Artificial Intelligence Enables Earlier Detection of Lung Cancer in General Practice

Strategic Insights -

A new study published in the British Journal of General Practice has found that artificial intelligence can facilitate earlier detection of lung cancer in general practice using routine patient data. According to the research, the journey of 80% of patients diagnosed with lung cancer starts in general practice, with 75% of patients diagnosed at an advanced stage, leading to 80% mortality within one year. The study used natural language processing and machine learning to analyze electronic patient data from four networks in the Netherlands, identifying 2386 patients with lung cancer.

Key Takeaways:

  • The study found that artificial intelligence-based support enables earlier detection of lung cancer in general practice using readily available text in patient files.
  • The research used natural language processing and machine learning to analyze electronic patient data from 525,526 patients, identifying 2386 patients with lung cancer.
  • The algorithm was able to predict lung cancer diagnosis 5 months before referral, with an area under the curve of 0.88 (95% CI = 0.86 to 0.89) during internal validation.
  • The study's conclusion is that artificial intelligence-based support is a valuable tool for earlier detection of lung cancer in general practice, but requires additional prospective clinical evaluation to establish its effectiveness.
  • The research was funded by KWF Kankerbestrijding and the Departments of Medical Informatics and General Practice of Amsterdam University Medical Center (UMC).
  • The study's findings were validated by the Dutch cancer registry, with both structured and free-text data used to predict lung cancer diagnosis.

Statistics:

  • 80% of patients diagnosed with lung cancer start their journey in general practice.
  • 75% of patients are diagnosed with lung cancer at an advanced stage (3 or 4).
  • 80% mortality rate within one year for patients diagnosed with lung cancer at an advanced stage.
  • 2386 patients with lung cancer were identified from a dataset of 525,526 patients.
  • The algorithm was able to predict lung cancer diagnosis 5 months before referral, with an area under the curve of 0.88 (95% CI = 0.86 to 0.89) during internal validation.

Sources:

  • "Artificial Intelligence for Early Detection of Lung Cancer In Gps' Clinical Notes: a Retrospective Observational Cohort Study." British Journal of General Practice, 2025.
  • Health & Medicine Week. May 30, 2025; p 4009.
  • Royal Coll General Practitioners, 14 Princes Gate, Hyde Park, London SW7 1PU, England (publisher of the British Journal of General Practice).
  • Martijn C. Schut, Vrije Universiteit Amsterdam, Amsterdam Univ Med Ctr Umc, Dept. of Laboratory Medicine, Amsterdam, Netherlands (contact for additional information).