DETECTION OF BRAIN TUMOR USING VGG16 AS AUTOENCODER WITH BI-LONG SHORT TERM MEMORY METHOD

The fast proliferation of aberrant brain cells that distinguishes a Brain Tumor (BT) poses a significant health concern to adults because they can cause serious organ dysfunction and possibility of causing death. These tumors have a wide range in location, size and texture. When at
tempting to find malignant tumors, Magnetic Resonance Imaging (MRI) is an essential technique. However, BT detection is manually complex and task with more time consumption that may result in mistakes. The aesthetically pleasing appearance of MRI scans is improved by enhancing
image technologies that employ various filters to the raw images. The paper focus on overcoming the existing gaps by introducing the integration of synergism among Bi-Long Short Term Memory (BiLSTM) as well as Convolutional Autoencoder (CAE) using VGG16 layer to optimize its united impact over predictive performance. The research’s objectives include en
hancing the conceptual integration of BiLSTM and CAE with VGG16 as the hyperparameter tuning to improve the model’s efficiency for capturing temporal and spatial nterdependence among heterogeneous datasets. Furthermore, this research concentrated on improving the prediction method
output interpretability for assuring their practical application in clinical environments. The CAE has trained from the source dataset and perform in optimizing during testing using a test subject for effective computation. Moreover, the BiLSTM is utilized as RNN model with CAE VGG16 for providing improved detection of BT in healthcare industries. Hence, the proposed CAE with BiLSTM is compared to traditional AE and CAE with LSTM for evaluating BT detection using MRI dataset with various BT classes.