Neural Network Modeling to Predict Intact Parathyroid Hormone in Uremic Patients on Continuous Ambulatory Peritoneal Dialysis

Jainn-Shiun Chiu, Wei-Tung Lin, Yu-Chuan Li, Yuh-Feng Wang

Research output: Contribution to journalArticle

Abstract

Background: Measuring plasma intact parathyroid hormone (iPTH) concentration is essential to evaluate renal osteodystrophy. Although frequent measurement is needed to avoid inadequate prescription of phosphate binder and vitamin D preparations, it is not cost- effective in some clinics. For this purpose, we developed an artificial neural network (ANN) to predict plasma iPTH concentration in uremic patients on continuous ambulatory peritoneal dialysis (CAPD). Methods: The study population consisted of 23 stable patients (11 male and 12 female, aged 48.8±15.3 years) on CAPD for more than 3 months. Among ANN models, the predictors included plasma calcium, phosphate, alkaline phosphatase concentrations, and calcium-phosphate product. The dependent variable was plasma iPTH concentration measured by radioimmunoassay (RIA-iPTH). Leave-one-out cross-validation was adopted to iron out generalization problems caused by finite population. The least ratio of standard deviation (SDR) was used to choose the best ANN model. For comparing the performance between predictive plasma iPTH concentration by ANN (ANN-iPTH) and RIA-iPTH, the correlation coefficient (r), mean error, and Passing and Bablok regression were evaluated. Results: The generalized regression neural network (SDR=0.74) was the final best ANN model. The relationship between RIA-iPTH and ANN-iPTH is described by Passing and Bablok regression ANN-iPTH=90.52+0.55×RIA-iPTH, with 95% confidence interval for intercept 23.08 to 122.83 and for slope 0.30 to 1.16, indicating that both methods are interchangeable without statistically significant deviation (P>0.10). Conclusion: ANN can accurately predict plasma iPTH concentration in uremic patients on CAPD. It is useful and beneficial to assess renal osteodystrophy frequently and led to proper treatment.
Original languageEnglish
Pages (from-to)135-141
Number of pages7
Journal核子醫學雜誌
Volume18
Issue number3
DOIs
Publication statusPublished - 2005

Fingerprint

Continuous Ambulatory Peritoneal Dialysis
Parathyroid Hormone
Neural Networks (Computer)
Chronic Kidney Disease-Mineral and Bone Disorder
Vitamin D
Population
Radioimmunoassay
Prescriptions
Alkaline Phosphatase
Iron
Phosphates

Keywords

  • 類神經網路
  • 完整副甲狀腺素
  • 連續性可攜式腹膜透析
  • neural network
  • intact parathyroid hormone
  • continuous ambulatory peritoneal dialysis

Cite this

Neural Network Modeling to Predict Intact Parathyroid Hormone in Uremic Patients on Continuous Ambulatory Peritoneal Dialysis. / Chiu, Jainn-Shiun ; Lin, Wei-Tung; Li, Yu-Chuan; Wang, Yuh-Feng .

In: 核子醫學雜誌, Vol. 18, No. 3, 2005, p. 135-141.

Research output: Contribution to journalArticle

@article{84596c4c703b4882a332a83c44686ff4,
title = "Neural Network Modeling to Predict Intact Parathyroid Hormone in Uremic Patients on Continuous Ambulatory Peritoneal Dialysis",
abstract = "Background: Measuring plasma intact parathyroid hormone (iPTH) concentration is essential to evaluate renal osteodystrophy. Although frequent measurement is needed to avoid inadequate prescription of phosphate binder and vitamin D preparations, it is not cost- effective in some clinics. For this purpose, we developed an artificial neural network (ANN) to predict plasma iPTH concentration in uremic patients on continuous ambulatory peritoneal dialysis (CAPD). Methods: The study population consisted of 23 stable patients (11 male and 12 female, aged 48.8±15.3 years) on CAPD for more than 3 months. Among ANN models, the predictors included plasma calcium, phosphate, alkaline phosphatase concentrations, and calcium-phosphate product. The dependent variable was plasma iPTH concentration measured by radioimmunoassay (RIA-iPTH). Leave-one-out cross-validation was adopted to iron out generalization problems caused by finite population. The least ratio of standard deviation (SDR) was used to choose the best ANN model. For comparing the performance between predictive plasma iPTH concentration by ANN (ANN-iPTH) and RIA-iPTH, the correlation coefficient (r), mean error, and Passing and Bablok regression were evaluated. Results: The generalized regression neural network (SDR=0.74) was the final best ANN model. The relationship between RIA-iPTH and ANN-iPTH is described by Passing and Bablok regression ANN-iPTH=90.52+0.55×RIA-iPTH, with 95{\%} confidence interval for intercept 23.08 to 122.83 and for slope 0.30 to 1.16, indicating that both methods are interchangeable without statistically significant deviation (P>0.10). Conclusion: ANN can accurately predict plasma iPTH concentration in uremic patients on CAPD. It is useful and beneficial to assess renal osteodystrophy frequently and led to proper treatment.",
keywords = "類神經網路, 完整副甲狀腺素, 連續性可攜式腹膜透析, neural network, intact parathyroid hormone, continuous ambulatory peritoneal dialysis",
author = "Jainn-Shiun Chiu and Wei-Tung Lin and Yu-Chuan Li and Yuh-Feng Wang",
year = "2005",
doi = "10.6332/ANMS.1803.002",
language = "English",
volume = "18",
pages = "135--141",
journal = "核子醫學雜誌",
issn = "1022-923X",
publisher = "中華民國核子醫學學會",
number = "3",

