Detecting Ovarian Cancer Using AI Nanotech Sensors
[Please note that this page contains affiliate links. If you choose to purchase after clicking a link, I may receive a commission at no extra cost to you.]
A study published in the journal Nature Biomedical Engineering reveals how artificial intelligence (AI) machine learning and nanotechnology can find ovarian cancer signals in blood.
Cancer is the Leading Cause of Death
The World Health Organization (WHO) website lists cancer as the leading cause of death globally, and cancer took 10 million patient lives in 2020.
According to the National Library of Medicine, for women who are diagnosed with gynecological cancers, the leading cause of death in ovarian cancer.
Finding early and quick intervention can significantly improve outcomes and increase survival rates for cancer patients.
Unfortunately, ovarian cancer is tough to detect early as it causes very few symptoms. In most cases, people are diagnosed at the late stages, which leads to poor patient outcomes.
80% of Ovarian Cancers are Not Detected Early
80% of ovarian cancers aren’t found in the early stages, when the tumour is small and hasn’t spread to a person’s lymph nodes or other tissues, according to the American Cancer Society (ACS).
The study was done by scientists from Cornell University, Weill Cornell Medicine, Lehigh University, University of Maryland, the Memorial Sloan Kettering Cancer Center, Albert Einstein College of Medicine, National Institute of Standards and Technology, and Hunter College High School.
The study authors wrote, “Serum biomarkers are insufficiently sensitive or diagnostic testing or specific to facilitate cancer screening.”
“In ovarian cancer, the very few found serum biomarkers are specific, yet insufficiently sensitive to find in the early stage of cancer and the impact to mortality rates of people with ovarian cancer.”
Figuring Out Lack of Biomarkers for Ovarian Cancer
To get around the lack of biomarkers for ovarian cancer, scientists developed carbon nanotubes using an AI-enabled nanosensor. Then, retaining over 260 blood serum samples, the scientists validated and trained several machine-learning finders to spot ovarian cancer.
Also named buckytubes, carbon nanotubes are lightweight hollow tubes consisting of nanoscale diameter carbon.
The chemically neutral nanotubes are three nanometers in size, and the length is usually just a few micrometres. Although they consist of two-dimensional folded graphene, the corrosion-resistant carbon nanotubes are a higher thermal capacity and are more robust than even steel.
Developing Artificial Neural Networks (ANN)
The researchers created artificial neural networks (ANN), Support Vector Machine models, and Random Forest for binary classes, logistic regression, and decision trees.
Bayesian optimization was also used, and custom MATLAB and Python code. According to the scientists, their answer had 87% sensitivity at 98% specificity and can adapt to identify other types of cancer.
This AI machine learning dramatically increases the accuracy of finding ovarian cancer versus current biomarker-based techniques. By combining the idea of nanotechnology and artificial intelligence machine learning, researchers have found a new way to find ovarian cancer that is much better than existing biomarker technologies.