This paper presents a system supporting clinical decisions for patients with Chronic Obstructive Pulmonary Disease (COPD). The system should partially fill the gaps highlighted during an analysis of the current state of the art of Clinical Decision Support Systems (CDSS) for telemonitoring patients affected by COPD. The first step taken was to replicate the performance of similar decision support systems found in the scientific literature. Using physiological parameters drawn from respiratory function tests on 414 patients, two predictive models were created using two machine-learning algorithms: neural network and support vector machine. Performance was comparable to that described in the literature. The results made it possible to affirm that the data available were sufficient to evaluate the extent of ventilatory deficit. The next step was to create a new predictive model with better performance than previously obtained. The C5.0 Machine Learning Algorithm was chosen for the development of the model. The resulting performance on the data available was significantly better than with the two previous models. This new predictive model, called COPD, was then implemented in a user interface created using Java programming language. The new software developed, which enables the evaluation and classification of respiratory test results and which can be used in many clinical applications, provides excellent performance compared to the current state of the art.
An Automatic System Supporting Clinical Decision for Chronic Obstructive Pulmonary Disease / Ernesto Iadanza, Vlad Mudura, Paolo Melillo, Monica Gherardelli. - In: HEALTH AND TECHNOLOGY. - ISSN 2190-7188. - ELETTRONICO. - 10:(2019), pp. 487-498. [10.1007/s12553-019-00312-9]
An Automatic System Supporting Clinical Decision for Chronic Obstructive Pulmonary Disease
Ernesto Iadanza;Vlad Mudura;Monica Gherardelli
2019
Abstract
This paper presents a system supporting clinical decisions for patients with Chronic Obstructive Pulmonary Disease (COPD). The system should partially fill the gaps highlighted during an analysis of the current state of the art of Clinical Decision Support Systems (CDSS) for telemonitoring patients affected by COPD. The first step taken was to replicate the performance of similar decision support systems found in the scientific literature. Using physiological parameters drawn from respiratory function tests on 414 patients, two predictive models were created using two machine-learning algorithms: neural network and support vector machine. Performance was comparable to that described in the literature. The results made it possible to affirm that the data available were sufficient to evaluate the extent of ventilatory deficit. The next step was to create a new predictive model with better performance than previously obtained. The C5.0 Machine Learning Algorithm was chosen for the development of the model. The resulting performance on the data available was significantly better than with the two previous models. This new predictive model, called COPD, was then implemented in a user interface created using Java programming language. The new software developed, which enables the evaluation and classification of respiratory test results and which can be used in many clinical applications, provides excellent performance compared to the current state of the art.File | Dimensione | Formato | |
---|---|---|---|
HEAL-D-19-00034.pdf
Accesso chiuso
Descrizione: Articolo principale
Tipologia:
Versione finale referata (Postprint, Accepted manuscript)
Licenza:
Tutti i diritti riservati
Dimensione
1.14 MB
Formato
Adobe PDF
|
1.14 MB | Adobe PDF | Richiedi una copia |
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.