Multi-center synergetic cancer prognosis prediction system based on multi-source migration learning

A multi-center synergetic cancer prognosis prediction system based on multi-source migration learning, comprising a model parameter setting process, a data screening terminal, and a multi-source migration learning processor wherein

• the model parameter setting process is arranged at a management center and is configured to set cancer prognosis prediction model parameters by a central processor, comprising a cancer category, a source center and a target center, sample characteristics, a sample data preprocessing method, and a prognosis index;
• the management center performs coordination management of resources of all clinical centers and accepts access of a user;
• the source center is one or more clinical centers with labeled samples for a specific cancer category, and is responsible for training of a source cancer prognosis prediction model;
• the target center is a clinical center with unlabeled samples for a specific cancer category, and is responsible for training a target cancer prognosis prediction model;
• the clinical center is one or more institutions that actually holds clinical data, and is responsible for screening sample data and training a cancer prognosis prediction model;
• the data screening terminal is arranged at a clinical center, the management center transmits a set model parameter to each of the clinical centers by a signal transmitter, each of the clinical centers screens data by utilizing the data screening terminal, inquires a sample feature and prognosis index data from a local database of the clinical center according to the model parameter, and preprocesses sample data according to a set sample data preprocessing method, and the source center obtains a labeled sample set and the target center obtains an unlabeled sample set;
• the multi-source migration learning processor comprises a source model training unit, a migration weight calculation unit, and a target model calculation unit;
• the source model training unit is arranged at each source center, it is assumed that there are K source centers denoted by S1,S2,S3. . . SK, and a terminal processor of an ith source center trains a local source cancer prognosis prediction model fSi(⋅) through its source model training unit and transmits a trained source model back to the management center by the signal transmitter;
• the migration weight calculation unit is arranged at the target center, and a terminal process of the target center receives the K source cancer prognosis prediction models sent by the management center by a signal receiver, it is assumed that there nT unlabeled samples at the target center, an ith unlabeled sample is expressed as xiT, the K source cancer prognosis prediction models are utilized to perform prognosis prediction of the sample xiT, respectively, to obtain a prediction label vector HiS:
• H
• i
• S
• =
• [
• f
• S
• 1
• (
• x
• i
• T
• )
• ,
• f
• S
• 2
• (
• x
• i
• T
• )
• ,
• …
• ,
• f
• S
• K
• (
• x
• i
• T
• )
• ]
• weighted summation is performed on the K prediction labels in the prediction label vector HiS to obtain a pseudo label ŷiT of the sample xiT:
• y
• ^
• i
• T
• =
• H
• i
• S
• ⁢
• θ
• =
• ∑
• j
• =
• 1
• K
• θ
• S
• j
• ⁢
• f
• S
• j
• (
• x
• i
• T
• )
• where θ=[θS1, θS2, θS2. . . , θSK]T represents a migration weight of each source model, a weight that minimizes a difference between two samples in a target center sample set is sought based on a smoothness assumption on the sample data of the target center, which is expressed as the following optimization problem:
• min
• θ:
• θ
• ′
• ⁢
• e
• =
• 1
• ,
• θ
• ≥
• 0
• ⁢
• ∑
• i
• ,
• j
• =
• 1
• n
• T
• (
• H
• i
• S
• ⁢
• θ
• -
• H
• j
• S
• ⁢
• θ
• )
• 2
• ⁢
• W
• ij
• where θ′ is a transposition of θ, e is a unit vector, and Wij indicates the similarity among samples;
• the aforementioned optimization problem is transformed into:
• min
• θ:
• θ
• ′
• ⁢
• e
• =
• 1
• ,
• θ
• ≥
• 0
• ⁢
• θ
• ′
• ⁢
• H
• S
• l
• ⁢
• L
• T
• ⁢
• H
• S
• ⁢
• θ
• where HS is a nT ×K matrix, and LT represents a graph Laplacian operator related to the target center, which is obtained by calculating according to LT =D−W, where W is a similarity matrix of samples of the target center, and D is a diagonal matrix obtained by calculating according to Dii=Σj=1nT Wij,
• therefore, the optimization problem is transformed into a standard quadratic programming problem, which is solved to obtain a migration weight θ;
• the target model calculation unit is arranged at the target center, obtains a sample pseudo label according to the migration weight θ, a terminal processor of the target model calculation unit trains the target cancer prognosis prediction model at the target center by utilizing the sample pseudo label, and transmits the trained target model back to the management center by a signal sender.