AI-Enhanced Lung Cancer Prediction: A Hybrid Model's Precision Triumph.

Area of Science:

• Oncology
• Artificial Intelligence
• Medical Informatics

Background:

• Lung cancer is a leading global cause of cancer mortality, necessitating early detection for effective treatment.
• Early detection of lung cancer remains a significant challenge, driving the need for innovative diagnostic approaches.
• Artificial intelligence (AI) presents transformative potential in cancer prediction and healthcare.

Purpose of the Study:

• To develop and evaluate a novel hybrid deep learning model for detecting lung cancer from electronic health records.
• To assess the model's performance against established benchmarks and state-of-the-art methods.

Main Methods:

• A hybrid deep learning architecture combining Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory Networks (BiLSTM) was developed.
• The model was trained and validated using patient medical notes from the MIMIC IV dataset.
• Performance was evaluated using metrics such as accuracy and Matthews Correlation Coefficient (MCC), with comparisons to LSTM and BioBERT models.

Main Results:

• The hybrid CNN-BiLSTM model achieved a high accuracy of 98.1% and an MCC of 96.2% on the MIMIC IV dataset.
• The proposed model demonstrated superior performance compared to standalone LSTM (MCC 93.5%, Accuracy 97.0%) and BioBERT (MCC 95.5%, Accuracy 98.0%).
• Further validation on the Yelp Review Polarity dataset confirmed the model's significant outperformance of other compared methods.

Conclusions:

• The developed hybrid deep learning model shows exceptional efficacy in early lung cancer detection from medical notes.
• This advancement represents a significant step towards enhancing precision medicine through AI-driven diagnostic tools.
• Continuous refinement of AI models is crucial for improving early cancer detection and patient outcomes.