}

TY - JOUR

T1 - Neural Network Modeling to Predict Intact Parathyroid Hormone in Uremic Patients on Continuous Ambulatory Peritoneal Dialysis

AU - Chiu, Jainn-Shiun

AU - Lin, Wei-Tung

AU - Li, Yu-Chuan

AU - Wang, Yuh-Feng

PY - 2005

Y1 - 2005

N2 - Background: Measuring plasma intact parathyroid hormone (iPTH) concentration is essential to evaluate renal osteodystrophy. Although frequent measurement is needed to avoid inadequate prescription of phosphate binder and vitamin D preparations, it is not cost- effective in some clinics. For this purpose, we developed an artificial neural network (ANN) to predict plasma iPTH concentration in uremic patients on continuous ambulatory peritoneal dialysis (CAPD). Methods: The study population consisted of 23 stable patients (11 male and 12 female, aged 48.8±15.3 years) on CAPD for more than 3 months. Among ANN models, the predictors included plasma calcium, phosphate, alkaline phosphatase concentrations, and calcium-phosphate product. The dependent variable was plasma iPTH concentration measured by radioimmunoassay (RIA-iPTH). Leave-one-out cross-validation was adopted to iron out generalization problems caused by finite population. The least ratio of standard deviation (SDR) was used to choose the best ANN model. For comparing the performance between predictive plasma iPTH concentration by ANN (ANN-iPTH) and RIA-iPTH, the correlation coefficient (r), mean error, and Passing and Bablok regression were evaluated. Results: The generalized regression neural network (SDR=0.74) was the final best ANN model. The relationship between RIA-iPTH and ANN-iPTH is described by Passing and Bablok regression ANN-iPTH=90.52+0.55×RIA-iPTH, with 95% confidence interval for intercept 23.08 to 122.83 and for slope 0.30 to 1.16, indicating that both methods are interchangeable without statistically significant deviation (P>0.10). Conclusion: ANN can accurately predict plasma iPTH concentration in uremic patients on CAPD. It is useful and beneficial to assess renal osteodystrophy frequently and led to proper treatment.

AB - Background: Measuring plasma intact parathyroid hormone (iPTH) concentration is essential to evaluate renal osteodystrophy. Although frequent measurement is needed to avoid inadequate prescription of phosphate binder and vitamin D preparations, it is not cost- effective in some clinics. For this purpose, we developed an artificial neural network (ANN) to predict plasma iPTH concentration in uremic patients on continuous ambulatory peritoneal dialysis (CAPD). Methods: The study population consisted of 23 stable patients (11 male and 12 female, aged 48.8±15.3 years) on CAPD for more than 3 months. Among ANN models, the predictors included plasma calcium, phosphate, alkaline phosphatase concentrations, and calcium-phosphate product. The dependent variable was plasma iPTH concentration measured by radioimmunoassay (RIA-iPTH). Leave-one-out cross-validation was adopted to iron out generalization problems caused by finite population. The least ratio of standard deviation (SDR) was used to choose the best ANN model. For comparing the performance between predictive plasma iPTH concentration by ANN (ANN-iPTH) and RIA-iPTH, the correlation coefficient (r), mean error, and Passing and Bablok regression were evaluated. Results: The generalized regression neural network (SDR=0.74) was the final best ANN model. The relationship between RIA-iPTH and ANN-iPTH is described by Passing and Bablok regression ANN-iPTH=90.52+0.55×RIA-iPTH, with 95% confidence interval for intercept 23.08 to 122.83 and for slope 0.30 to 1.16, indicating that both methods are interchangeable without statistically significant deviation (P>0.10). Conclusion: ANN can accurately predict plasma iPTH concentration in uremic patients on CAPD. It is useful and beneficial to assess renal osteodystrophy frequently and led to proper treatment.

KW - 類神經網路

KW - 完整副甲狀腺素

KW - 連續性可攜式腹膜透析

KW - neural network

KW - intact parathyroid hormone

KW - continuous ambulatory peritoneal dialysis

U2 - 10.6332/ANMS.1803.002

DO - 10.6332/ANMS.1803.002

M3 - Article

VL - 18

SP - 135

EP - 141

JO - 核子醫學雜誌

JF - 核子醫學雜誌

SN - 1022-923X

IS - 3

ER -