Content overflows the browser. Try zooming out (Ctrl + −) to see the full page. PDF Download
User Benefits Application News Peakintelligence peak integration software substantially reduces the time and effort required for data analysis in PFAS analysis. Since it does not require parameter settings, Peakintelligence allows even inexperienced users to perform peak integration and achieve results on par with experienced workers. Efficient Data Analysis with Peakintelligence™ —Application for Analyzing Per- and Polyfluoroalkyl Substances (PFAS)— Takanari Hattori and Hiroyuki Niwa High Performance Liquid Chromatograph Mass Spectrometer LCMS™-8045/8050/8060RX Introduction Per- and polyfluoroalkyl substances (PFAS) are lyophobic, heat resistant, and chemically resistant compounds used extensively as coatings, surface treatment agents, emulsifiers, and fire- extinguishing agents. Concerns about PFAS and their persistence in the environment, accumulation in living organisms, toxicity to living organisms, and ability to travel long distances have recently led to fact-finding studies on PFAS and the introduction of regulations. As of October 2023, only three PFAS are subject to regulatory restrictions in Japan (PFOS, PFOA, and PFHxS), but more are subject to restrictions in the USA and Europe. Throughout the world, an increasing number of PFAS are expected to require analysis. These changes will create more analytical data that will have to be processed, creating a need for more efficient methods of data analysis. Peakintelligence is a peak integration software equipped with peak-finding algorithms that have been developed with the help of artificial intelligence (AI). This Application News describes a case example in which it was used to process chromatogram data of PFAS analysis and reduce the effort required and increase the efficiency of the analysis. • Peakintelligence is a product developed through joint research between Shimadzu and Fujitsu. Limitations of Previous Algorithms Previous peak integration algorithms required the user to adjust and fine-tune numerous detection and integration parameters based on the chromatogram (Fig. 1). When the algorithms failed to detect and integrate peaks correctly, manual corrections were needed, which increased the workload for users. The skill level required to carry out this work and the challenge of standardizing this process also posed major issues when applying these algorithms. Fig. 1 Parameter Setting Window for Previous Peak Integration Algorithm Peakintelligence Peak Integration Algorithms Peakintelligence is a new peak integration technology that has been developed using deep learning, a type of AI. AI is the name used for the general concept and technology, while machine learning and deep learning are the techniques on which AI is built. Machine learning trains models with target attributes that must first be extracted by a person, but deep learning trains models with target attributes that are extracted mechanically by software on a computer. For this reason, deep learning can train models with large volumes of data and without the variability caused by person involvement. The trained model used by Peakintelligence was created from approximately 13,000 chromatograms and with peak detection and integration verified by experienced specialists. A training dataset was prepared by labeling the chromatogram data and peak start and end points, tuning the training and hyperparameters, and assessing model performance (Fig. 2). Installing this trained model on a laboratory PC gives the user the ability to analyze LC-MS data automatically with minimal effort. • Peakintelligence does not have the functionality to perform additional learning using newly acquired data. Machine (PC) Training Trained model Peakintelligence Training Output Analysis Trained PC (Peakintelligence installed) Fig. 2 Illustrative Overview of Peakintelligence Input Estimation
