ScaleSurfer base all-data Stats Prediction Model

This repository contains a ScaleSurfer multi-head model for predicting FreeSurfer-style .stats targets from a T1w image and an aparc+aseg segmentation.

Files

  • stats_model.pt: full PyTorch stats checkpoint, including frozen encoder weights, prediction heads, target normalization, and target columns.
  • config.json: architecture and feature-schema metadata needed by ScaleSurferStatsPredictor.
  • metadata.json: checksums and training/evaluation metadata.
  • summary.csv, history.csv, target_metrics.csv: copied training diagnostics.

Test Summary

Group Targets Values Normalized MAE Median absolute percent error
aseg 119 39699 0.4714784485426931 7.629659175872803
global 64 16070 0.212405763403243 1.7018315196037292
lh_aparc 306 106742 0.4562009682892021 8.101013660430908
rh_aparc 306 106738 0.45695450692734807 7.982769966125488

Loading

from scalesurfer.stats import ScaleSurferStatsPredictor

predictor = ScaleSurferStatsPredictor.from_pretrained('base')
features = predictor.predict_subjects(subjects_dir, subjects, return_format="wide")

This model is intended for research workflows and is not a clinical diagnostic device.

Downloads last month
51
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support