• Srinivasa Raghavan S
• Muthukumaran S
• Manju Bala P

To diagnose chest radiographs, doctors and radiologists still employ manual and visual methods. This necessitates the development of an independent, intelligent system that can interpret X-rays of the chest. A deep neural network is used in this thesis will be used dubbed a region-based convolutional neural network to identify X-ray of the chest picture as normal or abnormal. In order to train and evaluate the region-based convolutional neural network, We used X-rays of the chest radiography from public databases, including both healthy and unhealthy radiographs. When comparing the accuracy of two networks, the first one makes use of raw X-ray of the chest data as input, while the second employs enhanced and treated images using histogram equalisation. The Region based Convolutional Neural Network demonstrated a reasonable degree of generalisation when identifying whether unseen X-rays of the chest were normal or abnormal, according to experimental data. Furthermore, the use of histogram equalisation to improve images has been shown to assist systems owing to improved accuracy that results from enhanced pictures, learning and performance

Deep neural network, Convolutional Neural Network with Regions, radiography, classification, generalization

"R-CNN based Diagnosis of Tuberculosis from Chest X-ray Images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 5, page no.c187-c232, May-2023, Available :http://www.jetir.org/papers/JETIRTHE2034.pdf

"R-CNN based Diagnosis of Tuberculosis from Chest X-ray Images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 5, page no. ppc187-c232, May-2023, Available at : http://www.jetir.org/papers/JETIRTHE2034.pdf