IEEE BASE PAPER TITLE:
Breast Cancer Classification using CNN with Transfer Learning Models
OUR PROPOSED PROJECT TITLE:
Advancing Breast Cancer Diagnosis: Deep Learning for Accurate Tumor Classification using DenseNet201
IEEE BASE PAPER ABSTRACT:
Breast cancer is the deadliest and most common cancer in the world. Early treatment of this cancer can help to nip it in the bud. In present medical setting, this cancer is identified by manual clinical procedures, which can lead to human errors and further delay the treatment procedure. So, we propose a Convolutional Neural Network (CNN) model employed with transfer learning approach with RESNET50, VGG19 and InceptionV3 algorithms. The histopathological image dataset is used to detect cancer cells in the tissues of the breast. We examine the performance of different models based on their accuracy, by varying different optimizers (Adam, SGDM and RMSProp) for each transfer learning model. The results show that the Inception-V3 model with Adam optimizer outperforms VGG19 and RESNET-50 in terms of accuracy.
PROJECT OUTPUT VIDEO:
ALGORITHM / MODEL USED:
OUR PROPOSED ABSTRACT:
Breast Cancer is a major health concern worldwide, affecting millions of women each year. Early and accurate detection of breast cancer plays a crucial role in improving patient outcomes and survival rates. In this final year project titled “Breast Cancer Detection using Deep Learning,” we propose a novel approach to automatically diagnose breast tumors as benign or malignant by utilizing the power of DenseNet201, a powerful deep learning architecture. The primary objective of this project is to develop a robust and efficient deep learning model capable of accurately classifying breast tumor images based on microscopic biopsy data. For this purpose, we have leveraged the Breast Cancer Histopathological 400X (BreakHis 400X) dataset, sourced from Kaggle. This dataset consists of 1,693 microscopic biopsy images of both benign and malignant breast tumors, providing a diverse and challenging set of samples for training and evaluation. The DenseNet201 architecture was chosen due to its ability to efficiently capture intricate features and patterns within images. By leveraging transfer learning, we fine-tuned the pre-trained DenseNet201 model on the BreakHis dataset to achieve optimal performance for breast cancer classification. During the training phase, we observed remarkable results with a training accuracy of 96.00% and a validation accuracy of 89.00%. These outcomes demonstrate the model’s ability to learn complex representations from the data and effectively generalize to unseen samples, making it a promising tool for breast cancer diagnosis. The proposed deep learning-based breast cancer detection system offers several key advantages, including automation, reproducibility, and scalability. If deployed in clinical settings, this model could serve as an invaluable assistant to medical practitioners, aiding them in making accurate and timely decisions in diagnosing breast cancer. Furthermore, this project’s success opens avenues for future research and improvements in medical image analysis using deep learning. By incorporating larger datasets, fine-tuning hyperparameters, and exploring other state-of-the-art architectures, we believe even higher accuracy levels can be achieved, ultimately contributing to advancements in breast cancer diagnosis and treatment. The potential impact of this work in the field of medical image analysis and breast cancer diagnosis cannot be understated, offering a ray of hope towards improved healthcare outcomes for breast cancer patients worldwide.
- System : Pentium i3 Processor.
- Hard Disk : 500 GB.
- Monitor : 15’’ LED
- Input Devices : Keyboard, Mouse
- Ram : 6 GB
- Operating system : Windows 10 Pro.
- Coding Language : Python 3.10.9
- Web Framework : Flask
Ch.Rajendra Prasad; Banothu Arun; Soma Amulya; Preethi Abboju; Sreedhar Kollem; Srikanth Yalabaka, “Breast Cancer Classification using CNN with Transfer Learning Models”, 2023 International Conference for Advancement in Technology (ICONAT), IEEE Conference, 2023.