Identifying epidermal growth factor receptor mutation status in patients with lung adenocarcinoma by

Jun-Feng Xiong1, Tian-Ying Jia2, Xiao-Yang Li2

  • 11 Department of Biomedical Engineering, School of Biomedical Engineering, Shanghai Jiao Tong University , Shanghai , China.

Abstract

Insights

This study used deep learning (CNN) to analyze CT scans for predicting epidermal growth factor receptor (EGFR) gene mutations in lung adenocarcinoma patients, aiding treatment decisions.

Area of Science:

  • Oncology
  • Radiology
  • Artificial Intelligence

Background:

  • Genetic mutations, particularly in the epidermal growth factor receptor (EGFR) gene, are crucial for guiding treatment decisions in lung adenocarcinoma.
  • Identifying EGFR mutation status non-invasively can significantly impact patient management and treatment selection.

Purpose of the Study:

  • To develop and evaluate three-dimensional convolutional neural networks (CNNs) for predicting EGFR gene mutation status in lung adenocarcinoma patients using CT images.
  • To assess the potential of CNNs to identify imaging patterns indicative of EGFR mutations that are not discernible to the human eye.

Main Methods:

  • A cohort of 503 Chinese lung adenocarcinoma patients who underwent surgery between 2012 and 2015 was retrospectively analyzed.
  • Three-dimensional CNNs were trained and validated on CT images acquired before surgery to predict EGFR mutations (exon 19 deletion or exon 21 L858R).
  • Performance was evaluated using metrics including area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity.

Main Results:

  • The CNN model achieved an AUC of 0.776 in the independent validation set.
  • A fusion model combining CNNs with clinical features (sex, smoking history) improved performance, yielding an AUC of 0.838, with 77.2% accuracy, 75.8% sensitivity, and 79.1% specificity.
  • These results indicate the CNN's ability to identify potential indicators of EGFR mutation status from CT images.

Conclusions:

  • CNNs demonstrate potential in predicting EGFR mutation status in lung adenocarcinoma, offering a non-invasive approach to aid clinical decision-making.
  • Further improvements in CNN performance can be achieved through larger training datasets and optimized network architectures.
  • AI-based analysis of medical images holds promise for revealing spatial heterogeneity relevant to cancer characteristics.

Related Concept Videos

Mutations01:39

Mutations

Overview
94.5K
Factors Affecting Protein-Drug Binding: Patient-Related Factors01:29

Factors Affecting Protein-Drug Binding: Patient-Related Factors

Protein-drug binding, a pivotal aspect of pharmacokinetics, is subject to considerable variability influenced by an array of patient-related factors. The intricate interplay of age, individual differences, and pathological conditions significantly impact the binding dynamics and subsequent pharmacological effects.
Age stands as a key determinant in protein-drug binding. Neonates, characterized by low albumin content, experience heightened concentrations of unbound drugs such as phenytoin and...
327
Convolution Properties II01:17

Convolution Properties II

The important convolution properties include width, area, differentiation, and integration properties.
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
589
Role of Hematopoietic Growth Factors01:28

Role of Hematopoietic Growth Factors

Hematopoietic growth factors are molecules that regulate the differentiation rate of hematopoietic stem cells (HSCs). Erythropoietin (EPO), primarily produced by the kidneys, plays a crucial role in erythrocyte production. When oxygen levels in the blood are low, EPO is released into the bloodstream, reaching the bone marrow, where it stimulates HSCs to differentiate and mature into erythrocytes, which are vital for oxygen transport.
Thrombopoietin (TPO), mainly released by the liver,...
3.9K
Factors Influencing Microbial Growth: pH01:29

Factors Influencing Microbial Growth: pH

Microorganisms are classified as acidophiles, neutrophiles, or alkaliphiles based on their pH growth preferences, reflecting their adaptations to specific environments. Maintaining a stable intracellular pH is critical for macromolecular stability and enzymatic activity, which can be challenged by external pH variations.Neutrophiles, such as Escherichia coli, grow optimally between pH 5.5 and 8.0. These microorganisms inhabit neutral or slightly acidic environments and employ mechanisms like...
1.2K
Network Covalent Solids02:18

Network Covalent Solids

Network covalent solids contain a three-dimensional network of covalently bonded atoms as found in the crystal structures of nonmetals like diamond, graphite, silicon, and some covalent compounds, such as silicon dioxide (sand) and silicon carbide (carborundum, the abrasive on sandpaper). Many minerals have networks of covalent bonds.
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
16.2K