Abstract:Lung cancer is the second most prevalent and deadliest cancer in the United States, primarily due to its elusive early symptoms that hinder timely diagnosis. Lung nodules, minute anomalous areas, hold potential significance in lung cancer occurrence. Swift identification of these nodules can significantly enhance patient survival rates. Thoracic thin-sliced Computed Tomography (CT) scanning has emerged as a widely adopted approach for radiologists' diagnosing and prognosticating lung abnormalities. However, human factors can be prone to errors stemming from the multifarious causes underlying nodule formation, including factors like pollutants and infections. The domain of deep learning algorithms has recently showcased remarkable prowess in classifying and segmenting medical images. This study is geared towards the creation of a comprehensive framework that seamlessly integrates explainable AI techniques to achieve precise pulmonary nodule detection for computer-aided detection and diagnosis (CAD) systems. The framework's execution is underpinned by an explanation supervision approach, employing radiologists' nodule contours as guiding signals to imbue the model with an understanding of nodule morphologies. The nodule detection is achieved by augmentation of the model's learning capabilities, particularly on smaller datasets. Furthermore, two imputation techniques are deployed to refine the nodule region annotations, thereby attenuating noise inherent in human annotations. These techniques empower the model with steadfast attributions that align with human expectations. Our approach consistently elevates ResNet18's classification performance and explanation quality regarding accuracy and specificity by 22% and 23% using training datasets with 100 samples. For small dataset learning using 50 samples, the proposed framework still outperforms ResNet18 by 16% regarding accuracy and specificity. This framework seamlessly integrates a robust explanation supervision technique, guaranteeing nodule classification and morphology assessment accuracy. Reducing the burden on radiologists allows them to focus on diagnosing and prognosticating potentially cancerous pulmonary nodules at an early stage, consequently augmenting survival rates in lung cancer.

AIAT: Adaptive Iteration Adversarial Training for Robust Pulmonary Nodule Detection

A 3D Multi-scale Virtual Adversarial Network for False Positive Reduction in Pulmonary Nodule Detection

A Simple Method to Train the AI Diagnosis Model of Pulmonary Nodules.

Semi-supervised lung nodule detection with adversarial learning

No Surprises: Training Robust Lung Nodule Detection for Low-Dose CT Scans by Augmenting With Adversarial Attacks

Lung Nodule Detection via Deep Reinforcement Learning

Robust and accurate pulmonary nodule detection with self-supervised feature learning on domain adaptation

AI in Lung Health: Benchmarking Detection and Diagnostic Models Across Multiple CT Scan Datasets

Class-Aware Adversarial Lung Nodule Synthesis in CT Images

3D ARCNN: an Asymmetric Residual CNN for False Positive Reduction in Pulmonary Nodule

Robust Deep Learning from Incomplete Annotation for Accurate Lung Nodule Detection

A-DARTS: Attention-Guided Differentiable Architecture Search for Lung Nodule Classification

Explainable AI for Lung Nodule Detection and Classification in CT Images

A Deep-Learning Based Automatic Pulmonary Nodule Detection System

A kind of deep learning acceleration method for pulmonary nodule detection

AFA: adversarial frequency alignment for domain generalized lung nodule detection

Research progress of artificial intelligence based on deep learning in the field of pulmonary nodules

LDNNET: Towards Robust Classification of Lung Nodule and Cancer Using Lung Dense Neural Network

Accurate and Robust Pulmonary Nodule Detection by 3D Feature Pyramid Network with Self-supervised Feature Learning

Anatomical Structure-Aware Pulmonary Nodule Detection via Parallel Multi-task RoI Head.

An End-to-End Framework for Integrated Pulmonary Nodule Detection and False Positive Reduction.