Effective Lung Diseases Detection using Contrast Enhancement and Efficiennet

lung diseases including Covid-19 become major factor due to deadly human in the earth. Late early detection was influencing the fatal impact for patient. Majority of medical employed adopted image material to detect lung diseases. However, medical imaging experts are very limited. Image diagnosis operated diagnosis with traditional human vision. This is very hard to classify the classification of lung diseases. According to the problem above, artificial intelligent (AI) has been implemented to detect lung diseases effective. However, the result of lung disease detection faces the mis detection in high number. This problem emerges the serious problem for patients. Aim to handle of high error detection lung diseases classification, we employed several image pre-processing and adoption of deep learning model in the term of Efficienet. The pre-processing model including augmentation, white balance enhancement and contrast enhancement. According to previous research, majority of medical imaging process face with low quality of image. According to experiment report, our proposed model achieved significantly in reducing mistake detection on lung diseases classification. Where accuracy result was 0,97, precision was 0,96, recall was 0,98, F1-score was 0,97. We consider adopting our proposed model in multi-class classification. We evaluated our proposed model using evaluation metric and AUC Curve.