Performance Analysis on Various Tuberculosis Detection Techniques and Remedial Suggestions Using Deep Learning

Developing deep learning systems for TB detection faces challenges including data imbalance, limited and noisy data, interpretability issues, and the need for computational resources. Generalizing models across diverse populations and integrating them into clinical workflows while addressing ethical considerations further complicates their implementation and adoption in healthcare settings. This study presents an integrated approach for Tuberculosis (TB) detection, remedial suggestions, and air quality assessment. Leveraging machine learning techniques, the system employs Random Forest to gauge lung quality and ResNet-50 for TB detection via X-ray imaging analysis. The Random Forest algorithm assesses lung health by analyzing various parameters, aiding in early identification of abnormalities. Simultaneously, ResNet-50, a deep learning model, accurately detects TB manifestations in X-ray images, enabling swift diagnosis. Furthermore, this research investigates the correlation between air quality and TB prevalence, emphasizing environmental factors' impact on respiratory health. The integrated system holds promise in early TB detection, providing remedial suggestions based on severity levels, and highlights the critical role of air quality in respiratory health. Its multifaceted approach offers a comprehensive framework for mitigating TB risks and improving public health strategies.