Chest X-Ray Abnormality Detection by Using Artificial Intelligence: A Single-Site Retrospective Study of Deep Learning Model Performance

KVAK, Daniel, Anna CHROMCOVÁ, Marek BIROŠ, Robert HRUBÝ, Karolína KVAKOVÁ, Marija PAJDAKOVIĆ a Petra OVESNÁ. Chest X-Ray Abnormality Detection by Using Artificial Intelligence: A Single-Site Retrospective Study of Deep Learning Model Performance. Online. BioMedInformatics. Basel, Switzerland: MDPI, 2023, roč. 3, č. 1, s. 82-101. Dostupné z: https://dx.doi.org/10.3390/biomedinformatics3010006.

Základní údaje

• Originální název: Chest X-Ray Abnormality Detection by Using Artificial Intelligence: A Single-Site Retrospective Study of Deep Learning Model Performance
• Název anglicky: Chest X-Ray Abnormality Detection by Using Artificial Intelligence: A Single-Site Retrospective Study of Deep Learning Model Performance
• Autoři: KVAK, Daniel, Anna CHROMCOVÁ, Marek BIROŠ, Robert HRUBÝ, Karolína KVAKOVÁ, Marija PAJDAKOVIĆ a Petra OVESNÁ.
• Vydání: BioMedInformatics, Basel, Switzerland, MDPI, 2023.

Další údaje

• Typ výsledku: Článek v odborném periodiku
• Utajení: není předmětem státního či obchodního tajemství
• Doi: http://dx.doi.org/10.3390/biomedinformatics3010006
• Klíčová slova anglicky: artificial intelligence; computer-aided detection; deep learning; chest X-ray; patient prioritization
• Štítky: RIV ne
• Příznaky: Mezinárodní význam, Recenzováno

Anotace

Chest X-ray (CXR) is one of the most common radiological examination for both non-emergent and emergent clinical indications, but human error or lack of prioritization of patients can hinder timely interpretation. Deep learning (DL) algorithms have been proven to be useful in the assessment of various abnormalities including tuberculosis, lung parenchymal lesions or pneumothorax. Carebot AI CXR (v1.22) was developed to detect visual patterns on CXR for 12 preselected findings. To evaluate the proposed system, we designed a single-site retrospective study comparing the DL algorithm with the performance of 5 differently experienced radiologists. On the assessed dataset (n=127) collected from municipal hospital in the Czech Republic, Carebot AI CXR achieved sensitivity (Se) of 0.925 and specificity (Sp) of 0.644, compared to bootstrapped radiologists’ Se of 0.661 and Sp of 0.803, respectively, with statistically significant difference. The negative likelihood ratio (NLR) of the proposed software (0.12 (0.04-0.32)) was significantly lower than radiologists’ assessment (0.42 (0.4-0.43), p<0.0001). No critical findings were missed by the software.

Anotace anglicky

Chest X-ray (CXR) is one of the most common radiological examination for both non-emergent and emergent clinical indications, but human error or lack of prioritization of patients can hinder timely interpretation. Deep learning (DL) algorithms have been proven to be useful in the assessment of various abnormalities including tuberculosis, lung parenchymal lesions or pneumothorax. Carebot AI CXR (v1.22) was developed to detect visual patterns on CXR for 12 preselected findings. To evaluate the proposed system, we designed a single-site retrospective study comparing the DL algorithm with the performance of 5 differently experienced radiologists. On the assessed dataset (n=127) collected from municipal hospital in the Czech Republic, Carebot AI CXR achieved sensitivity (Se) of 0.925 and specificity (Sp) of 0.644, compared to bootstrapped radiologists’ Se of 0.661 and Sp of 0.803, respectively, with statistically significant difference. The negative likelihood ratio (NLR) of the proposed software (0.12 (0.04-0.32)) was significantly lower than radiologists’ assessment (0.42 (0.4-0.43), p<0.0001). No critical findings were missed by the software.