Interpretable Artificial Intelligence for COVID-19 Diagnosis from Chest CT Reveals Specificity of Ground-Glass Opacities

Anmol Warman, Pranav Warman, Ayushman Sharma, Puja B. Parikh, Roshan Warman, Narayan Viswanadhan, Lu Chen, Subhra Mohapatra, Shyam S. Mohapatra, Guillermo SapiromedRxiv 2020 · DOI

Abstract

Background: The use of CT imaging enhanced by artificial intelligence to effectively diagnose COVID-19, instead of or in addition to reverse transcription-polymerase chain reaction (RT-PCR), can improve widespread COVID-19 detection and resource allocation. Methods: 904 axial lung window CT slices from 338 patients in 17 countries were collected and labeled. The data included 606 images from COVID-19 positive patients, 224 images of other pulmonary diseases, and 74 images of normal patients. We developed an object detection model to identify ground-glass opacities (GGOs) for COVID-19, GGOs for non-COVID-19 diseases, and features inconsistent with COVID-19. These features were processed by an interpretable decision tree model. Results: On an independent test of 219 images, the model achieved 96.80% accuracy and 98.33% sensitivity. High performance was also maintained for asymptomatic patients (97.06% accuracy). Conclusion: We present an interpretable AI tool that differentiates COVID-19 GGOs from similar pathologies, suggesting that GGOs can be disease-specific.

Introduction

Early in the COVID-19 pandemic, the gold standard for diagnosis was the RT-PCR test. However, RT-PCR faced significant hurdles, including limited availability, slow turnaround times, and a false-negative rate ranging from 10% to 30%. While Computed Tomography (CT) scans offered a more sensitive alternative, they presented a different challenge: the radiological presentation of COVID-19 often overlaps significantly with other viral pneumonias.

Furthermore, most AI tools developed to bridge this gap have been "black boxes." While these models might output a diagnosis, they fail to explain the underlying reasoning. For clinicians, a diagnosis without visual evidence or a clear logic path is difficult to trust. This research addresses two major hurdles: improving the specificity of CT-based diagnosis and making the AI’s decision-making process transparent and interpretable.

Methodology

The researchers developed a two-stage "transparent" pipeline that combines object detection with a logical decision framework:

  1. Object Detection (YOLOv3): Instead of analyzing the image as a whole, the model uses the YOLOv3 (You Only Look Once) architecture to identify and localize specific radiological features. The model was trained to recognize three distinct categories:
  2. COVID-19 GGOs: Ground-glass opacities specifically associated with COVID-19.
  3. Non-COVID-19 GGOs: Opacities associated with other lung diseases.
  4. Disqualifying (DQ) Features: Radiological signs inconsistent with COVID-19, such as cavitations or pleural effusions.
  5. Interpretable Decision Tree: The features detected by YOLOv3 are fed into a decision tree. This model follows a human-verifiable logic path—for example, if COVID-specific GGOs are present and disqualifying features are absent, the model suggests a positive COVID-19 diagnosis.

To optimize performance with a specialized medical dataset, the team utilized transfer learning, initializing the model with weights from the COCO dataset.

Data and Validation

The study utilized a diverse, global dataset to ensure the model could generalize across different populations and hospital protocols. Data consisted of 904 axial lung window CT slices from 338 patients across 17 countries. The composition included 606 images from RT-PCR-confirmed COVID-19 patients, 224 images representing various other pulmonary diseases, and 74 images from healthy patients.

To ensure high-quality training data, regions were labeled by two upper-level radiology residents and confirmed by a board-certified radiologist with 10 years of experience. This rigorous labeling process allowed the model to learn the subtle phenotypical nuances of different lung pathologies.

