Claim: We claim:
1. A Non-Invasive Dehydration Monitoring and Alert System Using Electro dermal Activity System comprises:wearable Electrodermal Activity (EDA) sensors in conjunctionwith signal processing and machine learning techniquesto develop first time ever a dehydration self-monitoring tool,Monitoring My Dehydration (MMD), that can instantly detect thehydration level of human skin.
2. The Device as claimed in claim I, further comprising of alternating arrays of printed light-emitting diodes and photodetectors, can detect blood oxygen levels in any part of the body. The sensor uses light-emitting diodes to emit red and nearinfrared light, penetrating the skin and detecting the proportion of reflected light. The sensor made of biodegradable materials utilizes edge-field capacitance technology to monitor arterial blood and then transmits the data wirelessly.
3. The Device as claimed in claim 1, further comprising of , the system of the wireless battery-free sweat sensor, the process of detecting during sweating, and the wireless transmission and analysis interface of sweat composition on the smart phone.A high sensitivity throat microphone is used to record audio signals from the throat area. The microphone absorbs vibrations from the wearer's throat instead of sound signals which allows picking up sounds even in extremely noisy and windy environments.
4. The Device as claimed in claim lused various machine learningmethods, such as logistic regression, support vector machines,decision trees, and K-Nearest Neighbor (KNN) classifiersto model the data and predict the dehydration of the user. Additionally, a variety of physiological measures were takenof the subjects to confirm that they were, indeed, mildlydehydrated.
5. The Device as claimed in claim 1 starts with the Empatica E4 device, which collectsraw data about the user and does some basic preprocessing - calculating Skin Conductance from EDA signals,non-negative sparse deconvolution to extract components ofEDA signal. It then sends this data to the user's smartphoneover Bluetooth.
6. The Device as claimed in claim 1-6 contains Android application on the user's phonethen preprocesses the data by using basic statistical methodsto help remove noise from the data. Once this is done, thedata is fed to a pre-trained machine learning model using theWaikato Environment for Knowledge Analysis.
7. The Device as claimed in claim 1-7 has the process as, if the machine learning model predicts achange in hydration level, it will trigger a method that sendsthe user a notification to alert them of their changed hydrationlevel.
8. The Device as claimed in claim 1-7 has Skinconductance Base Level (SBL), which changes slowly overtime (tonic changes) and indicates the general activation of thesympathetic nervous system.
9. The Device as claimed in claim 1-7 has Skin Conductance Responses(SCRs), changes that last for shorter periods (phasic changes). SCRs indicate the activation of the somatic nervous system(SNS) but also reflect responses to events that are new, unexpected,relevant, and/or aversive.