Yaroslav Balytskyi*, Nataliia Kalashnyk, Inna Hubenko, Alina Balytska and Kelly McNear,
{"title":"通过特征正则化增强开放世界细菌拉曼光谱识别能力,提高对未知类别的复原力","authors":"Yaroslav Balytskyi*, Nataliia Kalashnyk, Inna Hubenko, Alina Balytska and Kelly McNear, ","doi":"10.1021/cbmi.4c00007","DOIUrl":null,"url":null,"abstract":"<p >The combination of deep learning techniques and Raman spectroscopy shows great potential offering precise and prompt identification of pathogenic bacteria in clinical settings. However, the traditional closed-set classification approaches assume that all test samples belong to one of the known pathogens, and their applicability is limited since the clinical environment is inherently unpredictable and dynamic, unknown, or emerging pathogens may not be included in the available catalogs. We demonstrate that the current state-of-the-art neural networks identifying pathogens through Raman spectra are vulnerable to unknown inputs, resulting in an uncontrollable false positive rate. To address this issue, first we developed an ensemble of ResNet architectures combined with the attention mechanism that achieves a 30-isolate accuracy of 87.8 ± 0.1%. Second, through the integration of feature regularization by the Objectosphere loss function, our model both achieves high accuracy in identifying known pathogens from the catalog and effectively separates unknown samples drastically reducing the false positive rate. Finally, the proposed feature regularization method during training significantly enhances the performance of out-of-distribution detectors during the inference phase improving the reliability of the detection of unknown classes. Our algorithm for Raman spectroscopy empowers the identification of previously unknown, uncataloged, and emerging pathogens ensuring adaptability to future pathogens that may surface. Moreover, it can be extended to enhance open-set medical image classification, bolstering its reliability in dynamic operational settings.</p>","PeriodicalId":53181,"journal":{"name":"Chemical & Biomedical Imaging","volume":"2 6","pages":"442–452"},"PeriodicalIF":0.0000,"publicationDate":"2024-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://pubs.acs.org/doi/epdf/10.1021/cbmi.4c00007","citationCount":"0","resultStr":"{\"title\":\"Enhancing Open-World Bacterial Raman Spectra Identification by Feature Regularization for Improved Resilience against Unknown Classes\",\"authors\":\"Yaroslav Balytskyi*, Nataliia Kalashnyk, Inna Hubenko, Alina Balytska and Kelly McNear, \",\"doi\":\"10.1021/cbmi.4c00007\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p >The combination of deep learning techniques and Raman spectroscopy shows great potential offering precise and prompt identification of pathogenic bacteria in clinical settings. However, the traditional closed-set classification approaches assume that all test samples belong to one of the known pathogens, and their applicability is limited since the clinical environment is inherently unpredictable and dynamic, unknown, or emerging pathogens may not be included in the available catalogs. We demonstrate that the current state-of-the-art neural networks identifying pathogens through Raman spectra are vulnerable to unknown inputs, resulting in an uncontrollable false positive rate. To address this issue, first we developed an ensemble of ResNet architectures combined with the attention mechanism that achieves a 30-isolate accuracy of 87.8 ± 0.1%. Second, through the integration of feature regularization by the Objectosphere loss function, our model both achieves high accuracy in identifying known pathogens from the catalog and effectively separates unknown samples drastically reducing the false positive rate. Finally, the proposed feature regularization method during training significantly enhances the performance of out-of-distribution detectors during the inference phase improving the reliability of the detection of unknown classes. Our algorithm for Raman spectroscopy empowers the identification of previously unknown, uncataloged, and emerging pathogens ensuring adaptability to future pathogens that may surface. Moreover, it can be extended to enhance open-set medical image classification, bolstering its reliability in dynamic operational settings.</p>\",\"PeriodicalId\":53181,\"journal\":{\"name\":\"Chemical & Biomedical Imaging\",\"volume\":\"2 6\",\"pages\":\"442–452\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-05-06\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://pubs.acs.org/doi/epdf/10.1021/cbmi.4c00007\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Chemical & Biomedical Imaging\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://pubs.acs.org/doi/10.1021/cbmi.4c00007\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Chemical & Biomedical Imaging","FirstCategoryId":"1085","ListUrlMain":"https://pubs.acs.org/doi/10.1021/cbmi.4c00007","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Enhancing Open-World Bacterial Raman Spectra Identification by Feature Regularization for Improved Resilience against Unknown Classes
The combination of deep learning techniques and Raman spectroscopy shows great potential offering precise and prompt identification of pathogenic bacteria in clinical settings. However, the traditional closed-set classification approaches assume that all test samples belong to one of the known pathogens, and their applicability is limited since the clinical environment is inherently unpredictable and dynamic, unknown, or emerging pathogens may not be included in the available catalogs. We demonstrate that the current state-of-the-art neural networks identifying pathogens through Raman spectra are vulnerable to unknown inputs, resulting in an uncontrollable false positive rate. To address this issue, first we developed an ensemble of ResNet architectures combined with the attention mechanism that achieves a 30-isolate accuracy of 87.8 ± 0.1%. Second, through the integration of feature regularization by the Objectosphere loss function, our model both achieves high accuracy in identifying known pathogens from the catalog and effectively separates unknown samples drastically reducing the false positive rate. Finally, the proposed feature regularization method during training significantly enhances the performance of out-of-distribution detectors during the inference phase improving the reliability of the detection of unknown classes. Our algorithm for Raman spectroscopy empowers the identification of previously unknown, uncataloged, and emerging pathogens ensuring adaptability to future pathogens that may surface. Moreover, it can be extended to enhance open-set medical image classification, bolstering its reliability in dynamic operational settings.
期刊介绍:
Chemical & Biomedical Imaging is a peer-reviewed open access journal devoted to the publication of cutting-edge research papers on all aspects of chemical and biomedical imaging. This interdisciplinary field sits at the intersection of chemistry physics biology materials engineering and medicine. The journal aims to bring together researchers from across these disciplines to address cutting-edge challenges of fundamental research and applications.Topics of particular interest include but are not limited to:Imaging of processes and reactionsImaging of nanoscale microscale and mesoscale materialsImaging of biological interactions and interfacesSingle-molecule and cellular imagingWhole-organ and whole-body imagingMolecular imaging probes and contrast agentsBioluminescence chemiluminescence and electrochemiluminescence imagingNanophotonics and imagingChemical tools for new imaging modalitiesChemical and imaging techniques in diagnosis and therapyImaging-guided drug deliveryAI and machine learning assisted imaging