Systems and methods for building an electronic data capture system
Disclosed are methods and systems for building an electronic data capture (EDC) system. The method includes forming a training dataset from a set of protocol documents and corresponding EDC builds. The method includes using the set of protocol documents and corresponding EDC builds of the training dataset to fine tune a pre-trained language model to produce an EDC build model. The method includes inputting a protocol document to the EDC build model to produce an EDC build prediction. The method includes generating data structures of the EDC system based at least in part on the EDC build prediction.
1 . A method of building an electronic data capture (EDC) system, the method comprising:
forming a training dataset from a set of clinical trial protocol documents and corresponding EDC builds, each EDC build comprising a machine-readable specification file defining electronic case report form (eCRF) data fields, edit checks for validating data entered into the eCRF data fields, and folder structures for storing study data within the EDC system;
using the set of clinical trial protocol documents and corresponding EDC builds of the training dataset to tune a pre-trained language model to produce an EDC build model;
inputting a clinical trial protocol document to the EDC build model to produce an EDC build prediction comprising a corresponding machine-readable specification file; and
generating data structures of the EDC system based at least in part on the EDC build prediction by loading the machine-readable specification file into an EDC system platform to instantiate the eCRF data fields, the edit checks for validating the data entered into the eCRF data fields, and the folder structures for storing study data within the EDC system.
2 . The method of claim 1 , wherein said generating data structures of the EDC system comprises generating one or more electronic case report forms.
3 . The method of claim 1 , wherein said generating data structures of the EDC system comprises generating one or more folder structures for managing EDC system information.
4 . The method of claim 1 , wherein said using the set of protocol documents and corresponding EDC builds of the training dataset to tune the pre-trained language model comprises adjusting hyperparameters of the pre-trained language model.
5 . The method of claim 1 , wherein the protocol document is a natural language text document.
6 . The method of claim 5 , wherein the protocol document comprises a schedule of assessments in tabular form.
7 . The method of claim 1 , wherein the EDC build prediction is in the form of an extensible markup language (XML) document.
8 . The method of claim 1 , wherein said forming the training dataset comprises dividing the set of protocol documents and corresponding EDC builds into the training dataset and a testing dataset.
9 . The method of claim 8 , further comprising:
inputting the protocol documents of the testing dataset to the EDC build model to produce EDC build predictions; and
analyzing the EDC build predictions produced from the protocol documents of the testing dataset to determine accuracy of the EDC build model.
10 . The method of claim 9 , further comprising performing tuning of the EDC build model based at least in part on said analyzing the EDC build predictions produced from the protocol documents of the testing dataset.
11 . A system for building an electronic data capture (EDC) system, the system comprising:
a computer having one or more processors in communication with a memory, the memory storing instructions executable by said one or more processors to perform:
forming a training dataset from a set of clinical trial protocol documents and corresponding EDC builds, each EDC build comprising a machine-readable specification file defining electronic case report form (eCRF) data fields, edit checks for validating data entered into the eCRF data fields, and folder structures for storing study data within the EDC system;
using the set of clinical trial protocol documents and corresponding EDC builds of the training dataset to tune a pre-trained language model to produce an EDC build model;
inputting a clinical trial protocol document to the EDC build model to produce an EDC build prediction comprising a corresponding machine-readable specification file; and
generating data structures of the EDC system based at least in part on the EDC build prediction by loading the machine-readable specification file into an EDC system platform to instantiate the eCRF data fields, the edit checks for validating the data entered into the eCRF data fields, and the folder structures for storing study data within the EDC system.
12 . The system of claim 11 , wherein said generating data structures of the EDC system comprises generating one or more electronic case report forms.
13 . The system of claim 11 , wherein said generating data structures of the EDC system comprises generating one or more folder structures for managing EDC system information.
14 . The system of claim 11 , wherein the EDC build prediction is in the form of an extensible markup language (XML) document.
15 . The system of claim 11 , wherein said forming the training dataset comprises dividing the set of protocol documents and corresponding EDC builds into the training dataset and a testing dataset.
16 . The system of claim 15 , further comprising:
inputting the protocol documents of the testing dataset to the EDC build model to produce EDC build predictions; and
analyzing the EDC build predictions produced from the protocol documents of the testing dataset to determine accuracy of the EDC build model.
17 . The system of claim 16 , further comprising performing tuning of the EDC build model based at least in part on said analyzing the EDC build predictions produced from the protocol documents of the testing dataset.
18 . A non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a computer, cause said one or more processors to perform a method for building an electronic data capture (EDC) system, the method comprising:
forming a training dataset from a set of clinical trial protocol documents and corresponding EDC builds, each EDC build comprising a machine-readable specification file defining electronic case report form (eCRF) data fields, edit checks for validating data entered into the eCRF data fields, and folder structures for storing study data within the EDC system;
using the set of clinical trial protocol documents and corresponding EDC builds of the training dataset to tune a pre-trained language model to produce an EDC build model;
inputting a clinical trial protocol document to the EDC build model to produce an EDC build prediction comprising a corresponding machine-readable specification file; and
generating data structures of the EDC system based at least in part on the EDC build prediction by loading the machine-readable specification file into an EDC system platform to instantiate the eCRF data fields, the edit checks for validating the data entered into the eCRF data fields, and the folder structures for storing study data within the EDC system.
19 . The computer-readable medium of claim 18 , wherein said generating data structures of the EDC system comprises generating one or more electronic case report forms.
20 . The computer-readable medium of claim 18 , wherein said generating data structures of the EDC system comprises generating one or more folder structures for managing EDC system information.
21 . The computer-readable medium of claim 18 , wherein the EDC build prediction is in the form of an extensible markup language (XML) document.
22 . The computer-readable medium of claim 18 , wherein said forming the training dataset comprises dividing the set of protocol documents and corresponding EDC builds into the training dataset and a testing dataset.
23 . The computer-readable medium of claim 22 , further comprising:
inputting the protocol documents of the testing dataset to the EDC build model to produce EDC build predictions; and
analyzing the EDC build predictions produced from the protocol documents of the testing dataset to determine accuracy of the EDC build model.
24 . The computer-readable medium of claim 23 , further comprising performing tuning of the EDC build model based at least in part on said analyzing the EDC build predictions produced from the protocol documents of the testing dataset.