above. Additionally, the color curves may be color-normalized before use. Finally, the last module (
Expected outcomes and alternatives: The various aspects described herein that can be applied to a range of sensors sharing the temporary tattoo-like form factor, including the cumulative alcohol monitoring biosensor described herein, or the previously developed instantaneous alcohol monitoring sensor. Modules may be added or improved while also training and calibrating the pipeline for the specific cumulative alcohol sensing application described herein. Highly qualitative data of cumulative alcohol values may be collected and used for training and calibrating until there is a more robust and semi-quantitative solution in place, which may be specific to a user. Reference colors may be used on the temporary tattoo, using the color versus concentration curves recorded as discussed above, thus allowing for an increase in accuracy. As described herein, calibrating the pipeline may allow for capture of a gradient tattoo UI (i.e. the ‘energy bars’ approach), where assessment of biosensor values may be easier. Also, because different smartphone models have different native RGB color spaces, color transformation techniques may be applied to a common color space using the full camera profile when available, or by using a standard color matrix.
Convolutional Neural Network (CNN) described for the detection module may be trained, thereby further increasing its robustness to detect a temporary tattoo at a wider range of camera angles, lighting conditions, skin types, and parts of the body. Other aspects of the pipeline may be enhanced including local storage, which may happen under a commercial setting integrating with services provided by third party companies or platforms such as Apple HealthKit (34). In general, the companion application may be robust enough to be ready to be used by consumers, initially during controlled and observational studies, and later in a commercial setting. Being user privacy protection a high concern for any consumer-oriented alcohol monitoring application, in addition to the use of standard user data encryption methods, commercially available services that involve Differential Privacy (35) may be integrated in systems described herein, where the information stored in remote servers is randomly distorted in known statistical quantities, thus preventing anybody from pinpointing specific user data to an individual.
Computer vision interpretation of data color changes in sensor representing the cumulative levels of simulated transdermal alcohol concentrations over a 48-hour period with 90% accuracy.
The detailed description of the technology herein has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the technology to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. The described embodiments were chosen in order to best explain the principles of the technology and its practical application to thereby enable others skilled in the art to best utilize the technology in various embodiments and with various modifications as are suited to the particular use contemplated. It is intended that the scope of the technology be defined by the claim.
The present patent application is a continuation of U.S. patent application Ser. No. 16/528,362 filed Jul. 31, 2019, now U.S. Pat. No. 11,350,875, which is a continuation of International patent application number PCT/US18/16281 filed Jan. 31, 2018, which claims the priority benefit of U.S. provisional patent application No. 62/452,579 filed Jan. 31, 2017, the disclosures of which are incorporated by reference herein.
| Number | Date | Country | |
|---|---|---|---|
| 62452579 | Jan 2017 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 16528362 | Jul 2019 | US |
| Child | 17834143 | US | |
| Parent | PCT/US2018/016281 | Jan 2018 | US |
| Child | 16528362 | US |
