English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Falsification of medicines is a big problem in many developing countries,where technological infrastructure is inadequate to detect these harmfulproducts. We have developed a set of inexpensive paper cards, called PaperAnalytical Devices (PADs), which can efficiently classify drugs based on theirchemical composition, as a potential solution to the problem. These cards havedifferent reagents embedded in them which produce a set of distinctive colordescriptors upon reacting with the chemical compounds that constitutepharmaceutical dosage forms. If a falsified version of the medicine lacks theactive ingredient or includes substitute fillers, the difference in color isperceivable by humans. However, reading the cards with accuracy takes trainingand practice, which may hamper their scaling and implementation in low resourcesettings. To deal with this, we have developed an automatic visual recognitionsystem to read the results from the PAD images. At first, the optimal set ofreagents was found by running singular value decomposition on the intensityvalues of the color tones in the card images. A dataset of cards embedded withthese reagents is produced to generate the most distinctive results for a setof 26 different active pharmaceutical ingredients (APIs) and excipients. Then,we train two popular convolutional neural network (CNN) models, with the cardimages. We also extract some "hand-crafted" features from the images and traina nearest neighbor classifier and a non-linear support vector machine withthem. On testing, higher-level features performed much better in accuratelyclassifying the PAD images, with the CNN models reaching the highest averageaccuracy of over 94\%.

Visual Recognition of Paper Analytical Device Images for Detection of Falsified Pharmaceuticals