Agricultural products are prone to aflatoxin (AF)-producing moulds (Aspergillus flavus, A. parasiticus) during harvesting, drying, processing and also storage. AF is a mycotoxin that may cause liver cancer when consumed in amounts higher than allowed limits. Figs, like other agricultural products, are mostly affected by AF-producing moulds and these moulds usually produce kojic acid together with AF. Kojic acid is a fluorescent compound and exhibiting bright greenish yellow fluorescence (BGYF) under ultraviolet (UV) light. Using this fluorescence property, fig-processing plants manually select and remove the BGYF+ figs to reduce the AF level of the processed figs. Although manual selection is based on subjective criteria and strongly depends on the expertise level of the workers, it is known as the most effective way of removing AF-contaminated samples. However, during manual selection, workers are exposed to UV radiation and this brings skin health problems. In this study, we individually investigated the figs to measure their fluorescence level, surface mould concentration and AF levels and noted a strong correlation between mould concentration and BGYF and AF, and BGYF and surface. In addition to a pairwise correlation, we proposed a machine-vision and machine-learning approach to detect the AF-contaminated figs using their multispectral images under UV light. The figs were classified in two different approaches considering their surface mould and AF level with error rates of 9.38% and 11.98%, respectively.

Early intervention can delay the progress of Alzheimer’s Disease (AD), but currently, there are no effective prediction tools. The goal of this study is to generate a reliable artificial intelligence (AI) model capable of detecting the high risk of AD, based on gene expression arrays from blood samples. To that end, a novel image-formation method is proposed to transform single-dimension gene expressions into a discriminative 2-dimensional (2D) image to use convolutional neural networks (CNNs) for classification. Three publicly available datasets were pooled, and a total of 11,618 common genes’ expression values were obtained. The genes were then categorized for their discriminating power using the Fisher distance (AD vs. control (CTL)) and mapped to a 2D image by linear discriminant analysis (LDA). Then, a six-layer CNN model with 292,493 parameters were used for classification. An accuracy of 0.842 and an area under curve (AUC) of 0.875 were achieved for the AD vs. CTL classification. The proposed method obtained higher accuracy and AUC compared with other reported methods. The conversion to 2D in CNN offers a unique advantage for improving accuracy and can be easily transferred to the clinic to drastically improve AD (or any disease) early detection.