A Convolutional Neural Network for COVID-19 Diagnosis: An Analysis of Coronavirus Infections through Chest X-rays
Bibliographic record
Abstract
Coronavirus (COVID-19) disease has not only become a pandemic but also an overwhelming strain on the healthcare industry. The conventional diagnostic methods include Antigen Rapid Kits and Reverse Transcription–Polymerase Chain Reaction (RT-PCR) tests. However, they entail several drawbacks such as low precision in diagnosis, increased time in obtaining test results, increased human–patient interaction, and high inaccuracy in the diagnosis of asymptomatic individuals, thus posing a significant challenge in today’s medical practice in curbing an extremely infectious disease such as COVID-19. To overcome these shortcomings, a machine learning (ML) approach was proposed to aid clinicians in more accurate and precise infection diagnoses. A Convolutional Neural Network was built using a sample size of 1920 chest X-rays (CXR) of healthy individuals and COVID-19-infected patients. The developed CNN’s performance was further cross-checked using the clinical results of the validation dataset comprising 300 CXRs. By converting the final output to binary, an intuitive classification of whether a specific CXR is of a healthy or a COVID-infected patient was accomplished. The statistical analysis of the CNN was: Accuracy: 95%; Precision: 96%; Specificity: 95%; Recall: 95%, and F1 score: 95%, thus, proving it to be a promising diagnostic tool in comparison to the other existing ML-based models. The datasets were obtained from Kaggle, GitHub, and European Institute for Biomedical Imaging Research repositories. The prospects of the proposed CNN lie in its flexibility to be altered and extrapolated in diagnosing other lung infections, such as pneumonia and bacterial infections, with relevant training algorithms and inputs. Additionally, the usage of other bio-imaging modalities as input datasets such as CT scans, Lung Ultrasounds and Heat Maps gives the CNN immense potential to assess for better insights on the severity of infection in both infected and asymptomatic patients as well as other related medical diagnoses.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".