A Multimodal Deep Learning Model Integrating Cnn And Transformer For Predicting Chemotherapy Induced Cardiotoxicity

Project Code :TCPGPY1965

Objective

The main objective of this project is to design and implement a multimodal deep learning framework that can accurately predict chemotherapy-induced cardiotoxicity by analyzing Temporal Dynamic Imaging data. Specifically, the project aims to extract spatial features using Convolutional Neural Networks (CNN), model sequential dependencies using LSTM, GRU, and Transformer-based architectures, and evaluate the classification performance of various model combinations, including CNN + LSTM, CNN + GRU, and CNN + Transformer. The objective also includes identifying the best-performing architecture based on accuracy, F1-score, and other relevant metrics, with the intent to support clinical decision-making and early intervention strategies.

Abstract

This project focuses on the automated evaluation of subjective answers by utilizing a pretrained language model such as Gemini AI. The system takes a question, its correct answer, and a student’s answer as input. It then compares the student’s response with the correct answer using semantic understanding to generate a score. To ensure secure handling of user data and answers, the system uses cloud-based infrastructure and algorithms. This approach reduces manual grading efforts, improves scoring consistency. It is designed with modular components for user management, answer input, AI-based evaluation, and result display. By integrating machine learning with secure system design, the project offers a scalable and efficient method for evaluating descriptive answers in academic environments.

Keywords: subjective answer evaluation, machine learning, Gemini AI, semantic analysis, student assessment, score prediction, cloud computing, automated grading, NLP, data security

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.