August 28, 2017 — 20/20 GeneSystems Inc. recently released in China what is believed to be the world’s first machine learning algorithm commercialized for use in conjunction with blood tests to aid in the early detection of lung cancer. The cloud-accessible software, marketed in China by MyBioMed Ltd., is expected to be utilized over the next six months by at least 20 hospitals in five or more cities to help in distinguishing between benign versus malignant pulmonary nodules following computed tomography (CT) imaging.
The algorithm, built and validated with data from over 1,000 patients tested at the Tianjin Cancer Institute combines the values of biomarker — tumor-associated proteins in the blood — and individual patient characteristics such as their age, smoking status and nodule size. It could potentially increase the number of cancers detected by more than 50 percent over current methods. Artificial neural networks are incorporated to permit continuous improvements as more data is acquired over time.
Lung cancer is a significant burden in China as more than half of adult men in that country are current or former smokers and heavy air pollution further contributes to the disease. "Lung cancer in China has a very high disease incidence and the current solution is to do early screening which often leads to good outcomes at a low cost,” said Yilong Wu, M.D. chair of the Chinese Society of Clinical Oncology and himself a chest surgeon and lung cancer expert. “Low-dose CT (LDCT) screening has been recognized worldwide as an approach to reduce lung cancer mortality but these scans uncover many nodules that turn out to be benign, and unneeded biopsies and surgery can be dangerous to many patients,” he added. “I am very optimistic about the approach introduced by 20/20 and MBM of adding bioinformatics and artificial intelligence methods to improve the sensitivity and specificity of detection and look forward to seeing continued expansion of their data sets and confirmatory studies."
20/20’s cloud-based lung nodule assessment software is the company’s first product marketed using a software-as-a-service (SaaS) revenue model, according to 20/20 CEO Jonathan Cohen.
For more information: www.2020gene.com