Research breakthrough possible @S-Logix

Office Address

  • 2nd Floor, #7a, High School Road, Secretariat Colony Ambattur, Chennai-600053 (Landmark: SRM School) Tamil Nadu, India
  • +91- 81240 01111

Social List

Research Topics in Cancer Detection and Diagnosis using Deep Learning

Cancer Detection and Diagnosis using Deep Learning

PhD Research and Thesis Topics in Cancer Detection and Diagnosis using Deep Learning

Cancer is one of the leading causes of death worldwide, and its early detection is the main focus of the scientific and medical research community. The manual analysis of medical images requires high time consumption and is error-prone. For this reason, computer-aided diagnosis systems are introduced to help doctors enhance the efficacy of medical image interpretation for cancer disease diagnosis.

Owing to the weakness of machine learning in feature extraction, deep learning models are applied to diagnose cancer with the help of image processing techniques. As a sub-discipline of artificial intelligence, deep learning models are utilized to analyze the complex biology of cancer. Deep learning models own the ability to analyze medical images with high-level feature representation.

Impressive deep learning architectures utilized for a variety of cancer diagnosis applications are Convolutional Neural Networks, Multi-Scale Convolutional Neural Networks (M-CNN), Fully Convolutional Networks (FCNs), U-Net Fully Convolutional Neural Networks, Generative Adversarial Networks (GANs), Recurrent Neural Networks (RNNs), Long Short-Term Memory (LTSM), Restricted Boltzmann Machine (RBM), Auto-encoders (AEs), Sparse Auto-encoders (SAE), Convolutional Auto-encoders (CAE), Deep Belief Networks (DBN) and Adaptive Fuzzy Inference Neural Network (AFINN). Significant types of cancer detection using deep learning are highlighted below;

Breast Cancer - Various studies have been conducted on breast cancer detection using deep learning, convolutional neural networks, and auto-encoders. Their variants are applied to distinguish cystic, benign, or malignant labels of breast cancer.

Lung Cancer - Deep learning models such as pretrained CNN, deep convolutional neural networks, multi-view CNN, multi-variant Convolutional neural network (Mc-CNN), and fine-tuned CNN are exploited for lung cancer detection.

Brain Cancer - In Brain cancer detection, segmentation of the healthy part of the brain is challenging. Deep learning models help to automate medical image segmentation, and recently fully connected convolutional neural networks, 2D CNN and 3D CNN have been employed.

Skin Cancer - In the early stages, skin cancer detection using deep learning models differentiates benign and malignant melanoma. Recently employed deep learning models for skin cancer analysis are the multi-task convolutional neural network, U-net CNN, fully convolutional residual network (FCRN), self-advised semi-supervised learning, transfer learning, and pre-trained CNN

Prostate Cancer - Accurate segmentation is essential in prostate cancer detection. Fully convolutional networks, Patch-based CNN, and Sparse auto-encoder are exploited for prostate cancer detection from the medical image, especially MRI.