[ad_1]
When medical companies produce pills and tablets that treat any disease, ache and pain, they need to isolate the active pharmaceutical ingredient from the suspension and dry it. The process requires a human operator to monitor the industrial dryer, agitate the material, and get the compound to the correct compression properties in the medicine. The work is highly dependent on the operator’s observations.
Methods to make this process less subjective and much more efficient are recent subjects Communications of nature The paper, authored by researchers at MIT and Takeda. The authors of the paper developed a way to use physics and machine learning to categorize the rough surfaces that characterize the particles in the mixture. A technique that uses a physics-enhanced autocorrelation-based estimator (PEACE) could transform pharmaceutical manufacturing processes for pills and powders, increasing efficiency and accuracy and resulting in fewer failed batches of pharmaceutical products.
“Failure batches or failed steps in the pharmaceutical process are very serious,” says Alan Myerson, a professor of practice in MIT’s Department of Chemical Engineering and one of the study’s authors. “Anything that improves the reliability of pharmaceutical manufacturing, reduces time and improves compliance is a big deal.”
The team’s work is part of an ongoing collaboration between Takeda and MIT. Launched in 2020. The MIT-Takeda program aims to leverage the expertise of both MIT and Takeda to solve problems at the intersection of medicine, artificial intelligence and healthcare.
In pharmaceutical manufacturing, determining whether a compound is adequately mixed and dried usually requires stopping an industrial-sized dryer and taking samples from the production line for testing. Takeda researchers thought that artificial intelligence could improve the task and reduce the downtime that slows down production. Initially, the research team planned to use the videos to train a computer model to replace a human operator. But deciding which videos to use for model training still turned out to be very subjective. Instead, the MIT-Takeda team decided to shine a laser on the particles during filtration and drying and measure the particle size distribution using physics and machine learning.
“We just shine a laser beam on top of this drying surface and observe,” says Qihang Zhang, a doctoral student in MIT’s Department of Electrical Engineering and Computer Science and first author of the study.
An equation derived from physics describes the interaction between the laser and the mixture, while machine learning characterizes the particle sizes. The process requires no stopping and starting, which means the entire job is safer and more efficient than standard operating procedure, says George Barbastatis, MIT professor of mechanical engineering and corresponding author of the study.
A machine learning algorithm also doesn’t need a lot of data sets to learn its work because physics allows for fast neural network training.
“We use physics to compensate for the lack of training data to train the neural network in an efficient way,” says Zhang. “Only a small amount of experimental data is enough to get a good result.”
Today, the only in-house processes used for particle measurement in the pharmaceutical industry are those used for liquid products, where the crystals float in a liquid. There is no method to measure the particles during mixing inside the powder. Powders can be made from rotting, but when the liquid is filtered and dried, its composition changes, requiring new measurements. In addition to making the process faster and more efficient, using the PEACE mechanism makes the job safer because it requires less handling of potentially powerful materials, the authors say.
The implications for pharmaceutical manufacturing can be significant, allowing drug production to be more efficient, sustainable and cost-effective by reducing the number of experiments that companies must conduct when making products. Monitoring the properties of the drying mixture is an issue the industry has long grappled with, says Charles Papageorgiou, director of Takeda’s Process Chemistry Development group and one of the study’s authors.
“It’s a problem that a lot of people are trying to solve, and it’s not a good sensor,” says Papageorgiou. “I think this is a pretty big step change in terms of being able to do particle size distributions in real time.”
Papageorgiou said the mechanism could have applications in other industrial pharmaceutical operations. At some point, laser technology can produce a video image, allowing manufacturers to use a camera for analysis rather than laser measurements. The company is now working to evaluate the tool in its lab on different compounds.
The results come directly from a collaboration between Takeda and three MIT departments: mechanical engineering, chemical engineering, and electrical engineering and computer science. Over the past three years, MIT and Takeda researchers have worked together on 19 projects focusing on the application of machine learning and artificial intelligence to problems in the healthcare and medical industries as part of the MIT-Takeda program.
Often, academic research can take years to translate into industrial processes. But the researchers hope that direct collaboration can shorten those timelines. Takeda is walking distance from the MIT campus, which allowed researchers to set up tests in the company’s lab, and real-time feedback from Takeda helped MIT researchers structure research around the company’s equipment and operations.
Combining the expertise and mission of both organizations helps researchers ensure that their experimental results have real-world implications. The team has already filed two patents and is planning a third.
[ad_2]
Source link