Toronto, ON – Altis Labs, Inc. (“Altis”), the computational imaging platform for advancing precision medicine, is announcing that results of its imaging-based stratification study in early-stage lung cancer patients will be presented at the American Society of Clinical Oncology 2021.
The study describes how imaging-based prognostication (IPRO), a fully-automated deep learning system, can be applied to pretreatment computed tomography (CT) scans to stratify stage I & II lung cancer patients into “low” and “high” risk groups.
Of the 260 patients that underwent surgical resection in the study’s withheld validation set, the 52 patients in the highest risk quintile had a 14.2-fold (95% CI 4.3-46.8, p < 0.001) increase in 5-year mortality hazard compared to the 52 patients in the lowest risk quintile.
IPRO was trained on 2,924 pretreatment CTs of 1,689 lung cancer patients with known 5-year survival outcomes. IPRO first localizes the lungs and subsequently evaluates imaging features present within the thorax using a three-dimensional convolutional neural network (3DCNN) to generate a mortality risk prediction. Predictions are entirely derived from imaging features.
IPRO’s ability to stratify early-stage lung cancer patients in an easily accessible, low-cost manner has the potential to inform treatment decision making, as well as enhance covariate adjustments and treatment effect quantification in clinical trials.
The study's lead author Felipe S. Torres, MD, PhD, cardiothoracic radiologist from the Joint Department of Medical Imaging at Toronto General Hospital, will be presenting the results in a poster discussion session (abstract #1552).
Altis is a clinical information company providing computational imaging tools to advance precision medicine. Altis’ software platform Nota enables researchers to operationalize imaging data and leverage predictive imaging insights at scale. Life sciences companies use Nota to accelerate and optimize R&D of their most promising therapies across all stages of clinical development. Altis is headquartered in Toronto, a city recognized for its deep learning research and medical institutions.