[論文レビュー] Automatic Detection of Coronavirus Disease (COVID-19) in X-ray and CT Images: A Machine Learning-Based Approach

Sara Hosseinzadeh Kassani, Peyman Hosseinzadeh Kassasni+3 more|arXiv (Cornell University)|Apr 22, 2020

COVID-19 diagnosis using AI参考文献 41被引用数 62

ひとこと要約

tldr: 本論文は、転移学習に基づく深層CNN特徴抽出器のアンサンブルを用い、それを伝統的な ML分類器と組み合わせて胸部X線およびCT画像からCOVID-19を検出する手法を評価し、公開データセットで最大99%の精度を達成した。

ABSTRACT

The newly identified Coronavirus pneumonia, subsequently termed COVID-19, is highly transmittable and pathogenic with no clinically approved antiviral drug or vaccine available for treatment. The most common symptoms of COVID-19 are dry cough, sore throat, and fever. Symptoms can progress to a severe form of pneumonia with critical complications, including septic shock, pulmonary edema, acute respiratory distress syndrome and multi-organ failure. While medical imaging is not currently recommended in Canada for primary diagnosis of COVID-19, computer-aided diagnosis systems could assist in the early detection of COVID-19 abnormalities and help to monitor the progression of the disease, potentially reduce mortality rates. In this study, we compare popular deep learning-based feature extraction frameworks for automatic COVID-19 classification. To obtain the most accurate feature, which is an essential component of learning, MobileNet, DenseNet, Xception, ResNet, InceptionV3, InceptionResNetV2, VGGNet, NASNet were chosen amongst a pool of deep convolutional neural networks. The extracted features were then fed into several machine learning classifiers to classify subjects as either a case of COVID-19 or a control. This approach avoided task-specific data pre-processing methods to support a better generalization ability for unseen data. The performance of the proposed method was validated on a publicly available COVID-19 dataset of chest X-ray and CT images. The DenseNet121 feature extractor with Bagging tree classifier achieved the best performance with 99% classification accuracy. The second-best learner was a hybrid of the a ResNet50 feature extractor trained by LightGBM with an accuracy of 98%.

研究の動機と目的

• Demonstrate a generic, non-handcrafted feature extraction approach using deep CNNs for COVID-19 classification on X-ray and CT images.
• Assess whether transferring knowledge from large pre-trained CNNs improves detection with limited COVID-19 data.
• Avoid extensive pre-processing to promote generalization across heterogeneous imaging sources.
• Identify combinations of CNN feature extractors and ML classifiers that maximize accuracy and efficiency.
• Provide a web-based CAD aid for rapid screening of suspected cases.

提案手法

• Use pre-trained CNN architectures (MobileNet, DenseNet, Xception, InceptionV3, InceptionResNetV2, ResNet, VGGNet, NASNet) as feature extractors (transfer learning) to encode images into low-dimensional feature vectors.
• Train multiple traditional ML classifiers (Decision Tree, Random Forest, XGBoost, AdaBoost, Bagging, LightGBM) on the CNN-derived features.
• Evaluate with 10-fold cross-validation on a public dataset of chest X-ray and CT images; weights initialized from ImageNet.
• Apply two minimal pre-processing steps: resizing to uniform dimensions and image normalization (ImageNet mean subtraction and min-max normalization).
• Report accuracy, precision, recall, and F1-score to compare CNN+ML combinations.

実験結果

リサーチクエスチョン

• RQ1 How well do deep CNN-derived feature representations combined with ML classifiers detect COVID-19 vs healthy controls in X-ray and CT images?
• RQ2 Which CNN architectures and ML classifiers yield the highest accuracy and reliability for this task?
• RQ3 Does minimal pre-processing and transfer learning provide robust generalization across heterogeneous image sources?
• RQ4 What are the trade-offs in computation time between feature extraction and classifier training for practical CAD deployment?

