The signature-based model for early detection of sepsis from electronic health records in the intensive care unit
Optimal feature selection leads to enhanced efficiency and accuracy when developing both supervised and unsupervised machine-learning models. In this work, a new signature-based regression model is proposed to automatically identify a patient's risk of sepsis based on physiological data streams...
| Main Authors: | , , , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
IEEE
2020
|
| _version_ | 1797063817685893120 |
|---|---|
| author | Morrill, J Kormilitzin, A Nevado-Holgado, A Swaminathan, S Howison, S Lyons, T |
| author_facet | Morrill, J Kormilitzin, A Nevado-Holgado, A Swaminathan, S Howison, S Lyons, T |
| author_sort | Morrill, J |
| collection | OXFORD |
| description | Optimal feature selection leads to enhanced efficiency and accuracy when developing both supervised and unsupervised machine-learning models. In this work, a new signature-based regression model is proposed to automatically identify a patient's risk of sepsis based on physiological data streams and to make a positive or negative prediction of sepsis for every time interval since admission to the intensive care unit. The gradient boosting machine algorithm that uses the features at the current time-points and the signature features extracted from the time-series to model the longitudinal effects of sepsis yields the utility function score of 0.360 (officially ranked 1st, team name: ‘Can I get your Signature?’) on the full test set. The signature method shows a systematic and competitive approach to model sepsis by learning from health data streams. |
| first_indexed | 2024-03-06T21:05:20Z |
| format | Conference item |
| id | oxford-uuid:3c3fb505-eaba-4310-8ea8-a377b7f25dce |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T21:05:20Z |
| publishDate | 2020 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:3c3fb505-eaba-4310-8ea8-a377b7f25dce2022-03-26T14:12:39ZThe signature-based model for early detection of sepsis from electronic health records in the intensive care unitConference itemhttp://purl.org/coar/resource_type/c_5794uuid:3c3fb505-eaba-4310-8ea8-a377b7f25dceEnglishSymplectic Elements at OxfordIEEE2020Morrill, JKormilitzin, ANevado-Holgado, ASwaminathan, SHowison, SLyons, TOptimal feature selection leads to enhanced efficiency and accuracy when developing both supervised and unsupervised machine-learning models. In this work, a new signature-based regression model is proposed to automatically identify a patient's risk of sepsis based on physiological data streams and to make a positive or negative prediction of sepsis for every time interval since admission to the intensive care unit. The gradient boosting machine algorithm that uses the features at the current time-points and the signature features extracted from the time-series to model the longitudinal effects of sepsis yields the utility function score of 0.360 (officially ranked 1st, team name: ‘Can I get your Signature?’) on the full test set. The signature method shows a systematic and competitive approach to model sepsis by learning from health data streams. |
| spellingShingle | Morrill, J Kormilitzin, A Nevado-Holgado, A Swaminathan, S Howison, S Lyons, T The signature-based model for early detection of sepsis from electronic health records in the intensive care unit |
| title | The signature-based model for early detection of sepsis from electronic health records in the intensive care unit |
| title_full | The signature-based model for early detection of sepsis from electronic health records in the intensive care unit |
| title_fullStr | The signature-based model for early detection of sepsis from electronic health records in the intensive care unit |
| title_full_unstemmed | The signature-based model for early detection of sepsis from electronic health records in the intensive care unit |
| title_short | The signature-based model for early detection of sepsis from electronic health records in the intensive care unit |
| title_sort | signature based model for early detection of sepsis from electronic health records in the intensive care unit |
| work_keys_str_mv | AT morrillj thesignaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT kormilitzina thesignaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT nevadoholgadoa thesignaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT swaminathans thesignaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT howisons thesignaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT lyonst thesignaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT morrillj signaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT kormilitzina signaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT nevadoholgadoa signaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT swaminathans signaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT howisons signaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit AT lyonst signaturebasedmodelforearlydetectionofsepsisfromelectronichealthrecordsintheintensivecareunit |