New Diagnostic Modality Combining Mass Spectrometry and Machine Learning for the Discrimination of Malignant Intraductal Papillary Mucinous Neoplasms

Sep 24, 2026

Background. Intraductal papillary mucinous neoplasm (IPMN) has malignant potential, yet effective diagnosis of malignant conversion is not established. The study evaluates a combined strategy of probe electrospray ionization-mass spectrometry (PESI-MS) and machine learning. Methods. Serum from 42 IPMN patients was analyzed by PESI-MS, converting mass spectra into continuous variables, then used partial least square regression or support vector machine to classify low-grade dysplasia versus advanced IPMN. Results. Both methods distinguished disease classes, yielding 88.1% diagnostic accuracy and an AUC of 0.924; each method required about 10 minutes. Conclusion. PESI-MS with machine learning enables rapid, on-site discrimination.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages8

Reading Time20 min

AccessFree to download

Abstract

Background

Methods

Results

Conclusion

Materials and Methods

Ann Surg Oncol (2023) 30:3150–3157 [https://doi.org/10.1245/s10434-022-13012-y](https://doi.org/10.1245/s10434-022-13012-y) ORIGINAL ARTICLE – TRANSLATIONAL RESEARCH New Diagnostic Modality Combining Mass Spectrometry and Machine Learning for the Discrimination of Malignant Intraductal Papillary Mucinous Neoplasms Sho Kiritani, MD, PhD1, Tomohiko Iwano, PhD2, Kentaro Yoshimura, PhD3, Ryo Saito, MD, PhD4, Takashi Nakayama, MD, PhD4, Daisuke Yamamoto, MD1, Hiroyuki Hakoda, MD1, Genki Watanabe, MD, PhD1, Nobuhisa Akamatsu, MD, PhD1, Junichi Arita, MD, PhD1, Junichi Kaneko, MD, PhD1, S´en Takeda, MD, PhD5,2, Daisuke Ichikawa, MD, PhD4, and Kiyoshi Hasegawa, MD, PhD1 1Hepato-Biliary-Pancreatic Surgery Division, Department of Surgery, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; 2Department of Anatomy and Cell Biology, Faculty of Medicine, The University of Yamanashi, Yamanashi, Japan; 3Division of Molecular Biology, Center for Medical Education and Sciences, Interdisciplinary Graduate School of Medicine, University of Yamanashi, Yamanashi, Japan; 4Department of Digestive Surgery and Breast and Endocrine Surgery, The University of Yamanashi, Yamanashi, Japan; 5Department of Anatomy, Teikyo University School of Medicine, Tokyo, Japan ABSTRACT Background. An intraductal papillary mucinous neoplasm (IPMN) is a pancreatic tumor with malignant potential. Although we anticipate a sensitive method to diagnose the malignant conversion of IPMN, an effective strategy has not yet been established. The combination of probe electrospray ionization-mass spectrometry (PESI-MS) and machine learning provides a promising solution for this purpose. Methods. We prospectively analyzed 42 serum samples obtained from IPMN patients who underwent pancreatic resection between 2020 and 2021 . Based on the postoperative pathological diagnosis, patients were classiﬁed into two groups: IPMN-low grade dysplasia (n = 17) and advanced-IPMN (n = 25) . Serum samples were analyzed by PESI-MS, and the obtained mass spectral data were converted into continuous variables. These variables were used to discriminate advanced-IPMN from IPMN-low grade dysplasia by partial least square regression or support 􀀂 The Author(s) 2022 First Received: 13 October 2022 Accepted: 9 December 2022 Published Online: 8 January 2023 K. Hasegawa, MD, PhD e-mail: [kihase-tky@umin.ac.jp](kihase-tky@umin.ac.jp); [Hasegawa-2su@h.u-tokyo.ac.jp](Hasegawa-2su@h.u-tokyo.ac.jp) vector machine analysis. The areas under receiver operating characteristics curves were obtained to visualize the difference between the two groups. Results. Partial least square regression successfully discriminated the two disease classes. From another standpoint, we selected 130 parameters from the entire dataset by PESI-MS, which were fed into the support vector machine. The diagnostic accuracy was 88.1%, and the area under the receiver operating characteristics curve was 0.924 by this method. Approximately 10 min were required to perform each method. Conclusion. PESI-MS combined with machine learning isan easy-to-use tool with the advantage of rapid on-site analysis. Here, we show the great potential of our system to diagnose the malignant conversion of IPMN, which would be a promising diagnostic tool in clinical settings. An intraductal papillary mucinous neoplasm (IPMN) of the pancreas is characterized by the production of mucin with a dilated pancreatic duct.1-3 This category of pancreatic tumor was ﬁrst conceptualized in the 2nd section of the World Health Organization classiﬁcation in 1996, which made clinicians aware of IPMN as a differential diagnosis. Since then, the number of IPMN cases has increased because of the enhanced accuracy of imaging technology and recognition of the disease.4,5 An intraductal papillary mucinous neoplasm (IPMN) of the pancreas is characterized by the production of mucin with a dilated pancreatic duct.1-3 This category of panc

FAQs

What clinical problem does the study address for IPMN?

The study targets the need for an effective approach to diagnose malignant conversion of intraductal papillary mucinous neoplasms.

How were serum samples analyzed in the study?

Serum from 42 IPMN patients was measured using probe electrospray ionization-mass spectrometry (PESI-MS), producing mass spectral data converted into continuous variables.

What performance did the machine-learning diagnostic approach achieve?

Using a support vector machine with 130 parameters selected from the dataset yielded 88.1% diagnostic accuracy and an AUC of 0.924; each method took about 10 minutes.

Sep 24, 2026

Uploaded by Emma Wilson