An Efficient Breast Cancer Disease Prediction Method using Deep Learning
Deepti Sharma, Rajneesh Kumar, Anurag Jain · 2023
Breast Cancer is considerable concern among health issues in women. It is one of the higher mortality factors when compared to other cancers. Early prediction of cancer patients helps doctors to give a better diagnosis and prognosis of the disease. The study's goal is to assess how well proposed CNN-based models perform at classifying patients' tumour types as either malignant or benign depending on whether they are cancerous or non-cancerous. The goal of the study is to identify the key factors, such as the number of convolutional layers, the calibre of the training data, and the dependent variable, that influence the model's performance during training. The study makes use of the Kaggle-available Breast Cancer Histopathological data set. It is commonly used to assess CNN-based models in the health care sector. The classification performances of the models are analysed, and the models' training efficacy is assessed. The work demonstrates that robust feature representation and precise patient predictions can be achieved by utilising deep learning techniques, notably CNN models. The metrics for estimation performance were 97.42% accuracy, 97.39%Precision, and 97.45%Recall, respectively. The study's findings support the idea that using deep learning techniques can help doctors make accurate diagnoses, choose the best course of therapy, and monitor patients' prognoses. In comparison to conventional procedures, it substantially offers clinicians a solution. According to the study, managing and interpreting healthcare data may be done much more effectively when machine learning and deep learning techniques are used.