Corinna Wegner, Zita I. Zarandy, Nico Feiler, Lea Gigou, Timo Halenke, Niklas Leopold-Kerschbaumer, Maik Krusche, Weronika Skibicka, Kosmas V. Kepesidis
{"title":"基于血液的红外光谱的无监督深度学习信息表示。","authors":"Corinna Wegner, Zita I. Zarandy, Nico Feiler, Lea Gigou, Timo Halenke, Niklas Leopold-Kerschbaumer, Maik Krusche, Weronika Skibicka, Kosmas V. Kepesidis","doi":"10.1002/jbio.70011","DOIUrl":null,"url":null,"abstract":"<p>This study explores using unsupervised deep learning to find a low-dimensional representation of infrared molecular fingerprints of human blood. We developed a fully convolutional denoising autoencoder to process Fourier transform infrared (FTIR) spectroscopy data, aiming to condense the spectra into a set of latent variables. By utilizing the autoencoder's bottleneck architecture and a custom loss function, we effectively reduced noise while retaining essential molecular information. This method improved lung cancer detection accuracy by 2.6 percentage points in a case–control study. The resulting latent space not only compacts spectral data, but also highlights variables linked to disease presence, offering potential for improving diagnostics.</p><p>\n <b>Trial Registration:</b> German Clinical Trials Register (DRKS): DRKS00013217</p>","PeriodicalId":184,"journal":{"name":"Journal of Biophotonics","volume":"18 8","pages":""},"PeriodicalIF":2.0000,"publicationDate":"2025-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/jbio.70011","citationCount":"0","resultStr":"{\"title\":\"Toward Informative Representations of Blood-Based Infrared Spectra via Unsupervised Deep Learning\",\"authors\":\"Corinna Wegner, Zita I. Zarandy, Nico Feiler, Lea Gigou, Timo Halenke, Niklas Leopold-Kerschbaumer, Maik Krusche, Weronika Skibicka, Kosmas V. Kepesidis\",\"doi\":\"10.1002/jbio.70011\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>This study explores using unsupervised deep learning to find a low-dimensional representation of infrared molecular fingerprints of human blood. We developed a fully convolutional denoising autoencoder to process Fourier transform infrared (FTIR) spectroscopy data, aiming to condense the spectra into a set of latent variables. By utilizing the autoencoder's bottleneck architecture and a custom loss function, we effectively reduced noise while retaining essential molecular information. This method improved lung cancer detection accuracy by 2.6 percentage points in a case–control study. The resulting latent space not only compacts spectral data, but also highlights variables linked to disease presence, offering potential for improving diagnostics.</p><p>\\n <b>Trial Registration:</b> German Clinical Trials Register (DRKS): DRKS00013217</p>\",\"PeriodicalId\":184,\"journal\":{\"name\":\"Journal of Biophotonics\",\"volume\":\"18 8\",\"pages\":\"\"},\"PeriodicalIF\":2.0000,\"publicationDate\":\"2025-03-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://onlinelibrary.wiley.com/doi/epdf/10.1002/jbio.70011\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Biophotonics\",\"FirstCategoryId\":\"101\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/jbio.70011\",\"RegionNum\":3,\"RegionCategory\":\"物理与天体物理\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"BIOCHEMICAL RESEARCH METHODS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Biophotonics","FirstCategoryId":"101","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/jbio.70011","RegionNum":3,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"BIOCHEMICAL RESEARCH METHODS","Score":null,"Total":0}
Toward Informative Representations of Blood-Based Infrared Spectra via Unsupervised Deep Learning
This study explores using unsupervised deep learning to find a low-dimensional representation of infrared molecular fingerprints of human blood. We developed a fully convolutional denoising autoencoder to process Fourier transform infrared (FTIR) spectroscopy data, aiming to condense the spectra into a set of latent variables. By utilizing the autoencoder's bottleneck architecture and a custom loss function, we effectively reduced noise while retaining essential molecular information. This method improved lung cancer detection accuracy by 2.6 percentage points in a case–control study. The resulting latent space not only compacts spectral data, but also highlights variables linked to disease presence, offering potential for improving diagnostics.
Trial Registration: German Clinical Trials Register (DRKS): DRKS00013217
期刊介绍：
The first international journal dedicated to publishing reviews and original articles from this exciting field, the Journal of Biophotonics covers the broad range of research on interactions between light and biological material. The journal offers a platform where the physicist communicates with the biologist and where the clinical practitioner learns about the latest tools for the diagnosis of diseases. As such, the journal is highly interdisciplinary, publishing cutting edge research in the fields of life sciences, medicine, physics, chemistry, and engineering. The coverage extends from fundamental research to specific developments, while also including the latest applications.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
