journal of biomedical informatics
All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

An Image Based Microtiter Plate Reader System for 96-well Format Fluorescence Assays

Author(s): Durai Arun P, Sankaran K, Muttan S

Background: 96-well microtiter plate assay are becoming popular analytical procedures in laboratory and clinical practices generating a demand for microtiter plate readers. The present colorimetric and fluorimetric based microtiter plate reader are efficient in general applications such as measuring reflectance, absorbance, and optical density of reaction mixtures. However, a dedicated system incorporating automation and computational techniques in specific microtiter plate readers for Drug discovery and Microbial viability, provides portability and compatibility with highthroughput analysis at affordable cost.

Objective: To develop image-based reader system for semi-quantitative measurements of transport assay performed in microtiter plate. In this work, we focus on deduction of illumination source and classification of fluorescence emitted by transport assay using this microtiter plate reader.

Method: We have taken the images of 96-well fluorescence efflux assay and developed an appropriate image analysis system to read the individual well images and categorize the results as fast efflux and slow efflux by the classifier. The images exhibited non-uniform illumination due to UV transilluminator, which then replaced by the circuit of blue LED after performing a study on its classification performance. Statistical descriptive features extracted from the image are classified using k-means clustering technique. The proposed system can be adapted to read any kind of microtiter plate assay.

Results: For uniform illumination of the source, of both UV transilluminator and blue LED, the color profile varied in accord to reaction mixture concentration. For classification performance of the source, out of four different images at various intensities of illumination from blue LED circuit controlled by potentiometer, the classification of efflux assay for image with maximum intensity gave high accuracy than the rest. For classification performance of the classifier, out of 1920-sub-images, the classifier misclassifies none.

Conclusion: For the microtiter plate reader, which classifies the liquid transport assay as fast efflux or slow efflux, the usage of UV transilluminator as illumination source is limited due to its non-uniform illumination and blue LED is apt for this application due its uniformity. The k-means clustering technique had been effective for this application when fed with statistical features describing the clusters.