Aiding infection analysis and diagnosis through temporally-contextualized matrix representations

Maksim Gomov, Jia Kai Chou, Jianping Kelvin Li, Soman Sen, Kiho Cho, Nam Tran, Kwan-Liu Ma

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Determining infections and sepsis of severely burned adults in a timely fashion allows clinicians to provide necessary treatments to critically ill patients, potentially reducing the chance of mortality. In current practice, clinicians examine large amounts of heterogeneous medical records using a spreadsheet-like representation and perform analysis of descriptive statistics to aid in their decision making for sepsis diagnosis. A more efficient approach is required to streamline such a process; accordingly, we developed an interactive visual interface for supporting quick inspection and comparison of patients' retrospective clinical trajectories. In particular, we employ a timeline representation to present entire treatment contexts for individual patients, and an aggregated matrix representation for summarizing multiple data variables of individual patients over time. This provides clinicians with a compact and intuitive way to discover the important trends, patterns, and events that occur in the context of multiple patients. We present several possible use cases identified by clinicians using our system and show that our preliminary results have the potential to greatly improve diagnostic timing and accuracy of sepsis identification in critically ill patients.

Original languageEnglish (US)
Title of host publication2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages31-38
Number of pages8
ISBN (Electronic)9781538631874
DOIs
StatePublished - Jun 15 2018
Event8th Annual IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017 - Phoenix, United States
Duration: Oct 1 2017 → …

Other

Other8th Annual IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017
CountryUnited States
CityPhoenix
Period10/1/17 → …

Fingerprint

Spreadsheets
Inspection
Decision making
Trajectories
Statistics
Infection
Sepsis
Critical Illness
Medical Records
Decision Making
Mortality
Therapeutics

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Information Systems and Management
  • Media Technology
  • Medicine (miscellaneous)
  • Health Informatics

Cite this

Gomov, M., Chou, J. K., Li, J. K., Sen, S., Cho, K., Tran, N., & Ma, K-L. (2018). Aiding infection analysis and diagnosis through temporally-contextualized matrix representations. In 2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017 (pp. 31-38). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/VAHC.2017.8387498

Aiding infection analysis and diagnosis through temporally-contextualized matrix representations. / Gomov, Maksim; Chou, Jia Kai; Li, Jianping Kelvin; Sen, Soman; Cho, Kiho; Tran, Nam; Ma, Kwan-Liu.

2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017. Institute of Electrical and Electronics Engineers Inc., 2018. p. 31-38.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gomov, M, Chou, JK, Li, JK, Sen, S, Cho, K, Tran, N & Ma, K-L 2018, Aiding infection analysis and diagnosis through temporally-contextualized matrix representations. in 2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017. Institute of Electrical and Electronics Engineers Inc., pp. 31-38, 8th Annual IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017, Phoenix, United States, 10/1/17. https://doi.org/10.1109/VAHC.2017.8387498
Gomov M, Chou JK, Li JK, Sen S, Cho K, Tran N et al. Aiding infection analysis and diagnosis through temporally-contextualized matrix representations. In 2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017. Institute of Electrical and Electronics Engineers Inc. 2018. p. 31-38 https://doi.org/10.1109/VAHC.2017.8387498
Gomov, Maksim ; Chou, Jia Kai ; Li, Jianping Kelvin ; Sen, Soman ; Cho, Kiho ; Tran, Nam ; Ma, Kwan-Liu. / Aiding infection analysis and diagnosis through temporally-contextualized matrix representations. 2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 31-38
@inproceedings{48b076ca07b541dea5b44b3c8783f0d6,
title = "Aiding infection analysis and diagnosis through temporally-contextualized matrix representations",
abstract = "Determining infections and sepsis of severely burned adults in a timely fashion allows clinicians to provide necessary treatments to critically ill patients, potentially reducing the chance of mortality. In current practice, clinicians examine large amounts of heterogeneous medical records using a spreadsheet-like representation and perform analysis of descriptive statistics to aid in their decision making for sepsis diagnosis. A more efficient approach is required to streamline such a process; accordingly, we developed an interactive visual interface for supporting quick inspection and comparison of patients' retrospective clinical trajectories. In particular, we employ a timeline representation to present entire treatment contexts for individual patients, and an aggregated matrix representation for summarizing multiple data variables of individual patients over time. This provides clinicians with a compact and intuitive way to discover the important trends, patterns, and events that occur in the context of multiple patients. We present several possible use cases identified by clinicians using our system and show that our preliminary results have the potential to greatly improve diagnostic timing and accuracy of sepsis identification in critically ill patients.",
author = "Maksim Gomov and Chou, {Jia Kai} and Li, {Jianping Kelvin} and Soman Sen and Kiho Cho and Nam Tran and Kwan-Liu Ma",
year = "2018",
month = "6",
day = "15",
doi = "10.1109/VAHC.2017.8387498",
language = "English (US)",
pages = "31--38",
booktitle = "2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Aiding infection analysis and diagnosis through temporally-contextualized matrix representations

AU - Gomov, Maksim

AU - Chou, Jia Kai

AU - Li, Jianping Kelvin

AU - Sen, Soman

AU - Cho, Kiho

AU - Tran, Nam

AU - Ma, Kwan-Liu

PY - 2018/6/15

Y1 - 2018/6/15

N2 - Determining infections and sepsis of severely burned adults in a timely fashion allows clinicians to provide necessary treatments to critically ill patients, potentially reducing the chance of mortality. In current practice, clinicians examine large amounts of heterogeneous medical records using a spreadsheet-like representation and perform analysis of descriptive statistics to aid in their decision making for sepsis diagnosis. A more efficient approach is required to streamline such a process; accordingly, we developed an interactive visual interface for supporting quick inspection and comparison of patients' retrospective clinical trajectories. In particular, we employ a timeline representation to present entire treatment contexts for individual patients, and an aggregated matrix representation for summarizing multiple data variables of individual patients over time. This provides clinicians with a compact and intuitive way to discover the important trends, patterns, and events that occur in the context of multiple patients. We present several possible use cases identified by clinicians using our system and show that our preliminary results have the potential to greatly improve diagnostic timing and accuracy of sepsis identification in critically ill patients.

AB - Determining infections and sepsis of severely burned adults in a timely fashion allows clinicians to provide necessary treatments to critically ill patients, potentially reducing the chance of mortality. In current practice, clinicians examine large amounts of heterogeneous medical records using a spreadsheet-like representation and perform analysis of descriptive statistics to aid in their decision making for sepsis diagnosis. A more efficient approach is required to streamline such a process; accordingly, we developed an interactive visual interface for supporting quick inspection and comparison of patients' retrospective clinical trajectories. In particular, we employ a timeline representation to present entire treatment contexts for individual patients, and an aggregated matrix representation for summarizing multiple data variables of individual patients over time. This provides clinicians with a compact and intuitive way to discover the important trends, patterns, and events that occur in the context of multiple patients. We present several possible use cases identified by clinicians using our system and show that our preliminary results have the potential to greatly improve diagnostic timing and accuracy of sepsis identification in critically ill patients.

UR - http://www.scopus.com/inward/record.url?scp=85050149109&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050149109&partnerID=8YFLogxK

U2 - 10.1109/VAHC.2017.8387498

DO - 10.1109/VAHC.2017.8387498

M3 - Conference contribution

SP - 31

EP - 38

BT - 2017 IEEE Workshop on Visual Analytics in Healthcare, VAHC 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -