Lung Cancer Identification with Deep Networks: Convolutional Neural Network or Transformers!
Bibliographic record
Abstract
Lung cancer is one of the most common and deadly cancers worldwide, with low survival rates due to delayed identification. Timely and accurate diagnosis can improve treatment outcomes and reduce mortality. Histopathological imaging remains the gold standard for diagnosis, but manual inspection is laborious and subjective. In this context, deep learning approaches can automate classification and support pathologists in making reliable decisions. This research presents a deep learning-based multi-class lung cancer classification framework using histopathological images. The dataset consists of three categories: benign, adenocarcinoma, and squamous cell carcinoma. Multiple CNN architectures, including MobileNetV2, ResNet-101, and EfficientNet-B3, were evaluated. Among them, MobileNetV2 achieved the highest classification accuracy of 99.1%, while EfficientNet-B3 achieved 97.2% with a strong balance across precision, recall, F1-score, and ROC-AUC. To address class imbalance and overfitting, transfer learning, class weighting, and data augmentation were applied. Confusion matrix analysis further confirmed the ability of the models to distinguish between similar cancer subtypes. The findings highlight the potential of EfficientNet-based CNNs for reliable medical image analysis and digital pathology. Future work will explore Vision Transformers and attention mechanisms to enhance diagnostic accuracy and interpretability. Overall, this research demonstrates the promise of AI-supported histological evaluation and its role in advancing clinical decision support.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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 source (direct Gemma or distilled Codex), 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".