ELIC Publishes Paper Showing Open Source System Supports Deep Learning AI and Quantitative Imaging Analyses on Cloud-Based Datasets
08 / 15 / 23
The International Association for the Study of Lung Cancer (IASLC), Accumetra, and numerous international CT lung cancer screening researchers published a paper in the Journal of Thoracic Oncology (JTO) showing that the open source Early Lung
Imaging Confederation (ELIC) system can support deep learning AI and quantitative imaging analyses on diverse and globally distributed cloud-based datasets. This paper provides results from deep learning and quantitative imaging experiments performed on 697 lung cancer screening cases with two CT image acquisition time points each. ELIC is designed to support the computational study of large collections of early lung cancer cases, where image data stays on the cloud within local regions while AI and quantitative imaging algorithms are distributed to those data locations. To read more about how this new global infrastructure can advance medical AI visit: https://us06web.zoom.us/j/3456879728
Accumetra Publishes Method That Predicts The Bias and Precision Performance of CT Lung Nodule Volume Measurements
08 / 03 / 23
Lung imaging researchers from Accumetra, Mount Sinai, and Cleveland Clinic published a manuscript in July indicating for the first time that CT lung volume measurements can be reliably predicted using a new calibration-based analysis method. The
study showed that good volume measurement prediction results (bias, precision) could be obtained even when using low-cost reference objects, specifically three rolls of 3M Scotch Tape (TM). This novel approach, which formed the underlying computer-vision, math, and physics foundation of Accumetra's CTLX1 and CTLX2 phantoms, is being used internationally to improve lung nodule measurements. Numerous research projects have been launched as a result of this work, investigating improved methods to predict how a CT scanner will perform when requested to measure the volume of small objects near the limits of CT resolution. To read more, visit https://qims.amegroups.org/article/view/115487/html . Contact Accumetra at info@accumetra.com for information on how this new approach is being used to better monitor and optimize CT image acquisition for precise small lung nodule volume measurement in standard clinical care, lung cancer screening, and clinical trials.
