Esophageal Cancer Classification in Initial Stages Using Deep and Transfer Learning

The sixth most frequent cancer and one with the highest death rate is esophageal cancer (EC). Initial diagnosis might increase the survival level of people with esophageal abnormalities. Esophagitis to non-dysplasia is the progression of EC in that order. Convolutional neural networks (CNN) equipped with effective feature extractors improve the prognosis
$\text{of Barrett's}$
esophagus, esophagitis as well as precancerous stage conditions. More relevant characteristics may be extracted for the mechanized Barrett's esophagus classification as well as esophagitis using transfer learning methods with CNN. This research reports on a study on the employment of Enhanced Convolution Neural Networks (ECNN) for the classification of esophagitis. The evaluation
$\text{will}$
be done with the K vasir open-access dataset, which delivers considered esophagitis along with Barrett's esophagus images. Several performance metrics, including Accuracy (acc), precision (prec), Recall (rec), and flscore, are used to assess the outcomes. To demonstrate the effectiveness of the proposed strategy, comparison studies are conducted with numerous different existing methodologies.