Tap water fingerprinting using a convolutional neural network built from images of the coffee-ring effect

Detection of milk adulteration using coffee ring effect and convolutional neural network

A low-cost and effective method is reported to identify water and synthetic milk adulteration of cow's milk using coffee ring patterns. The cow's milk samples were diluted with tap water (TW), distilled water (DW) and mineral water (MW) and drop cast onto glass slides to observe coffee ring patterns. The area of the ring, total particle area and average particle diameter were extracted from these patterns. For each ring, the ratio of total particle area versus total ring area was calculated. The area ratio, regardless of water adulterants, follows an exponential model with respect to average particle diameter. Unlike TW, the ratio for DW and MW adulterated milk are clustered and classified together with respect to the particle diameter. These results were independent of dilution level and are used for adulterant classification. The ring of milk adulterated using synthetic milk gave multiple concentric rings, flower-like structures, and oil globules throughout the dilution level. An Alexnet model was used to classify water and synthetic milk adulterants in authentic milk. The trained model could achieve 96.7% and 95.8% accuracy for binary and tertiary classification respectively. These results enable us to distinguish synthetic milk from pure milk and segregate DW and MW with respect to TW adulterated milk.

Drying of bio-colloidal sessile droplets: Advances, applications, and perspectives

Drying of biologically-relevant sessile droplets, including passive systems such as DNA, proteins, plasma, and blood, as well as active microbial systems comprising bacterial and algal dispersions, has garnered considerable attention over the last decades. Distinct morphological patterns emerge when bio-colloids undergo evaporative drying, with significant potential in a wide range of biomedical applications, spanning bio-sensing, medical diagnostics, drug delivery, and antimicrobial resistance. Consequently, the prospects of novel and thrifty bio-medical toolkits based on drying bio-colloids have driven tremendous progress in the science of morphological patterns and advanced quantitative image-based analysis. This review presents a comprehensive overview of bio-colloidal droplets drying on solid substrates, focusing on the experimental progress during the last ten years. We provide a summary of the physical and material properties of relevant bio-colloids and link their native composition (constituent particles, solvent, and concentrations) to the patterns emerging due to drying. We specifically examined the drying patterns generated by passive bio-colloids (e.g., DNA, globular, fibrous, composite proteins, plasma, serum, blood, urine, tears, and saliva). This article highlights how the emerging morphological patterns are influenced by the nature of the biological entities and the solvent, micro- and global environmental conditions (temperature and relative humidity), and substrate attributes like wettability. Crucially, correlations between emergent patterns and the initial droplet compositions enable the detection of potential clinical abnormalities when compared with the patterns of drying droplets of healthy control samples, offering a blueprint for the diagnosis of the type and stage of a specific disease (or disorder). Recent experimental investigations of pattern formation in the bio-mimetic and salivary drying droplets in the context of COVID-19 are also presented. We further summarized the role of biologically active agents in the drying process, including bacteria, algae, spermatozoa, and nematodes, and discussed the coupling between self-propulsion and hydrodynamics during the drying process. We wrap up the review by highlighting the role of cross-scale in situ experimental techniques for quantifying sub-micron to micro-scale features and the critical role of cross-disciplinary approaches (e.g., experimental and image processing techniques with machine learning algorithms) to quantify and predict the drying-induced features. We conclude the review with a perspective on the next generation of research and applications based on drying droplets, ultimately enabling innovative solutions and quantitative tools to investigate this exciting interface of physics, biology, data sciences, and machine learning.

Patterns from dried drops as a characterisation and healthcare diagnosis technique, potential and challenges: A review

When particulate-laden droplets evaporate, they leave behind complex patterns on the substrate depending on their composition and the dynamics of their evaporation. Over the past two decades, there has been an increased interest in interpreting these patterns due to their numerous applications in biomedicine, forensics, food quality analysis and inkjet printing. The objective of this review is to investigate the use of patterns from dried drops as a characterisation and diagnosis technique. The patterns left behind by dried drops of various complex fluids are categorised. The potential applications of these patterns are presented, focussing primarily on healthcare, where the future impact could be greatest. A discussion on the limitations which must be overcome and prospective works that may be carried out to allow for widespread implementation of this technique is presented in conclusion.

Texture Analysis of Dried Droplets for the Quality Control of Medicines

The quality control of medicines guarantees the effectiveness of treatments for diseases. We explore the use of texture analysis of patterns in dried droplets as a tool to readily detect both impurities and changes in drug concentration. Four types of medicines associated with different routes of administration were analyzed: Methotrexate, Ciprofloxacin, Clonazepam, and Budesonide. We use NaCl and a hot substrate at 63 ∘C to promote aggregate formation and to reduce droplet drying time. Depending on the medicine, optical microscopy reveals different complex aggregates such as circular to oval splatters, fern-like islands, crown shapes, crown needle-like and bump-like patterns as well as dendritic branched and star-like crystals. We use some physical features of the stains (as the stain diameter and superficial area) and gray level co-occurrence matrix (GLCM) to characterize patterns of dried droplets. Finally, we show that structural analysis of stains can achieve 95% accuracy in identifying medicines with 30% water dilution, while it achieves 99% accuracy in detecting drugs with 10% other substances.

A Novel Fingerprint Biometric Cryptosystem Based on Convolutional Neural Networks

Modern access controls employ biometrics as a means of authentication to a great extent. For example, biometrics is used as an authentication mechanism implemented on commercial devices such as smartphones and laptops. This paper presents a fingerprint biometric cryptosystem based on the fuzzy commitment scheme and convolutional neural networks. One of its main contributions is a novel approach to automatic discretization of fingerprint texture descriptors, entirely based on a convolutional neural network, and designed to generate fixed-length templates. By converting templates into the binary domain, we developed the biometric cryptosystem that can be used in key-release systems or as a template protection mechanism in fingerprint matching biometric systems. The problem of biometric data variability is marginalized by applying the secure block-level Bose–Chaudhuri–Hocquenghem error correction codes, resistant to statistical-based attacks. The evaluation shows significant performance gains when compared to other texture-based fingerprint matching and biometric cryptosystems.