Application News www.shimadzu.com/an/ Shimadzu Corporation © Shimadzu Corporation, 2024 For Research Use Only. Not for use in diagnostic procedures. This publication may contain references to products that are not available in your country. Please contact us to check the availability of these products in your country. The content of this publication shall not be reproduced, altered or sold for any commercial purpose without the written approval of Shimadzu. See http://www.shimadzu.com/about/trademarks/index.html for details. Third party trademarks and trade names may be used in this publication to refer to either the entities or their products/services, whether or not they are used with trademark symbol “TM” or “”. Shimadzu disclaims any proprietary interest in trademarks and trade names other than its own. The information contained herein is provided to you "as is" without warranty of any kind including without limitation warranties as to its accuracy or completeness. Shimadzu does not assume any responsibility or liability for any damage, whether direct or indirect, relating to the use of this publication. This publication is based upon the information available to Shimadzu on or before the date of publication, and subject to change without notice. LCMS and Peakintelligence are trademarks of Shimadzu Corporation and its affiliates in Japan and other countries. No parameter configuration Excellent Peak Integration with No Parameter Adjustment Because Peakintelligence is machine-trained with chromatograms processed by experienced specialists, it can analyze peaks to the same standard as experienced analysts. Unlike previous algorithms, It also requires no prior adjustment of parameters (Fig. 3). Users do not need to set multiple complex parameters. They can process chromatograms with Peakintelligence by simply specifying an algorithm and model. This ability also reduces variability arising from misconfigured parameters and other forms of human error. Fig. 3 Parameter Setting Window for Previous Algorithm and Peakintelligence Previous method Peakintelligence Reduced Data Analysis Times Industrial wastewater was analyzed for 72 PFAS (31 internal standard substances) and Peakintelligence was used to process the resulting chromatograms. As shown in Fig. 4, it successfully detected and integrated peaks with a poor S/N. As shown in Fig. 5, the previous method falsely detected baseline noise and undulations, while Peakintelligence detected peaks accurately with no false detections. Processing peaks by the previous method resulted in false detections in 26 of the 72 components, while Peakintelligence showed no false detections for any component (Fig. 6). When a peak is falsely detected, the peak must be removed manually, which increases the analysis time. Peakintelligence required only 45 seconds to analyze a single piece of data compared to 2 minutes for the previous method. Scaled up to 100 pieces of data, Peakintelligence reduces the analysis time from 3.3 hours to around just 1.3 hours, offering substantial savings in labor and time. 1.46e3 Q 312.95>269.00 (-) 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 0.0e0 2.0e2 4.0e2 6.0e2 8.0e2 1.0e3 1.2e3 1.4e3 Even with a poor S/N, peak integration with ease PFHxA Fig. 4 Example Peak Integration by Peakintelligence Fig. 6 Comparing Peak Integration Results for 72 PFAS in Industrial Wastewater Conclusion Using Peakintelligence to process peaks in PFAS analysis substantially reduced the number of false detections and falsely identified peaks. It also significantly reduced the number of peaks that required manual processing, and it did not need pre- analysis adjustment of parameters. These improvements enable substantial reductions in the time and effort required to perform data analysis. Avoiding parameter settings by the user also eliminates variability in the results between users, which helps eliminate dependency on specific personnel. As the number of PFAS requiring analysis increases throughout the world, a more efficient method of analyzing peaks will be needed; Peakintelligence is an effective solution for this need. Previous method Peakintelligence 46 26 72 0 No problems False detections Fig. 5 Comparison of Peak Integration by Previous Method and Peakintelligence 6.80e1 Q 498.95>79.95 (-) 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 0.0e0 1.0e1 2.0e1 3.0e1 4.0e1 5.0e1 6.0e1 2.10e2 Q 314.95>135.00 (-) 4.5 5.0 5.5 6.0 6.5 7.0 7.5 0.0e0 2.0e1 4.0e1 6.0e1 8.0e1 1.0e2 1.2e2 1.4e2 1.6e2 1.8e2 2.0e2 No falsely detected peaks 6.80e1 Q 498.95>79.95 (-) 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 0.0e0 1.0e1 2.0e1 3.0e1 4.0e1 5.0e1 6.0e1 2.10e2 Q 314.95>135.00 (-) 4.5 5.0 5.5 6.0 6.5 7.0 7.5 0.0e0 2.0e1 4.0e1 6.0e1 8.0e1 1.0e2 1.2e2 1.4e2 1.6e2 1.8e2 2.0e2 False detection of noise as peaks PFOS PFEESA Previous method Peakintelligence First Edition: Sep. 2024 01-00658-EN No parameter setting
Related Products Some products may be updated to newer models. LCMS-8060NX Liquid Chromatograph Mass Spectrometer LCMS-TQ RX Series Triple Quadrupole LC-MS/MS Related Solutions Environment Water Waste-Others PFAS Price Inquiry Product Inquiry Technical Service / Support Inquiry Other Inquiry Powered by TCPDF (www.tcpdf.org)