Optimized design technologiesOptimized design technologies
Yamatake has developed technologies for handling large-set data, but in the manufacturing field it is necessary to handle small-set experimental data, and in the pharmaceutical field it has proved particularly difficult to apply our technology.When our researchers were looking for ways in which to adapt our technologies so they could be used in pharmaceutical applications, they were approached by Prof. Kozo Takayama of the pharmacology department of Hoshi University, who was interested in the possibility of using small-set experimental data to generate models for drug design. Following research, we picked response surface methodology to develop a means for interpolating experimental data in a space configured from desired design factors and quantized characteristics. We also developed a method of searching for required design conditions on the response surface thus generated (Figure 1).
 Figure 1: Optimized design using response surface methodology *Click to zoom |
What follows is an explanation of this technology using drug manufacture as an example.
- Finding solutions to manufacturing recipe problems (example: basic drug formulation)
This methodology can be usefully applied to the formulation of drugs, as illustrated by the example of the bronchodilator Theophylline. It can also be used to solve similar problems in other manufacturing fields.
As demonstrated in Figure 2, Theophylline is manufactured in tablet form by mixing the non-medicinal ingredients controse and cornstarch in a ratio that determines the release (suppression vs. acceleration) of the principal agent, so that the desired release curve is achieved.
This release curve features a low initial disintegration rate (35.5 or less), fast disintegration in the primary phase (20 or more), and slow disintegration in the secondary phase (0.48 or less).
In other words, it was necessary to find the appropriate quantities of controse and cornstarch for achieving these release characteristics, and also the correct compression force to apply when pressing the tablets.
Figure 2: Determining Theophylline formulation
In tackling such problems, we would normally gather and examine experimental data (see Figure 3). However, using this new approach it is now possible to obtain the sort of surface seen in Figure 4, from which one can easily determine the optimum tablet formulation in order to meet the required specifications.
Figure 3: Experimental data for Theophylline formulation
Figure 4: Optimum Theophylline formulation
- Manufacturing processes (drug manufacture)
Following the completion of drug formulation, no problems may be encountered with manufacture under the sort of ideal conditions found in a laboratory, but with the transfer to mass production, yield is bound to fall below expectations because of the effect of disturbances that are unavoidable in real-world manufacturing facilities (Figure 5). To tackle this sort of problem, Yamatake's data mining can be used to analyze process disturbances and generate a response surface that includes those disturbances within the design factors of the basic recipe. This makes it possible to devise optimum manufacturing conditions that take into account the disturbances found in the mass production facilities.
Figure 5: Drug manufacture processes using mass production equipment
*Click to zoom
<<Measurement data processing
