Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics
Abstract Objective Differential diagnosis between epileptic seizures and psychogenic nonepileptic events (PNEEs) is a worldwide problem for neurologists. The present study aims to identify important characteristics from body fluid tests and develop diagnostic models based on them. Methods This is a...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2023-09-01
|
Series: | Epilepsia Open |
Subjects: | |
Online Access: | https://doi.org/10.1002/epi4.12775 |
_version_ | 1797729166247854080 |
---|---|
author | Yilin Xia Wanlin Lai Shihai Li Zhining Wen Lei Chen |
author_facet | Yilin Xia Wanlin Lai Shihai Li Zhining Wen Lei Chen |
author_sort | Yilin Xia |
collection | DOAJ |
description | Abstract Objective Differential diagnosis between epileptic seizures and psychogenic nonepileptic events (PNEEs) is a worldwide problem for neurologists. The present study aims to identify important characteristics from body fluid tests and develop diagnostic models based on them. Methods This is a register‐based observational study in patients with a diagnosis of epilepsy or PNEEs at West China Hospital of Sichuan University. Data from body fluid tests between 2009 and 2019 were used as a training set. We constructed models with a random forest approach in eight training subsets divided by sex and categories of tests, including electrolyte, blood cell, metabolism, and urine tests. Then, we collected data prospectively from patients between 2020 and 2022 to validate our models and calculated the relative importance of characteristics in robust models. Selected characteristics were finally analyzed with multiple logistic regression to establish nomograms. Results A total of 388 patients, including 218 with epilepsy and 170 with PNEEs, were studied. The AUROCs of random forest models of electrolyte and urine tests in the validation phase achieved 80.0% and 79.0%, respectively. Carbon dioxide combining power, anion gap, potassium, calcium, and chlorine in electrolyte tests and specific gravity, pH, and conductivity in urine tests were selected for the logistic regression analysis. C (ROC) of the electrolyte and urine diagnostic nomograms achieved 0.79 and 0.85, respectively. Significance The application of routine indicators of serum and urine may help in the more accurate identification of epileptic and PNEEs. |
first_indexed | 2024-03-12T11:25:06Z |
format | Article |
id | doaj.art-5621eed7bf8e449594e86f2620ebd42b |
institution | Directory Open Access Journal |
issn | 2470-9239 |
language | English |
last_indexed | 2024-03-12T11:25:06Z |
publishDate | 2023-09-01 |
publisher | Wiley |
record_format | Article |
series | Epilepsia Open |
spelling | doaj.art-5621eed7bf8e449594e86f2620ebd42b2023-09-01T09:39:51ZengWileyEpilepsia Open2470-92392023-09-018395996810.1002/epi4.12775Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristicsYilin Xia0Wanlin Lai1Shihai Li2Zhining Wen3Lei Chen4Department of Neurology, West China Hospital Sichuan University Chengdu ChinaDepartment of Neurology, West China Hospital Sichuan University Chengdu ChinaCollege of Chemistry Sichuan University Chengdu ChinaCollege of Chemistry Sichuan University Chengdu ChinaDepartment of Neurology, West China Hospital Sichuan University Chengdu ChinaAbstract Objective Differential diagnosis between epileptic seizures and psychogenic nonepileptic events (PNEEs) is a worldwide problem for neurologists. The present study aims to identify important characteristics from body fluid tests and develop diagnostic models based on them. Methods This is a register‐based observational study in patients with a diagnosis of epilepsy or PNEEs at West China Hospital of Sichuan University. Data from body fluid tests between 2009 and 2019 were used as a training set. We constructed models with a random forest approach in eight training subsets divided by sex and categories of tests, including electrolyte, blood cell, metabolism, and urine tests. Then, we collected data prospectively from patients between 2020 and 2022 to validate our models and calculated the relative importance of characteristics in robust models. Selected characteristics were finally analyzed with multiple logistic regression to establish nomograms. Results A total of 388 patients, including 218 with epilepsy and 170 with PNEEs, were studied. The AUROCs of random forest models of electrolyte and urine tests in the validation phase achieved 80.0% and 79.0%, respectively. Carbon dioxide combining power, anion gap, potassium, calcium, and chlorine in electrolyte tests and specific gravity, pH, and conductivity in urine tests were selected for the logistic regression analysis. C (ROC) of the electrolyte and urine diagnostic nomograms achieved 0.79 and 0.85, respectively. Significance The application of routine indicators of serum and urine may help in the more accurate identification of epileptic and PNEEs.https://doi.org/10.1002/epi4.12775differential diagnosisepilepsymachine learningpsychogenic nonepileptic events |
spellingShingle | Yilin Xia Wanlin Lai Shihai Li Zhining Wen Lei Chen Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics Epilepsia Open differential diagnosis epilepsy machine learning psychogenic nonepileptic events |
title | Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics |
title_full | Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics |
title_fullStr | Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics |
title_full_unstemmed | Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics |
title_short | Differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics |
title_sort | differentiation of epilepsy and psychogenic nonepileptic events based on body fluid characteristics |
topic | differential diagnosis epilepsy machine learning psychogenic nonepileptic events |
url | https://doi.org/10.1002/epi4.12775 |
work_keys_str_mv | AT yilinxia differentiationofepilepsyandpsychogenicnonepilepticeventsbasedonbodyfluidcharacteristics AT wanlinlai differentiationofepilepsyandpsychogenicnonepilepticeventsbasedonbodyfluidcharacteristics AT shihaili differentiationofepilepsyandpsychogenicnonepilepticeventsbasedonbodyfluidcharacteristics AT zhiningwen differentiationofepilepsyandpsychogenicnonepilepticeventsbasedonbodyfluidcharacteristics AT leichen differentiationofepilepsyandpsychogenicnonepilepticeventsbasedonbodyfluidcharacteristics |