Artificial intelligence cuts 12% of interval cancers and detects aggressive breast tumors earlier
Mammographic screening supported by artificial intelligence recorded a 12% reduction in interval cancers in a randomized study carried out at Suécia. The MASAI trial, coordinated by Lund University, involved more than 100 thousand women and demonstrated greater sensitivity in identifying tumors, especially the most aggressive ones. The findings indicate that the technology helps in early detection without compromising the specificity of the exams.
The application of AI occurred in regular mammographic screening programs in the European country. Pesquisadores compared the method with traditional double reading by radiologists and observed consistent benefits across different age groups and breast densities.
Interval cancers arise between rounds of exams and have more unfavorable characteristics. The decrease in these cases represents significant progress for public health programs.
- 12% reduction in overall interval cancers.
- 16% fewer invasive tumors detected late.
- 21% fewer large cases (T2+).
- 27% fewer non-luminal A subtypes, considered aggressive.
How technology works in screening
Artificial intelligence analyzes mammogram images and assigns a risk score on a scale of 1 to 10. Exames classified as low risk receive a single reading by a radiologist, while those at higher risk undergo double assessment.
This stratification maintains diagnostic accuracy and optimizes clinical workflow. The system highlights suspicious regions in the images to support medical decisions.
Researchers observed sensitivity of 80.5% in the AI group, compared to 73.8% in the control group. Specificity remained at 98.5% in both groups.
Main results of the MASAI trial
The study followed women with a median age of approximately 54 years. The AI intervention reduced the exam reading burden by 44% for radiologists.
Tumors identified early were smaller in size and lacked lymph node involvement in most cases. Esses factors are directly associated with better survival rates.
Detection occurred at the screening stage in 81% of cases with AI, compared to 74% with the conventional method.
Benefits observed in clinical practice
Professionals reported greater confidence in technologically supported analyses. The tool easily integrates with existing hospital IT systems.
Swedish regions have already adopted the model in their daily routine. Implementation only requires software compatible with digital mammography equipment.
Radiologists maintain a central role in the process. Pacientes express a preference for human assessment combined with technology.
The trial recorded 82 interval cancers in the intervention group versus 93 in the control group. The difference reinforces the effectiveness of the hybrid approach.
Application to different patient profiles
The superiority of sensitivity remained consistent regardless of breast density. Mulheres with dense breasts especially benefit from automated analysis.
Leia Também
Similar results appeared across varying age groups. The performance was valid for invasive cancers, but did not differ significantly in in situ cases.
Workload reduction and operational efficiency
Radiologists face increasing volume of exams in population-based programs. AI alleviates this demand without increasing false positive rates.
False positives generate anxiety and unnecessary procedures in patients. The balance achieved in the study avoids this additional risk.
The reduced load allows for greater focus on complex cases. Profissionais save time for multidisciplinary discussions.
Global expansion potential
European countries are evaluating similar implementation in public health systems. Technology adapts to contexts with a shortage of specialists.
Breast cancer records millions of new cases annually around the world. Early Detecção raises cure rates above 90% in early stages.
Additional studies confirm positive trends. Pesquisas retrospectives have already indicated superior AI accuracy in image analysis.
Organized tracking programs benefit directly. Gradual integration preserves quality of care.
Limitations and future perspectives
The trial emphasizes the need for constant human supervision. Pacientes value interaction with health professionals.
Continuous algorithm training requires diverse data. Atualizações guarantee performance in heterogeneous populations.
Additional research evaluates impact on long-term mortality. Prolonged Seguimento will clarify definitive clinical benefits.
The combination of AI and medical expertise represents a promising model. Avanços Technological technologies continue to improve screening tools.
The MASAI study sets the benchmark for future trials. Resultados encourage responsible adoption of innovation.
