AI and Machine Learning (ML) are being incorporated into many fields, but one area that they are literally able to assist in life-saving practices is the field of healthcare. AI and ML are currently being used in some of the most cutting-edge cancer studies to help with diagnostics and early detection.
ML Algorithm Determines Short-Term Mortality in Cancer Patients
A recent study demonstrated that an ML algorithm integrated into an electronic health record can accurately predict short-term mortality in cancer patients and may be used to help physicians and patients make decisions regarding behavioral interventions, goals of care, and end-of-life preferences. The analysis included 24,582 patients aged 18 years and above with outpatient oncology or hematology.
AI-Powered Cancer Diagnostics Solution
Another example of AI in the field of cancer diagnostics is the Galen Platform, a solution created by the Israeli company, Ibex Medical Analytics, and its partner, LDPath, a UK-based pathology services provider. The platform is the first AI-powered cancer diagnostics solution and can be used to detect breast and prostate cancer from biopsies. It’s built on deep learning algorithms that have been created by a team of pathologists, software engineers, and data scientists.
The solution comes at a time when there is a shortage of pathologists in the UK, which has led to a delay in cancer diagnoses. The current waiting time for a cancer diagnosis is 6 weeks, a crucial amount of time when it comes to treatment. The pressure on pathologists can also lead to wrong diagnoses. With an AI platform, the margin of error is significantly reduced, and the stress placed on pathologists can be lightened.
ML in Diagnostic Training to Detect Breast Cancer
In Australia, a new online platform was launched in November 2020. Called VIETRAD, the platform uses machine learning technology to test, monitor, and improve the radiologic diagnosis of breast cancer. The platform uses high-quality mammograms from Vietnam and Australia and then allows radiologists to practice reading images. The goal is to improve their skills and reduce the margin of human error. ML is used to identify errors in diagnosis and target the areas in which radiologists need more training. The approach, used by the University of Sydney, is credited with improving radiologists’ efficacy in detecting breast cancer from below 50% to 85%.
New research from the University of Sheffield in England shows that AI can help doctors better predict the risk of patients for developing oral cancer. Currently, physicians predict the likelihood of a patient’s pre-cancerous changes developing into oral cancer by analyzing a biopsy based on 15 different criteria. The criteria are assessed to create a score, and the score helps determine the course of treatment. The main challenge is that the score is subjective and can differ according to the pathologist. By using AI and ML, the subjectivity of determining the score is removed. This will result in scores that are more accurate, consistent, and objective, and it can revolutionize the diagnosis of oral cancer and management.
Saving Lives Through Early Detection
Using AI and ML to help in cancer diagnostics is a relatively young field, but the impact of this new technology is already apparent. New developments are happening every day, and they have the potential to change cancer mortality rates. The earlier cancer is detected, the easier it is to treat and the lower the mortality rate is likely to be. Using AI and ML to help in cancer diagnostics can save lives through early detection.