Visual Question Answering using Data Mining Techniques for Skeletal Scintigraphy in medical domain - VQADMSS

Understanding about the medical images of patients is a very tedious task. Doctors should convey their patient through the image of the questions asked by the patient. Large amounts of labeled data are required for training in traditional approaches for VQA (Visual Question Answering). Also, the description of clinic trial text in English and in multilingual contexts is one of the challenges in the medical field. To present the clarification about the images, doctors are required to provide the related images. It is better for comparison with the patient's previous report and current report. This paper contributes to solve the problems related to VQA for better description of the image and accuracy related images through the answer of the questions, also to make it easy to convey the users with any kind of images. Question answer process should be more descriptive for easy to understand and traceable. This system helps to identify the types of images which are captured by any scanner. The better accuracy is a visualization method which projects the answers as a baseline that shows the corresponding region with various colors, which is easier to note the answers in visual method for the appropriate questions. This proposed framework focuses on Radiology image for Skeletal Scintigraphy to transform and generate a model using Data Mining Techniques. This system suggests that the effective medical Visual Question answering techniques is better to assist doctors in clinical analysis and diagnosis. This also will help the hospital services to grow the medical domain.