US 12,224,043 B2
Methods for determining treatment for cancer patients
Li Yu, Gaithersburg, MD (US); Harry Yang, Gaithersburg, MD (US); Mohammed Dar, Gaithersburg, MD (US); Lorin Roskos, Gaithersburg, MD (US); Jean-Charles Soria, Gaithersburg, MD (US); Charles Ferte, Gaithersburg, MD (US); Wei Zhao, Gaithersburg, MD (US); Aline Gendrin Brokmann, Cambridge (GB); Jolyon Faria, Cambridge (GB); Pralay Mukhopadhyay, Wilmington, DE (US); and Ikbel Achour, Gaithersburg, MD (US)
Assigned to MEDIMMUNE, LLC, Gaithersburg, MD (US)
Filed by MEDIMMUNE, LLC, Gaithersburg, MD (US)
Filed on Sep. 13, 2023, as Appl. No. 18/367,923.
Application 18/367,923 is a continuation of application No. 16/657,375, filed on Oct. 18, 2019, granted, now 11,798,653.
Claims priority of provisional application 62/854,566, filed on May 30, 2019.
Claims priority of provisional application 62/747,420, filed on Oct. 18, 2018.
Prior Publication US 2023/0420081 A1, Dec. 28, 2023
Int. Cl. G16B 40/20 (2019.01); G06N 5/01 (2023.01); G06N 20/20 (2019.01); G16B 5/20 (2019.01); G16H 20/40 (2018.01); G16H 50/20 (2018.01); G16H 50/30 (2018.01); G16H 50/50 (2018.01); A61K 38/00 (2006.01); C07K 16/28 (2006.01); G01N 33/574 (2006.01)
CPC G16B 40/20 (2019.02) [G06N 5/01 (2023.01); G06N 20/20 (2019.01); G16B 5/20 (2019.02); G16H 50/20 (2018.01); G16H 50/30 (2018.01); G16H 50/50 (2018.01); A61K 38/00 (2013.01); C07K 16/2818 (2013.01); G01N 33/57484 (2013.01); G16H 20/40 (2018.01)] 20 Claims
OG exemplary drawing
 
1. A computing system comprising:
computer memory configured to store model training data, wherein the model training data involves a first set of cancer patients that underwent cancer treatments, wherein the model training data associates (i) results from laboratory tests conducted on the first set of cancer patients and tumor types of the first set of cancer patients with (ii) whether individuals from the first set of cancer patients died within a threshold number of weeks from initiation of the cancer treatments; and
one or more processors in communication with the computer memory and configured to execute program instructions to:
generate a plurality of variations of a gradient boosting machine learning model, wherein generating each of the variations comprises:
training, based on the model training data, a sequential series of decision trees until a stopping condition is reached using a respective set of hyperparameters, wherein each subsequent decision tree of the sequential series of decision trees is constructed based on residual values of its preceding decision tree of the sequential series of decision trees, and wherein the respective sets of hyperparameters are different for each of the variations; and
forming an additive function of the sequential series of decision trees, wherein each of the variations of the gradient boosting machine learning model is usable to predict whether a further cancer patient dies within the threshold number of weeks based on the additive function applied to results from the laboratory tests as conducted on the further cancer patient and a tumor type of the further cancer patient;
apply each of the variations of the gradient boosting machine learning model to the model training data; and
select, from among the variations of the gradient boosting machine learning model, a particular variation that provides predictions of whether the first set of cancer patients die within the threshold number of weeks from initiation of the cancer treatments within a threshold degree of accuracy.