Cumulative biosensor system to detect alcohol

1. A wearable system for detecting cumulative ethanol consumption, the system comprising:

an adhesive layer that adheres to skin and that allows sweat from the skin to pass through;

an ink layer comprising a plurality of sections that each reacts in response to a different level of ethanol content in the sweat along a gradient, wherein each of the sections changes in appearance based on the respective reaction at the respective different ethanol content level;

a processor that executes stored instructions in memory, wherein the processor compares an image captured of each of the sections under a set of environmental conditions to that of one or more reference shapes stored in a remote database by using a transformation function and interprets one or more changes based on use of artificial intelligence and the set of environmental conditions; and

a mobile application configured to receive customization input, wherein the ink layer is printed based on the customization input received through the mobile application, wherein the customization input defines one or more masked image regions that are compared to one or more reference parameters of one or more reference shapes stored in a database.

2. The system of claim 1 , further comprising a camera that captures the image of the plurality of sections after the reaction in response to detection of ethanol consumption.

3. The system of claim 2 , wherein the image is normalized to eliminate effects of illumination.

4. The system of claim 2 , wherein the image is normalized to preserve inherent colors of the image and colors of one or more reference colors stored in a database memory.

5. The system of claim 2 , wherein the image is normalized to preserve color variations associated with a cumulative amount of ethanol consumption.

6. The system of claim 2 , wherein the camera captures the image of the plurality of sections within a range of camera angles.

7. The system of claim 1 , wherein each of the plurality of sections are evaluated relative to a corresponding reference parameter of the one or more stored reference shapes.

8. The system of claim 1 , wherein each of the plurality of sections are classified based on a sliding window of a convolutional neural network for image patch labeling.

9. The system of claim 8 , wherein classifying each of the plurality of sections includes extracting one or more masked image regions that define the plurality of sections.

10. The system of claim 9 , wherein extracting the masked image regions includes detecting keypoints that define a shape of the extracted masked image regions.

11. The system of claim 10 , wherein the shape of the extracted masked image region is defined as a deformation of shape of the one or more reference shapes.

12. The system of claim 10 , wherein the shape of the one or more reference shapes is mapped to the shape of the extracted masked image region by the transformation function.

13. The system of claim 1 , wherein comparing the plurality of sections to the one or more reference shapes comprises:

detecting a color value of at least one section; and

matching the detected color value to a stored chart of calibrated color concentration values.

14. The system of claim 13 , wherein matching the detected color value to the stored chart includes comparing the detected color value to a representative sample of skin types along a Fitzpatrick phototyping scale.

15. The system of claim 1 , wherein the ink layer is printed based on customization input to include different salts that induce a different change in appearance in response to ethanol content in the sweat.

16. The system of claim 1 , wherein the ink layer is printed based on customization input to include a range of different shapes.