Prevention is key in detecting cancer early on.
“Cancer is among the leading causes of death worldwide,” . In 2018 alone, 9.5 million cancer-related deaths were recorded worldwide . Prevention of cancer has been continuously researched to better the health of our current and future generations. With medical technology advancing to improve the quality of health for many, artificial intelligence (AI) for predicting cancer has been widely studied. Studies have shown that using AI for early detection improves patient survival rates and may limit the need for extensive treatment. Some of the most common cancers are lung cancer, colon cancer, and breast cancer .
AI is detecting skin lesions, mammographic lesions, and colon cancer with accuracy levels comparable to dermatologists, radiologists, and pathologists . According to author Elizabeth Svoboda in her 2017 article “Artificial intelligence is improving the detection of lung cancer,” 70% of lung cancers are detected during the latest stages, making a 5-year survival rate uncommon . Getting routine screening can reduce the mortality rate by 20-30% . A team of researchers found that 3D computed tomography (CT) scanning had a 95% accuracy rate detecting cancerous tumors . Meanwhile, radiologists only have a 65% accuracy rate . Radiologists and AI systems can work together, however, there are issues to confront. An AI system’s analysis on why it scans an image as “benign” or “malignant” can be too complex for a human to understand . Outside of these issues, using CT scanning to check for lung cancers can help the survival rate of lung cancer patients in the future.
According to Wang et al. in the 2020 article “Application of artificial intelligence to the diagnosis and therapy of colorectal cancer,” the use of convolutional neural network (CNN) was effective in screening colonoscopy polyps with a specificity of 74.8%, sensitivity of 99.3%, and accuracy of 97.7% . The development of CNN for “autonomous detection and localization of colon polyps in colon capsule endoscopy” had accuracy levels of 96.4%, sensitivity of 97.1%, and specificity of 93.3% . Having the ability to achieve rates such as this makes a great method to identify cancers when medical professionals are performing coloscopies to examine the health of an individual’s colon and diagnosis any cancerous tumors.
In 2020, the National Breast Cancer Foundation found that more than 276,000 new cases of invasive breast cancer and more than 48,000 non-invasive cases were diagnosed that year in the United States . 64% of breast cancer cases are diagnosed in the early stages of development . In various approaches to examine cancerous tumors, convolutional neural network (CNN) is used. Supported convolutional neural network (SCNN) allows mammogram or breast images to be seen by utilizing adaptive median filter (AMF) to reduce noise, reduce distortion, and improve clarity . The 2020 study by Alfifi et al. titled “Enhanced Artificial Intelligence System for Diagnosing and Predicting Breast Cancer using Deep Learning” found that the SCNN had a 95% accuracy, and traditional convolutional neural network (TCNN) had a 92% accuracy in detecting breast cancers . Early diagnosis gave patients a 99% chance of survival . Using the most advanced AI to examine breast cancer images can decrease the number of future invasive cases.