Cancer screening model, and construction method and construction device of cancer screening model
A construction method and cancer technology, applied in the fields of medical data mining, character and pattern recognition, medical automatic diagnosis, etc., can solve problems such as difficulty in improving the accuracy of detection, and achieve the effect of improving accuracy.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0071] Objective: Through the processed data characteristics, model training, then use the test set verification model accuracy.
[0072] step:
[0073] 1. Establish a data set consisting of 29 healthy urine samples, 21 bladder cancer patients.
[0074]2. Baseline Baselines are constructed based on the CFDNA sample of normal population. Construction Process Step 1: Sequencing 60 normal populations CFDNA compared to the reference genome, and obtains the BAM file. Step 2: By referring to the Bed file of the genomic Fasta file with the WGS, build a new coordinate file preProcessed.InterVal.List using the preducessintervals command in Gatk. Step 3: By the coordinate file available in step 2 Preprocessed.Interval.List, some reference genome Fasta files, 60 normal populations CFDNA sequencing comparison BAM files, using the collection of the collectreadcounts command to get each sample The corresponding READS statistics file records the number of READs in each interval bin. Step 4: By r...
Embodiment 2
[0093] In this embodiment, a construction device for a cancer screening model is provided. The build device includes: CNVBaseline build module, set to construct a different CNV Baseline; genome feature acquisition module based on different data, setting SWGS data to detect samples with different CNV Baseline, and acquires multiple genome; The final genome feature determination module is set to establish a training set and test set using SWGS data to be detected, and the value of each genomic feature is used to select a genomic feature of the AUC value as the final genomic feature; model training module, set to Model training is performed on the final genomic characteristics to obtain cancer screening models.
[0094] Specifically, cancer is bladder cancer, and the sample to be tested includes a healthy population of urine peatdracery with bladder cancer patients with urine.
[0095] The CNV Baseline build module includes: CFCNV Baseline constructing a sub-module set to select a pr...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


