Development of a Pan-Cancer Screening Model Based on Blood Biomarkers

Condition(s) studied

Investigational drug(s) / intervention(s)

No intervention: This is an observational, retrospective study with no assigned interventions. Data are collected from existing medical records, including routine blood biomarkers (complete blood count, biochemistry, coagulation panel, tumor markers). No experimental drugs, devices, or procedures are administered. Only de-identified historical data are used for model development.

Study summary

This study aims to develop a pan-cancer screening model using routine blood biomarkers (including complete blood count, biochemical tests, coagulation panel, and tumor markers). The study is retrospective, collecting data from approximately 10,000,000 cancer patients diagnosed at multiple centers in China between January 2006 and September 2025. All patients have confirmed pathological diagnosis and complete blood test records. A Mixture of Experts (MoE) machine learning model will be built to predict the presence of various cancers (e.g., gastric, colorectal, liver, lung, ovarian cancer). The goal is to establish a low-cost, non-invasive screening tool suitable for large-scale population screening.

Eligibility

Primary outcome measure(s)

• Area under the ROC curve (AUC) — At study completion, approximately December 2030
  AUC of the MoE model for discriminating cancer from non-cancer controls.
• Sensitivity of the model — At study completion, approximately December 2030
  True positive rate of the pan-cancer screening model.
• Specificity of the model — At study completion, approximately December 2030
  True negative rate of the pan-cancer screening model.

Trial sites (1)

More Zhejiang Cancer Hospital trials in China

Other trials for the same condition