Deep Learning Algorithms for Diagnosis of Lung Cancer: A Systematic Review and Meta-Analysis

SIMPLE SUMMARY: Lung cancer screening has been shown to help reduce mortality in selected populations of smokers; however, performing screening programs at a larger scale with high accuracy is still a challenge. The use of artificial intelligence (AI) has been investigated to improve large scale scr...

Descripción completa

Detalles Bibliográficos
Autores principales: Forte, Gabriele C., Altmayer, Stephan, Silva, Ricardo F., Stefani, Mariana T., Libermann, Lucas L., Cavion, Cesar C., Youssef, Ali, Forghani, Reza, King, Jeremy, Mohamed, Tan-Lucien, Andrade, Rubens G. F., Hochhegger, Bruno
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2022
Materias:
Systematic Review
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9405626/
https://www.ncbi.nlm.nih.gov/pubmed/36010850
http://dx.doi.org/10.3390/cancers14163856
Descripción
Sumario:SIMPLE SUMMARY: Lung cancer screening has been shown to help reduce mortality in selected populations of smokers; however, performing screening programs at a larger scale with high accuracy is still a challenge. The use of artificial intelligence (AI) has been investigated to improve large scale screening. We have performed a meta-analysis of the diagnostic accuracy of deep learning (DL) algorithms to diagnose lung cancer. Combining six eligible studies, the pooled sensitivity and specificity of DL algorithms were 0.93 (95% CI 0.85–0.98) and 0.68 (95% CI 0.49–0.84), respectively. Despite remaining challenges in the field, AI is likely to play an important role in disease screening in the future. ABSTRACT: We conducted a systematic review and meta-analysis of the diagnostic performance of current deep learning algorithms for the diagnosis of lung cancer. We searched major databases up to June 2022 to include studies that used artificial intelligence to diagnose lung cancer, using the histopathological analysis of true positive cases as a reference. The quality of the included studies was assessed independently by two authors based on the revised Quality Assessment of Diagnostic Accuracy Studies. Six studies were included in the analysis. The pooled sensitivity and specificity were 0.93 (95% CI 0.85–0.98) and 0.68 (95% CI 0.49–0.84), respectively. Despite the significantly high heterogeneity for sensitivity (I(2) = 94%, p < 0.01) and specificity (I(2) = 99%, p < 0.01), most of it was attributed to the threshold effect. The pooled SROC curve with a bivariate approach yielded an area under the curve (AUC) of 0.90 (95% CI 0.86 to 0.92). The DOR for the studies was 26.7 (95% CI 19.7–36.2) and heterogeneity was 3% (p = 0.40). In this systematic review and meta-analysis, we found that when using the summary point from the SROC, the pooled sensitivity and specificity of DL algorithms for the diagnosis of lung cancer were 93% and 68%, respectively.

Ejemplares similares