主な発見

DT	RF	XGBoost	AdaBoost	Bagging	LightGBM
83.00 ± 0.26	93.00 ± 0.23	95.00 ± 0.16	80.00 ± 0.17	96.00 ± -0.11	82.00 ± 0.28
92.00 ± 0.15	90.00 ± 0.21	94.00 ± 0.16	92.00 ± 0.19	99.00 ± 0.07	96.00 ± 0.11
84.00 ± 0.26	90.00 ± 0.24	90.00 ± 0.18	87.00 ± 0.25	96.00 ± 0.11	87.00 ± 0.17
95.00 ± 0.17	90.00 ± 0.19	96.00 ± 0.11	93.00 ± 0.20	96.00 ± 0.11	96.00 ± 0.11
82.00 ± 0.22	84.00 ± 0.29	88.00 ± 0.15	80.00 ± 0.12	95.00 ± 0.12	84.00 ± 0.16
84.00 ± 0.31	93.00 ± 0.16	93.00 ± 0.19	87.00 ± 0.33	94.00 ± 0.12	88.00 ± 0.21
89.00 ± 0.17	90.00 ± 0.15	93.00 ± 0.16	94.00 ± 0.12	93.00 ± 0.16	98.00 ± 0.09
93.00 ± 0.12	92.00 ± 0.16	93.00 ± 0.16	94.00 ± 0.17	91.00 ± 0.22	93.00 ± 0.20
90.00 ± 0.19	91.00 ± 0.19	88.00 ± 0.19	90.00 ± 0.19	90.00 ± 0.19	85.00 ± 0.19
82.00 ± 0.23	88.00 ± 0.19	89.00 ± 0.17	81.00 ± 0.23	93.00 ± 0.19	82.00 ± 0.26
87.00 ± 0.17	88.00 ± 0.22	94.00 ± 0.19	87.00 ± 0.17	93.00 ± 0.19	89.00 ± 0.17
86.00 ± 0.26	?	?	?	?	?
87.00 ± 0.12	96.00 ± 0.11	92.00 ± 0.19	90.00 ± 0.18	95.00 ± 0.12	88.00 ± 0.10
79.00 ± 0.32	89.00 ± 0.24	89.00 ± 0.28	76.00 ± 0.32	95.00 ± 0.12	78.00 ± 0.26
90.00 ± 0.27	86.00 ± 0.26	93.00 ± 0.16	89.00 ± 0.20	96.00 ± 0.11	88.00 ± 0.28

• The best result is 99.00% accuracy (±0.09) using DenseNet121 features with a Bagging classifier.
• A close second is 98.00% accuracy (±0.09) with DenseNet121 features (or ResNet50 features) paired with LightGBM.
• Top precision, recall, and F1-scores (99.00%) are achieved with MobileNet and InceptionV3 features when using Bagging as the classifier.
• Overall results show that deep CNN features (DenseNet121/DenseNet201/MobileNet/Xception/InceptionV3) with Bagging or XGBoost classifiers outperform several other CNN+ML combinations on the provided dataset.
• Extraction and training times indicate the approach is faster than training very deep CNNs from scratch, enabling near real-time inference for CAD systems.
• A web-based detection tool was implemented to simulate a clinical screening pipeline, though clinical validation is still required.]
• table_headers: ["DT","RF","XGBoost","AdaBoost","Bagging","LightGBM"]
• table_rows:[["83.00 ± 0.26","93.00 ± 0.23","95.00 ± 0.16","80.00 ± 0.17","96.00 ± -0.11","82.00 ± 0.28"],["92.00 ± 0.15","90.00 ± 0.21","94.00 ± 0.16","92.00 ± 0.19","99.00 ± 0.07","96.00 ± 0.11"],["84.00 ± 0.26","90.00 ± 0.24","90.00 ± 0.18","87.00 ± 0.25","96.00 ± 0.11","87.00 ± 0.17"],["95.00 ± 0.17","90.00 ± 0.19","96.00 ± 0.11","93.00 ± 0.20","96.00 ± 0.11","96.00 ± 0.11"],["82.00 ± 0.22","84.00 ± 0.29","88.00 ± 0.15","80.00 ± 0.12","95.00 ± 0.12","84.00 ± 0.16"],["84.00 ± 0.31","93.00 ± 0.16","93.00 ± 0.19","87.00 ± 0.33","94.00 ± 0.12","88.00 ± 0.21"],["89.00 ± 0.17","90.00 ± 0.15","93.00 ± 0.16","94.00 ± 0.12","93.00 ± 0.16","98.00 ± 0.09"],["93.00 ± 0.12","2 - 0.16","93.00 ± 0.16","94.00 ± 0.17","91.00 ± 0.22","93.00 ± 0.20"],["90.00 ± 0.19","91.00 ± 0.19","88.00 ± 0.19","90.00 ± 0.19","90.00 ± 0.19","85.00 ± 0.19"],["82.00 ± 0.23","88.00 ± 0.19","89.00 ± 0.17","81.00 ± 0.23","93.00 ± 0.19","82.00 ± 0.26"],["87.00 ± 0.17","88.00 ± 0.22","94.00 ± 0.19","87.00 ± 0.17","93.00 ± 0.19","89.00 ± 0.17"],["86.00 ± 0.26","?","?","?","?","?"] , ["87.00 ± 0.12","96.00 ± 0.11","92.00 ± 0.19","90.00 ± 0.18","95.00 ± 0.12","88.00 ± 0.10"],["79.00 ± 0.32","89.00 ± 0.24","89.00 ± 0.28","76.00 ± 0.32","95.00 ± 0.12","78.00 ± 0.26"],["90.00 ± 0.27","86.00 ± 0.26","93.00 ± 0.16","89.00 ± 0.20","96.00 ± 0.11","88.00 ± 0.28"]]}{
• table_headers natural language translation omitted to preserve structure: