Film Coating End-Point Determination for Colored Immediate Release Tablets Using Multivariate Image Analysis | AIChE

Film Coating End-Point Determination for Colored Immediate Release Tablets Using Multivariate Image Analysis


Conference Presentation

Conference Type

AIChE Annual Meeting

Presentation Date

November 10, 2009

Skill Level




Quality by design is an exercise that relies on quantitative metrics of the quality of a product. For immediate release (IR) film coated products, the uniformity of the coat is traditionally assessed by visual inspection. It is desired to substitute such assessment with a quantitative one.

In this work, Multivariate image analysis (MIA) was used to assess film coated product. The objective is to determine the end-point for cosmetic coating of the tablets, and calculate the distribution of coating across individual tablets as the process progresses; using digital RGB images as inputs. Knowing the distribution of the coating will allow the user to establish reasonable limits of coating material to be applied accounting for the mixing and spray inefficiencies in the coater.

This method has been successfully applied to snack food , and relies on the correlation of appearance with the property of interest. In the case of coated tablets there is a correlation between the weight increase in the tablets and their color. This correlation is broken when the coating is uniform, and the application of more coating material on the tablet does not result in a change in color (this point is referred to as the cosmetic end-point).

MIA is performed on digital images taken off-line, of a representative number of tablets sampled at different time points during the coating run; the percentage increase in weight is determined for each group of tablets. The images and the estimated weight increases are used in the MIA method to calibrate a predictive model to estimate the average coating weight for a given set of pixels in the image.

The average weight increase is calculated by applying this prediction to an image of multiple tablets. The coating distributions are calculated by subdividing the images into area units approximately equal to that of a single tablet. The prediction is done for each sub-image, and the set of predictions is used to produce a distribution.

Two lots of 100mg tablets from two different scales were studied to assess the off-line application of the method. As expected, the coating distributions on tablets produced at smaller scale exhibit narrower distributions that those at large scale. The estimated average amount of coating required for cosmetic purposes is the same in both cases, however differences in mixing (which result in different distributions) dictate a different cosmetic-end point for each one.

An initial assessment to carry out this analysis in-situ was also made using a digital video camera inside a laboratory scale coater. An adaptive PCA model is proposed as a method to monitor the changes in the appearance of the tablets. Experimental results are shown to illustrate the use of the method in successfully determining coating-end point for a 175mg round tablet.

[1] H. Yu, J. F. MacGregor, G. Haarsma, and W. Bourg, &'&'Digital Imaging for Online Monitoring and Control of Industrial Snack Food Processes,&'&' Ind. Eng. Chem. Res., vol. 42, pp. 3036-3044, 2003.&'


Once the content has been viewed and you have attested to it, you will be able to download and print a certificate for PDH credits. If you have already viewed this content, please click here to login.



Do you already own this?



AIChE Member Credits 0.5
AIChE Pro Members $15.00
AIChE Graduate Student Members Free
AIChE Undergraduate Student Members Free
AIChE Explorer Members $25.00
Non-Members $25.00