Proprietary AI Algorithm Reduces Data Processing Time by 75 % and Supports Implementing Work Style Reform in Research Workplaces: Release of New Liquid Chromatograph Software “Peakintelligence for LC”

Shimadzu Corporation announces the release of Peakintelligence for LC, software incorporating a proprietary for a liquid chromatography(LC)with algorithm developed using AI technology. The algorithm was developed based on AI-learning using LC data from Ajinomoto Co., Inc. (“Ajinomoto Co.”). The algorithm can chromatography process data with the same skill level as experienced analysts automatically (about a 90 % match rate) in about 75 % less time. Shimadzu has been developing an algorithm using high-quality analysis data and decided to collaborate in response to Ajinomoto Co.’s request to establish a data processing flow that eliminates dependence on specific personnel. Peakintelligence for LC will be sold to users of Shimadzu LC for R&D and quality control applications in the pharmaceutical, food, and chemical fields.

LC is used in a wide range of fields, including pharmaceuticals, foods, and chemicals. Until now, peak*1finding algorithms have required the user to configure multiple parameter*2settings before they can start data processing. Also, when the algorithm can’t deal with a chromatogram, an experienced user has to select the peak by hand. Configuring and adjusting peak integration parameters is a time-consuming process.

Peakintelligence for LC is an algorithm incorporating AI assistance to search for chromatography peaks. It has learned peak processing skills from analytical data supplied by Ajinomoto Co. from healthcare, food&wellness and green solution fields making it compatible with data processing in a wide variety of fields. The algorithm can be implemented immediately without adjusting any parameters and efficient workflows can be established without relying on specific individuals. Shimadzu has been offering Peakintelligence series software since 2019. Ajinomoto Co. has been involved in amino acid research and has deployed a wide range of businesses based on that advantage since the company was founded over 100 years ago. The collaboration with Ajinomoto Co. will generate synergies that ensure Shimadzu will continue contributing to the well-being of mankind and the Earth in the future as well.

1. Algorithm Developed by AI-Learning from Ajinomoto Co. Analytical Data Is Applicable for Data Analysis in a Wide Variety of Fields
The new software was developed using analytical data generated from peak integration*3 work by Ajinomoto Co. researchers. By teaching the algorithm with extensive analytical data generated from healthcare, food & wellness, and green solution fields, which are considered strengths of Ajinomoto Co., the algorithm is now compatible with analytical data from a wide variety of other fields.

2. High-Quality Peak Processing Can Be Performed by Anyone
Because users do not need to review and configure parameter settings, even users with limited experience can acquire high-quality analytical results quickly. Consequently, the software can achieve efficient workflows with fewer human errors, less risk of data falsification, and that are not dependent on specific individuals .

3. Reduces Data Processing Time and Supports Reforming Laboratory Working Style
Automatic data processing by Peakintelligence for LC on par with an experienced operator (by about 90 %) and with no configuring of parameter settings significantly reduces the time spent on peak integration corrections. Peakintelligence for LC will increase analytical productivity by allowing customers to focus on other tasks.

*1: Peak-shaped areas in the chromatogram (graph of measurement results) obtained by an analytical instrument. Their height and area indicate the amount of a compound present.
*2: Conditions for detecting the starting point and ending point of peaks.
*3: Adjustment of baseline settings used as a reference for peak signals and for determining the position of peaks.
For more details, visit
Peakintelligence for LC

< | >