A new AI tool is helping doctors make better decisions for treating small brain metastases with stereotactic radiosurgery (SRS). This machine learning model analyzes factors like radiation dose, patient details, and treatment information to predict the risk of treatment failure at 6 months, 1 year, and 2 years after SRS.
Treating brain metastases under 2 cm is challenging, and traditional dosing methods of 20 Gy, 22 Gy, or 24 Gy often follow general rules without considering specific patient needs. This AI tool changes that by offering personalized insights.
Dr. Rupesh Kotecha, a radiosurgery expert at Baptist Health Miami Cancer Institute, explained, “We wanted to assess patient characteristics and treatment details to determine the risk of failure at different time points.”
The study, presented at the 2024 ASTRO meeting, included data from 235 patients treated at Miami Cancer Institute between 2017 and 2022. These patients underwent 358 SRS sessions to treat 1,503 brain metastases. Key findings include:
- Median patient age: 65 years
- Most common cancers: lung cancer (58.5%) and breast cancer (24.6%)
- Radiation doses: 20 Gy (20%), 22 Gy (29%), and 24 Gy (51%)
- Local failure rate: 9.2% of lesions
The AI tool demonstrated high accuracy (88%) and specificity (91%), especially in predicting 1-year outcomes. This technology not only helps doctors choose the best radiation dose but also guides follow-up care by adjusting MRI scan frequency based on patient risk.
Dr. Kotecha added that as more data from other institutions is added, the model’s predictions will become even more reliable. This AI tool shows promise for improving SRS treatments and making patient care more personalized and efficient.
Source: targetedonc
