Buscar en
Radiología (English Edition)
Toda la web
Inicio Radiología (English Edition) Utility of artificial intelligence for detection of pneumothorax on chest radiop...
Journal Information
Vol. 66. Issue S1.
Suplement “"Advances in Thoracic Radiology”
Pages S40-S46 (April 2024)
Share
Share
Download PDF
More article options
Visits
1
Vol. 66. Issue S1.
Suplement “"Advances in Thoracic Radiology”
Pages S40-S46 (April 2024)
Original articles
Utility of artificial intelligence for detection of pneumothorax on chest radiopgraphs done after transthoracic percutaneous transthoracic biopsy guided by computed tomography
Utilidad de la inteligencia artificial para la detección de neumotórax en radiografías de tórax de control tras punción percutánea transtorácica guiada por tomografía computarizada
Visits
1
D. Ferrando Blanco
Corresponding author
davferbla@hotmail.com

Corresponding author.
, Ó. Persiva Morenza, L.B. Cabanzo Campos, A.L. Sánchez Martínez, D. Varona Porres, L.A. Del Carpio Bellido Vargas, J. Andreu Soriano
Servicio de Radiología, Hospital Universitari Vall d’Hebrón, Barcelona, Spain
This item has received
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Figures (5)
Show moreShow less
Tables (3)
Table 1. Sample characteristics included (n = 258 patients included and n = 284 procedures performed).
Table 2. Characteristics of complications.
Table 3. Overall performance of the program for detecting pneumothorax, with the radiologist reading as gold standard.
Show moreShow less
Special issue
This article is part of special issue:
Vol. 66. Issue S1

Suplement “"Advances in Thoracic Radiology”

More info
Abstract
Objetive

To assess the ability of an artificial intelligence software to detect pneumothorax in chest radiographs done after percutaneous transthoracic biopsy.

Material and methods

We included retrospectively in our study adult patients who underwent CT-guided percutaneous transthoracic biopsies from lung, pleural or mediastinal lesions from June 2019 to June 2020, and who had a follow-up chest radiograph after the procedure. These chest radiographs were read to search the presence of pneumothorax independently by an expert thoracic radiologist and a radiodiagnosis resident, whose unified lecture was defined as the gold standard, and the result of each radiograph after interpretation by the artificial intelligence software was documented for posterior comparison with the gold standard.

Results

A total of 284 chest radiographs were included in the study and the incidence of pneumothorax was 14.4%. There were no discrepancies between the two readers’ interpretation of any of the postbiopsy chest radiographs. The artificial intelligence software was able to detect 41/41 of the present pneumothorax, implying a sensitivity of 100% and a negative predictive value of 100%, with a specificity of 79.4% and a positive predictive value of 45%. The accuracy was 82.4%, indicating that there is a high probability that an individual will be adequately classified by the software. It has also been documented that the presence of Port-a-cath is the cause of 8 of the 50 of false positives by the software.

Conclusions

The software has detected 100% of cases of pneumothorax in the postbiopsy chest radiographs. A potential use of this software could be as a prioritisation tool, allowing radiologists not to read immediately (or even not to read) chest radiographs classified as non-pathological by the software, with the confidence that there are no pathological cases.

Keywords:
Chest radiograph
Artificial intelligence
Deep learning
Core needle biopsy
Fine needle aspiration
Pneumothorax
Resumen
Objetivo

Evaluar la capacidad de detección de neumotórax por parte de un programa de inteligencia artificial (IA) en las radiografías de tórax de control posbiopsia percutánea transtorácica.

Material y métodos

Se han incluido de forma retrospectiva los pacientes adultos sometidos a biopsia percutánea transtorácica de lesiones pulmonares, pleurales o mediastínicas guiada por TC desde junio de 2019 hasta junio de 2020 y que contaban con radiografía de tórax de control tras el procedimiento. Se procedió a la lectura independiente de las radiografías de tórax de control posprocedimiento por parte de un radiólogo torácico y un residente de radiodiagnóstico, cuya lectura unificada se definió como el gold estándar, y se anotó el resultado de cada radiografía de tórax tras la interpretación por un programa de IA para la detección de neumotórax.

Resultados

Se incluyeron 284 radiografías de tórax posprocedimiento, en las que la incidencia de neumotórax fue del 14,4%, sin existir discrepancias en el diagnóstico entre los dos lectores. El algoritmo de IA fue capaz de detectar 41/41 de los neumotórax presentes, suponiendo una sensibilidad del 100% y un valor predictivo negativo del 100%, a expensas de una especificidad del 79,4% y un valor predictivo positivo del 45%. La exactitud o valor global fue del 82,4%, indicando que hay una alta probabilidad de que un individuo sea clasificado adecuadamente por el modelo de IA. También se ha documentado que la presencia de Port-a-cath fue causa de 8 de los 50 falsos positivos.

Conclusiones

El programa ha detectado el 100% de casos de neumotórax en las radiografías de tórax de control. Un potencial uso de este software sería como herramienta de priorización, permitiendo postergar, o incluso obviar, la lectura de las radiografías de tórax clasificadas como no patológicas por el algoritmo con la garantía de que no haya casos patológicos.

Palabras clave:
Radiografía de tórax
Inteligencia artificial
Aprendizaje profundo
Biopsia con aguja gruesa
Punción aspiración con aguja fina
Neumotórax

Article

These are the options to access the full texts of the publication Radiología (English Edition)
Subscriber
Subscriber

If you already have your login data, please click here .

If you have forgotten your password you can you can recover it by clicking here and selecting the option “I have forgotten my password”
Purchase
Purchase article

Purchasing article the PDF version will be downloaded

Price 19.34 €

Purchase now
Contact
Phone for subscriptions and reporting of errors
From Monday to Friday from 9 a.m. to 6 p.m. (GMT + 1) except for the months of July and August which will be from 9 a.m. to 3 p.m.
Calls from Spain
932 415 960
Calls from outside Spain
+34 932 415 960
E-mail
Article options
Tools
es en pt

¿Es usted profesional sanitario apto para prescribir o dispensar medicamentos?

Are you a health professional able to prescribe or dispense drugs?

Você é um profissional de saúde habilitado a prescrever ou dispensar medicamentos