Fig 1.
(A) Construction of the UV sensor (from the top to the bottom): protective liner with adhesive, permeable polyurethane (TPU, 16 μm) with printed UV ink, UV blockers and reference colors, top skin adhesive layer (25 μm), NFC antenna (yellow, 18 μm) and polyimide film encapsulation (PI, 12.7 μm), NFC antenna and chip (0.5 mm), polyethylene terephthalate layer (PET, 12 μm), bottom skin adhesive layer (25 μm), and bottom liner. (B) The front of the UV patch. (C) The back of the UV patch. Bar = 10 mm. (D) Wearing the UV patch on the back of one’s hand. (E) Reading the UV patch using the My UV Patch app.
Fig 2.
The mechanism of the UV sensor color change and color change quantification.
(A) The UV patch is composed of a series of reference colors 1 to 10, UV variable ink squares 11 to 16, and UV reversible ink squares 17 and 18. The reference colors 1–10 correspond to the different colors of the UV ink squares when they are exposed to UV radiation. (B) The six UV sensitive ink squares change colors at distinctive rates when exposing to UVA radiation with square 11 being the most sensitive and square 14 being the least sensitive. The color change is quantified in CIE Lab color space. (C) Schematics showing the UV patch before and after exposure to UVA radiation.
Fig 3.
The app algorithm flowchart.
Table 1.
Personal UV daily sunstock.
Table 2.
Personal UV risk determination.
Fig 4.
Comparison of the UV patch and Scienterra UV dosimeter.
10 UV doses were examined, 3 replicates for each UV dose, p<0.00001.
Fig 5.
Clinical evaluation of the UV patch.
(A) The study subjects wore the UV patches and Scienterra dosimeters during regular city and beach activities. Both devices showed agreement in UV dose measurements. (B) The study subjects conducted controlled activity: single file walk in specified directions. The activity was repeated in the morning, afternoon, and evening. Each study subject wore one Scienterra dosimeter and two UV patches: one without sunscreen and the other one with sunscreen applied on it. Both the electronic dosimeter and the UV patch without sunscreen showed consistent results. p < 0.0001, n = 24. The UV patch covered with sunscreen showed significant reduction in measured UV radiation.
Fig 6.
Comparison of UV readings among UV patch image analysis, Scienterra dosimeter and the mobile application.
UVA readings by patch picture analysis showed high correlation with Scienterra dosimeter readings, which validates the UV sensor image technique p < 0.0001, r = 0.88, n = 30 (A). When compared between the patch picture analysis and app reading, the data still shows good correlation but the fast patch scanning requirement for improved user experience affected data quality p < 0.0001, r = 0.92, n = 30 (B). Similar result is shown between Scienterra dosimeter and app reading p < 0.0001, r = 0.92, n = 24 (C). The total UV dose shows a good correlation between the Scienterra dosimeter and patch picture analysis p < 0.0001, r = 0.87, n = 24 (D). The 95% prediction ellipse is shown. The strong correlation among the three measurements further validates the sensor system.
Fig 7.
(A) A world average UV exposure is generated based on the My UV Patch app user data from June 6th, 2016 to August 18th, 2016. Zoom in maps are shown for continental US (B) and part of Europe (C). The country and state that contributed the data are labeled in yellow to red, color map is generated by normalizing the UV exposures to range between 0 (minimum UV exposure, yellow) and 1 (maximum UV exposure, red).