NEW DEEP LEARNING LUNG CT TOOL
The new tool uses automated cancer classifiers seen in the conventional CT scans of the chest and applies them to LungDirect’s advanced learning algorithms.
LungDirect’s algorithms are built and validated from over 1,000 patients that were tested at the Tianjin Cancer Institute. The test combines the values of biomarkers and individual patient characteristics – e.g. age, smoking status and nodule size. The anticipation is that this will increase the number of cancers detected by more than 50 percent in comparison to current methods. Artificial neural networks have also been incorporated into the algorithm to continuously improve the system as more data in acquired.
LungDirect’s system creates a cancer classifier to direct the algorithm towards a more detailed outcome. Deep learning tools such as these have already achieved impressive diagnostic results, but LungDirect will be the first screening and incidentaloma system to fully include this technology. The company are still working on improving the system by collaborating with the top academic cancer centres to combine imaging techniques, machine learning and deep electronic health record integrations to develop big data channels that can train the LungDirect platform to achieve superior performance.
To find the latest developments surrounding machine learning in medical imaging, register for your free ticket to the Medical Imaging Convention here.
Additionally, if you are interested in exhibiting at next year’s convention and meeting thousands of decision makers actively looking to purchase products and services, please contact Nick Woore, Event Director, now on:
Tel: 01179 296 097