A ROBUST CT SCAN APPLICATION FOR PRIOR STAGE LIVER DISORDER PREDICTION WITH GOOGLENET DEEPLEARNING TECHNIQUE

Recent technologies mainly concentrate on medical applications based on image processing tools. The medical
image processing has recognized the different diseases with fast diagnoses, such as lung, heart, brain tumour and liver. The earlier stage of disease diagnosis helps to identify appropriate disease treatment. In this investigation, CT scan-based liver disease or disorders have been predicted and classified based on the Google Net CNN (Convolutional neural networks) deep learning algorithm. At the initial stage, the local threshold (LT) segmentation model and at the classification stage
Improved GoogleNet CNN deep learning model applied on selected real CT liver images. This work mainly focuses on
liver disorders prediction and disease identification using CT liver medical images. The proposed LT-GoogleNet CNN
deep learning model diagnosis the liver diseases with real and accurate manner. Here we used the two different algorithms
to identify the black and white pixels on the given data set CT images to remove noise on the practical image to get proper
and good accurate results. The performance measures such as precession, accuracy, PSNR, CC and time of diagnosis has
been improved. At final implemented LT-GoogleNet CNN deep learning model compared with existed methods, conclude
that this mechanism is efficient. After doing the practical values we got using the mentioned proposed method Google Net
CNN prediction probability is good accuracy as 98 and precession 98.6, Recall 98.3, F1 score 98.4, PSNR 59.8, CC
99.83predection of liver disease is verified using the different database ANDI-1 and ANDI-2