公开(公告)号：EP4361944A1

公开(公告)日：2024-05-01

申请号：EP22306652.3

申请日：2022-10-31

申请人： Sartorius Stedim Data Analytics AB ,  UNIVERSITE CLAUDE BERNARD - LYON 1 ,  Institut National des Sciences Appliquées de Lyon ,  Ecole Superieure de Chimie Physique Electronique de Lyon ,  Centre national de la recherche scientifique

发明人： SJÖGREN, Rickard ,  REIF, Oscar-Werner ,  TRYGG, Johan ,  PETIOT, Emma ,  MARQUETTE, Christophe ,  CHASTAGNIER, Laura

IPC分类号： G06T7/00

CPC分类号： G06T7/0012 ,  G06T2207/1005620130101 ,  G06T2207/1008820130101 ,  G06T2207/2007620130101 ,  G06T2207/2008420130101 ,  G06T2207/3002420130101 ,  G06T2207/1006420130101

摘要： Methods for monitoring a process for producing multicellular structures from a cell population in a 3D cell culture are described. The methods comprise obtaining one or more 3D images of the 3D cell culture that are acquired by imaging the cell culture at different depths and/or from different angles, processing the one or more images to obtain a plurality of image-derived features, and determining the value of one or more metrics indicative of the progress or outcome of the maturation process using a statistical model adapted to predict the values of the one or more metrics indicative of the progress or outcome of the maturation process using inputs comprising the plurality of image-derived features. Related methods, systems and products are also described.

公开(公告)号：EP4290412A2

公开(公告)日：2023-12-13

申请号：EP23204765.4

申请日：2018-09-05

申请人： Sartorius Stedim Data Analytics AB

发明人： SJÖGREN, Rickard ,  TRYGG, Johan

IPC分类号： G06N3/0442

摘要： A computer-implemented method for data analysis is provided. The method comprises: obtaining a deep neural network (100) for processing data and at least a part of a training dataset used for training the deep neural network, the deep neural network comprising a plurality of hidden layers, the training dataset including possible observations that can be input to the deep neural network, the deep neural network being trained using the training dataset; obtaining first sets of intermediate output values that are output from at least one of the plurality of hidden layers, each of the first sets of intermediate output values obtained by inputting a different one of the possible observations included in said at least the part of the training dataset; constructing a latent variable model using the first sets of intermediate output values, the latent variable model providing a mapping of the first sets of intermediate output values to first sets of latent variables for the latent variable model in a sub-space that has a dimension lower than a dimension of the sets of the intermediate outputs; and storing the latent variable model and the first sets of latent variables for the latent variable model in a storage medium.

公开(公告)号：EP3847586A1

公开(公告)日：2021-07-14

申请号：EP19762167.5

申请日：2019-09-05

申请人： Sartorius Stedim Data Analytics AB

发明人： SJÖGREN, Rickard ,  TRYGG, Johan

IPC分类号： G06N3/04

公开(公告)号：EP3831924A1

公开(公告)日：2021-06-09

申请号：EP19213037.5

申请日：2019-12-03

申请人： Sartorius Stedim Data Analytics AB

发明人： GRIMM, Christian ,  SCHLACK, Stefan ,  TRYGG, Johan

IPC分类号： C12M1/36 ,  G16B40/20

摘要： A computer implemented and a system for adapting control of a cell culture in a production-scale vessel with regard to a starting medium are provided. The method comprises providing multiple production-scale process trajectories, each derived from a successfully controlled cell culture. The method further comprises receiving a media lot for the cell culture. The method further comprises sampling first media from the media lot for possible use in the production-scale vessel. Moreover, the method comprises starting a seed train using the first media to achieve inoculation of the production-scale vessel. The method further comprises providing a plurality of micro-scale vessels in a process control device, wherein the production-scale is greater than the micro-scale. The method further comprises sampling second media from the media lot for the micro-scale vessels, wherein each of the micro-scale vessels receives a representative portion of the media lot. In addition, the method comprises introducing cells from the seed train into the micro-scale vessels to start cell cultures in each of the micro-scale vessels.

