You can monitor your hydration level using your smartphone camera

TL;DR

This work introduces a non-invasive dehydration monitoring method using video-PPG signals extracted from fingertip footage captured by a standard smartphone camera. It builds a labeled dataset from fasting subjects during Ramadan and develops ML, DL, and transformer models to perform dehydration detection (binary) and dehydration level classification on a 1–4 scale, achieving up to 99% accuracy. A low-complexity alternative using LSTM-FCN feature extraction plus t-SNE is proposed, along with SHAP-based explainability to interpret model decisions. The approach demonstrates high accuracy, feasibility for edge deployment, and broad practical impact for at-home health monitoring, smart homes, and portable health assessment.

Abstract

This work proposes for the first time to utilize the regular smartphone -- a popular assistive gadget -- to design a novel, non-invasive method for self-monitoring of one's hydration level on a scale of 1 to 4. The proposed method involves recording a small video of a fingertip using the smartphone camera. Subsequently, a photoplethysmography (PPG) signal is extracted from the video data, capturing the fluctuations in peripheral blood volume as a reflection of a person's hydration level changes over time. To train and evaluate the artificial intelligence models, a custom multi-session labeled dataset was constructed by collecting video-PPG data from 25 fasting subjects during the month of Ramadan in 2023. With this, we solve two distinct problems: 1) binary classification (whether a person is hydrated or not), 2) four-class classification (whether a person is fully hydrated, mildly dehydrated, moderately dehydrated, or extremely dehydrated). For both classification problems, we feed the pre-processed and augmented PPG data to a number of machine learning, deep learning and transformer models which models provide a very high accuracy, i.e., in the range of 95% to 99%. We also propose an alternate method where we feed high-dimensional PPG time-series data to a DL model for feature extraction, followed by t-SNE method for feature selection and dimensionality reduction, followed by a number of ML classifiers that do dehydration level classification. Finally, we interpret the decisions by the developed deep learning model under the SHAP-based explainable artificial intelligence framework. The proposed method allows rapid, do-it-yourself, at-home testing of one's hydration level, is cost-effective and thus inline with the sustainable development goals 3 & 10 of the United Nations, and a step-forward to patient-centric healthcare systems, smart homes, and smart cities of future.

Figures (13)

• Figure 1: Pictorial overview of our proposed smartphone-based method for non-invasive dehydration monitoring.
• Figure 2: Architectures of the purpose-built deep learning models (LSTM-FCN and BiLSTM-FCN) for dehydration level classification.
• Figure 3: Architecture of the purpose-built DistilBERT model for the downstream task of dehydration level classification.
• Figure 4: Architecture of the purpose-built 1D-ViT model for the downstream task of dehydration level classification.
• Figure 5: Accuracy comparison of the various ML models, i.e., KNN-Euclidean, KNN-Cosine, KNN-Cubic, Ensemble-Subspace KNN, Ensemble-Bagged Trees, Kernel-SVM, Kernel-LR, SVM-Fine Gaussian, 2x10 NN, 1x100 NN, and 3x10 NN classifiers for hydration level classification. Sub-fig. (a) compares the accuracy of the ML models for binary classification, while sub-fig. (b) compares the accuracy of the ML models for 4-class classification. For both sub-figures, we also evaluate the impact of data augmentation (by a factor of 2, 3, and 4) on the accuracy of the ML models.