Results and Discussion

The model demonstrated exceptional performance across several metrics on an independent test set of 219 images:

  • Accuracy: 96.80% (95% CI, 96.75 to 96.86)
  • Sensitivity: 98.33% (95% CI, 98.29 to 98.40)
  • Specificity: 94.95% (95% CI, 94.84 to 95.05)
  • AUC-ROC: 0.9664

Notably, the model maintained a 97.06% accuracy when tested on 34 images from asymptomatic patients. This suggests that AI-enhanced CT could be a powerful tool for identifying "silent" spreaders who show no clinical symptoms but have incidental CT findings. Additionally, the system is highly efficient, processing an image in less than 20 milliseconds on a standard Tesla P100 GPU.

The Specificity of GGOs

A key insight from this study is that ground-glass opacities, conventionally viewed as non-specific markers of inflammation, may actually be disease-specific. The AI's ability to differentiate between COVID-19 GGOs and non-COVID-19 GGOs suggests there are subtle physical characteristics in these opacities that the human eye may struggle to categorize consistently, but which a trained model can identify.

Conclusion and Future Work

This interpretable AI tool provides a "second pair of eyes" for radiologists, offering both a diagnosis and the visual evidence required to support it. By generating tight bounding boxes around relevant features, the model moves away from the "black box" paradigm toward a collaborative clinical tool.

Future work will focus on expanding the dataset to include rarer pulmonary conditions and exploring multimodal integration. Combining AI-driven CT analysis with Electronic Health Records (EHR) could further refine diagnostic accuracy and provide a more robust paradigm for managing respiratory pandemics.

Summary updated

Abstract

Background The use of CT imaging enhanced by artificial intelligence to effectively diagnose COVID-19, instead of or in addition to reverse transcription-polymerase chain reaction (RT-PCR), can improve widespread COVID-19 detection and resource allocation. Methods 904 axial lung window CT slices from 338 patients in 17 countries were collected and labeled. The data included 606 images from COVID-19 positive patients (confirmed via RT-PCR), 224 images of a variety of other pulmonary diseases including viral pneumonias, and 74 images of normal patients. We developed, trained, validated, and tested an object detection model which detects features in three categories: ground-glass opacities (GGOs) for COVID-19, GGOs for non-COVID-19 diseases, and features that are inconsistent with a COVID-19 diagnosis. These collected features are passed into an interpretable decision tree model to make a suggested diagnosis. Results On an independent test of 219 images from COVID-19 positive, a variety of pneumonia, and healthy patients, the model predicted COVID-19 diagnoses with an accuracy of 96.80 % (95% confidence interval [CI], 96.75 to 96.86) , AUC-ROC of 0.9664 (95% CI, 0.9659 to 0.9671) , sensitivity of 98.33% (95% CI, 98.29 to 98.40) , precision of 95.93% (95% CI, 95.83 to 95.99), and specificity of 94.95% (95% CI, 94.84 to 95.05). On an independent test of 34 images from asymptomatic COVID-19 positive patients, our model achieved an accuracy of 97.06% (95% CI, 96.81 to 97.06) and a sensitivity of 96.97% (95% CI, 96.71 to 96.97). Similarly high performance was also obtained for out-of-sample countries, and no significant performance difference was obtained between genders. Conclusion We present an interpretable artificial intelligence CT analysis tool to diagnose COVID-19 in both symptomatic and asymptomatic patients. Further, our model is able to differentiate COVID-19 GGOs from similar pathologies suggesting that GGOs can be disease-specific.
BibTeX
@misc{warman2020interpretable,
  author = {Anmol Warman and Pranav Warman and Ayushman Sharma and Puja B. Parikh and Roshan Warman and Narayan Viswanadhan and Lu Chen and Subhra Mohapatra and Shyam S. Mohapatra and Guillermo Sapiro},
  title = {{Interpretable Artificial Intelligence for COVID-19 Diagnosis from Chest CT Reveals Specificity of Ground-Glass Opacities}},
  year = {2020},
  doi = {10.1101/2020.05.16.20103408},
  howpublished = {medRxiv},
}
Figure