公开(公告)号：EP3620983A1

公开(公告)日：2020-03-11

申请号：EP18192649.4

申请日：2018-09-05

申请人： Sartorius Stedim Data Analytics AB

发明人： SJÖGREN, Rickard ,  TRYGG, Johan

IPC分类号： G06N3/04

摘要： A computer-implemented method for data analysis is provided. The method comprises: obtaining a deep neural network (100) for processing images and at least a part of a training dataset used for training the deep neural network, the deep neural network comprising a plurality of hidden layers, the training dataset including possible observations that can be input to the deep neural network; obtaining first sets of intermediate output values that are output from at least one of the plurality of hidden layers, each of the first sets of intermediate output values obtained by inputting a different one of the possible input images included in said at least the part of the training dataset; constructing a latent variable model using the first sets of intermediate output values, the latent variable model providing a mapping of the first sets of intermediate output values to first sets of projected values in a sub-space that has a dimension lower than a dimension of the sets of the intermediate outputs; receiving an observation to be input to the deep neural network; obtaining a second set of intermediate output values that are output from said at least one of the plurality of hidden layers by inputting the received observation to the deep neural network; mapping, using the latent variable model, the second set of intermediate output values to a second set of projected values; and determining whether or not the received observation is an outlier with respect to the training dataset based on the latent variable model and the second set of projected values.

公开(公告)号：EP4290412A3

公开(公告)日：2024-01-03

申请号：EP23204765.4

申请日：2018-09-05

申请人： Sartorius Stedim Data Analytics AB

发明人： SJÖGREN, Rickard ,  TRYGG, Johan

IPC分类号： G06N3/04 ,  G05B23/02 ,  G06N3/0442 ,  G06N3/0455 ,  G06N3/0464 ,  G06N3/048 ,  G06N3/092 ,  G06N3/088

摘要： A computer-implemented method for data analysis is provided. The method comprises: obtaining a deep neural network (100) for processing data and at least a part of a training dataset used for training the deep neural network, the deep neural network comprising a plurality of hidden layers, the training dataset including possible observations that can be input to the deep neural network, the deep neural network being trained using the training dataset; obtaining first sets of intermediate output values that are output from at least one of the plurality of hidden layers, each of the first sets of intermediate output values obtained by inputting a different one of the possible observations included in said at least the part of the training dataset; constructing a latent variable model using the first sets of intermediate output values, the latent variable model providing a mapping of the first sets of intermediate output values to first sets of latent variables for the latent variable model in a sub-space that has a dimension lower than a dimension of the sets of the intermediate outputs; and storing the latent variable model and the first sets of latent variables for the latent variable model in a storage medium.

公开(公告)号：EP3620986A1

公开(公告)日：2020-03-11

申请号：EP19180972.2

申请日：2019-06-18

申请人： Sartorius Stedim Data Analytics AB

发明人： SJÖGREN, Rickard ,  TRYGG, Johan

IPC分类号： G06N3/04

摘要： A computer-implemented method for analysis of cell images is provided. The method comprises: obtaining a deep neural network (100) for processing images and at least a part of a training dataset used for training the deep neural network, the deep neural network comprising a plurality of hidden layers and being trained using the training dataset, the training dataset including possible cell images that can be input to the deep neural network; obtaining first sets of intermediate output values that are output from at least one of the plurality of hidden layers, each of the first sets of intermediate output values obtained by inputting a different one of the possible input images included in said at least the part of the training dataset; constructing a latent variable model using the first sets of intermediate output values, the latent variable model providing a mapping of the first sets of intermediate output values to first sets of projected values in a sub-space that has a dimension lower than a dimension of the sets of the intermediate outputs; receiving a new cell image to be input to the deep neural network; obtaining a second set of intermediate output values that are output from said at least one of the plurality of hidden layers by inputting the received new cell image to the deep neural network; mapping, using the latent variable model, the second set of intermediate output values to a second set of projected values; and determining whether or not the received new cell image is an outlier with respect to the training dataset based on the latent variable model and the second set of projected values.