UConn Health Using AI for Early Identification of Lung Cancer

UConn Health has implemented an Artificial Intelligence (“AI”) system to assist with the identification, management, and treatment of patients with lung cancer.

UConn Health has implemented an Artificial Intelligence (“AI”) system to assist with the identification, management, and treatment of patients with lung cancer.

Lung cancer is the most common type of cancer and the leading cause of cancer deaths in the world. 1.8 million people die annually from lung cancer with 134,000 deaths in the US alone.  Survival rates are high when the disease is caught early but, sadly, very low when diagnosed later. A patient diagnosed at the earliest stage ‘1A’ has a 90% chance of five-year survival compared to only 15% chance when the disease has progressed to later stages III or IV.

Late diagnosis of lung cancer is common because symptoms don’t usually occur in the early stages. By the time symptoms such as weight loss, coughs, shoulder pain or shortness of breath cause the patient to seek medical help, the disease has already progressed. In the US, screening programs have also been introduced to help detect lung cancer early but unfortunately these still face challenges in recruitment: less than 5% of all eligible individuals have undergone screening.

The new AI system, Virtual Nodule Clinic (VNC) from software vendor Optellum, offers an easier way to identify and diagnose lung cancer early. CT scans are used where lung nodules, the first sign of lung cancer, have been reported in scans acquired for an unrelated reason. These ‘incidental findings’ account for 90% of all diagnosed lung nodules. Studies show that around 60% are not followed up as the patient is managed only for the condition for which the scan was acquired. Most of these incidental findings are benign, but some are malignant and require urgent action.

The VNC installation at UConn Health uses a unique Patient Discovery AI module that automatically analyzes all CT radiology reports. Using natural language processing, it finds the patients who have reported lung nodules and ensures that they are all added to an easy-to-use nodule management dashboard, ready to review by the team. From there, clinicians can ensure the patients are appropriately followed up.

However, this isn’t the only part of the system that employs AI to assist the care pathway. The Optellum software also includes the only Lung Cancer Prediction tool cleared by the FDA as a medical device. This AI assesses any given lung nodule and provides a quantified risk score from 1-10; a higher score indicates a higher probability of malignancy. This has been shown to increase clinicians’ diagnostic accuracy and improve the quality of the subsequent treatment decision. This innovative technology will enable UConn Health to find more cancers and treat those cancers at an earlier stage.

Connecticut has an impressive lung cancer survival rate, ranking second in the nation but there is still more work to be done. It only ranks as average for identification of new cases and for the screening rate for high-risk patients. Furthermore, African American, Latino and Asian patients are less likely than Caucasians to be diagnosed early.

The clinical team led by Dr. Omar Ibrahim, associate professor of Medicine and director of Interventional Pulmonary at UConn Health and the Carole and Ray Neag Comprehensive Cancer Center, hopes to use the new AI software platform to improve the state’s statistics and get patients diagnosed and treated faster.

“It’s all about stage shift,” he explained. “Using the AI technology of Virtual Nodule Clinic I can find the patients who need care – before they even realize it themselves – and get them treated at the earliest possible stage.”

The Carole and Ray Neag Cancer Center is located on the Farmington campus of UConn Health, the only research hospital in the area where medical staff and scientists collaborate to create new and better treatments for our patients. For an appointment call 800-579-7822.