AI reshaping oncology and cardiovascular health 

October 05, 2023

Deep-learning, detection of informative patterns from data, and artificial intelligence (AI) have been integrated into the classical analysis/diagnostic workflow to improve the accuracy of screening and detection of cancer and cardiovascular disease using the availability of large datasets. Historically radiology services and, in some instances, pathology is where machine learning, deep learning algorithms, and AI models have been used in the fields of cancer and cardiovascular health. Now, AI is revolutionizing direct patient care. Below we highlight two specialties where diagnoses are becoming laser-sharp and treatments tailor-made with the help of AI.

AI has changed the way patients with cancer are diagnosed, monitored, and treated, by detecting abnormalities in imaging that may otherwise have gone undetected by humans. Here are a few things on the horizon for deep learning and AI.  Johns Hopkins Kimmel Cancer Center is further developing a deep learning-based method for identifying tumors in patients with cancer and training a machine learning algorithm (an open-source program called DeepTCR) to predict patient response to treatment. Researchers are taking this a step further to diagnose and predict cancer through a concept called biology-guided deep learning, predicting treatment outcomes and identifying underlying tumor biology. Classifying tumor subtypes is key to cancer treatment, as the response to treatment can be vastly different based on the type of tumor.

AI and traditional clinical care are being used together to better care for patients at the Mayo clinic by combining human expertise and learnings from big data. Use cases include cutting time to diagnosis and limiting brain damage for patients suffering from strokes, and detection of a weak heart pump. Researchers are also developing a way to identify genes within a patient’s DNA that have a relationship with cardiovascular disease along with demographics such as age, ethnicity, and gender to assist with detecting the risk of heart disease before a serious event.

AI offers seemingly endless possibilities for transforming diagnosis, treatment, and patient care. Its remarkable ability to process complex data has led to earlier detection, personalized therapies, and more informed clinical decisions. However, as with any application of AI, its potential is caveated with concerns about data privacy, potential biases, and the need for human oversight – all of which underscore the importance of a balanced approach.