Non-enzymatic colorimetric detection of hydrogen peroxide using a μPAD coupled with a machine learning-based smartphone app

Abstract

Here, a smartphone app named Hi-perox Sens supported by machine learning classifiers was applied to a μPAD based on an iodide-mediated TMB-H2O2 reaction system for non-enzymatic colorimetric determination of H2O2.

A field-deployable water quality monitoring with machine learning-based smartphone colorimetry

“…Feature extraction is the process of obtaining distinctive features representing an object based on color, texture, size, shape, and location and is an essential step in training ML classifiers. 34,51 Here, image features are extracted based on color and texture information only, as it has been found to be adequate based on extensive experimentation. After the region of interest (ROI) was cropped, it was converted from RGB (red-green-blue) to HSV (hue-saturation-value) and L*a*b* (lightness, green-red, blue-yellow) for each concentration to obtain in R, G, B, H, S, V, L*, a*, and b* color channels separately as HSV is more robust towards external lighting changes, and L*a*b* is particularly useful for boosting colors in images due to the way it handles colors, which offers more distinguish features for the training of machine learning classifiers.…”

“…Feature extraction is the process of obtaining distinctive features representing an object based on color, texture, size, shape, and location and is an essential step in training ML classiers. 34,51 Here, image features are extracted based on color and texture information only, as it has been found to be adequate based on extensive experimentation. Aer the region of interest (ROI) was cropped, it was converted from RGB (red- Then, the input sample is assigned to the class label based on the majority voting of its nearest neighbors.…”

“…In this regard, articial intelligence (AI) methods have been used to enable the learning of computing devices without a need for human intervention. 30,31 The advances in AI help to improve the measurement techniques in various elds such as environmental data monitoring, [32][33][34][35] weather forecasting, 36,37 medical imaging, 38 diagnosis, [39][40][41] agricultural management, 42,43 and disaster prediction. 44,45 Tinelli et al simulated the bio-contamination risk in the water distribution systems.…”

Beyin Bilgisayarlı Tomografi Görüntülerinde Yapay Zeka Tabanlı Beyin Damar Hastalıkları Tespiti

“…The dataset plays a significant role to achieve a robust model in deep learning-based systems (Doğan, Yüzer, Kılıç, & Şen, 2021). This study was conducted with the publicly available dataset (Koç et al, 2022).…”

Machine Learning-Based Quantification of (−)-trans-Δ-Tetrahydrocannabinol from Human Saliva Samples on a Smartphone-Based Paper Microfluidic Platform

“…For practical applications, the amount of THC in saliva or blood inevitably varies from person to person . By uncovering the underlying patterns from training data, machine learning (ML) algorithms can overcome interferences generated by proteins and other molecules in saliva and blood as well as sample-to-sample variances. , Several well-known ML algorithms, such as support vector machines (SVMs), k -nearest neighbors ( k -NNs), random forests (RFs), Bayesian networks, and Gaussian networks, have been used in many genomics, proteomics, and other biological applications . For example, Nakano et al trained a deep learning model and an SVM algorithm to classify oral malodor and healthy breath from oral microbiota in saliva from 16S rRNA sequences .…”