Wings & Warnings: Advanced Illness Prediction in Birds Through CNN and Random Forest Integration

In this study, the authors provide a hybrid diagnosis model developed for the accurate diagnosis of avian diseases. To take advantage of the strengths of deep learning and ensemble approaches, their model consists of two stages of convolutional neural networks, each followed by a max-pooling layer, and a final classifier based on a random forest. When tested to differentiate between different bird ailments, the model regularly outperformed human experts. In the case of Avian Influenza, in particular, we found an astounding 99% accuracy, recall, and F1-score (precision + recall + F1-score = 100%). Avian botulism, psittacosis, and aspergillosis all showed about 99% accuracy, which is on par with the other two diseases. The detection rates for Newcastle disease or mycoplasma infections were both recorded at 98% accuracy. The aggregate results showed an estimated F1-score, recall, and precision for all disorders of 95.86%. Similarities between the weighted, micro, and macro averages further support the model's validity. The convolutional layers & random forest-based technique have shown impressive performance, outlining its potential to revolutionize avian healthcare diagnostics by delivering a fast and accurate solution for identifying diseases in birds.