AI Increases Breast Cancer Diagnosis Accuracy to 92 Percent
Artificial Intelligence (AI) technology is beginning to transform itself into a vital tool in increasing the accuracy of diagnostic examinations for cancer, particularly breast cancer.
This technology is considered capable of helping medical professionals assess Human Epidermal Growth Factor Receptor 2 (HER2) status more precisely in order to determine the most appropriate therapy for patients.
Dr. Patricia Diana Prasetyo, MSi.Med, Sp.PA, a specialist in anatomical pathology, emphasised that the accuracy of HER2 status assessment is a crucial factor in breast cancer management. According to her, anti-HER2 targeted therapy will only provide maximum benefit if based on accurate and consistent examination results.
“Anti-HER2 targeted therapy, when given appropriately, can extend patient survival and improve quality of life. For this reason, the assessment or scoring of HER2 status must be accurate and consistent,” Patricia stated in an official statement on Monday (9 March).
The step to strengthen early detection has become increasingly urgent given the projected surge in cancer cases in Indonesia. Based on recent studies, the number of cancer cases in the country is predicted to increase by more than 70 percent by 2050 if preventive and early detection efforts are not strengthened from the outset.
Currently, approximately 400,000 new cancer cases are recorded annually, with mortality figures reaching 240,000 lives.
Patricia explained that the integration of AI as a clinical companion has been proven to significantly improve examination quality. Data from studies presented at the ASCO Annual Meeting 2025 showed that the use of AI was able to boost assessment accuracy to 92 percent.
Beyond accuracy, AI plays a major role in improving consistency among examiners. Whilst conventional methods have inter-examiner consistency rates of around 66 percent, AI assistance can increase this figure to around 82 percent.
Furthermore, this technology excels at detecting difficult-to-identify categories, namely HER2-low and HER2-ultralow categories. The use of AI was recorded as increasing detection capability in these subcategories by up to 40 percent compared to conventional assessment methods.
With the support of this technology, the utilisation of AI is expected to strengthen the accuracy of breast cancer classification whilst helping medical professionals in making faster, more accurate and measurable clinical decisions for patient safety.
A recent study in The Lancet Oncology revealed a tragic disparity: breast cancer mortality rates are declining in developed countries but have surged by nearly 100 percent in poorer countries.
The Director of Prevention and Control of Non-Communicable Diseases (PTM) at the Ministry of Health (Kemenkes), Siti Nadia Tarmizi, outlined the urgency of improving breast cancer early detection systems.
Health Minister Budi Gunadi Sadikin urged all Indonesian women aged over 30 years to take advantage of the Free Health Check (CKG) programme to prevent breast cancer.