A Machine Learning Approach to Support Treatment Identification for Chiari I Malformation

Chiari I malformation is characterized by the herniation of cerebellar tonsils below foramen magnum. It often accompanied syringomyelia and neurosurgical management still controversial. In fact, it frequent that some symptomatic patients initially undergo bony decompression posterior fossa need in a short time more invasive surgery with higher morbility (e.g., dural plastic, or without tonsillar coarctation) because unsatisfactory results at MRI controls. This study proposes machine learning approach (based on SVM classifier), applied to different morphometric indices estimated from sagittal information patient (i.e., age symptoms diagnosis), recognize risk clinical deterioration. Our database includes 58 pediatric who underwent treatment. A negative outcome 1 year intervention was observed 38% them (accuracy 62%). algorithm allows us increase accuracy about 71%, showing be valid support neurosurgeons refining picture.