Cluster Heatmap, Nomenclature Changes, and Automatic Quality Control Update

Summary of Changes

• Added a new graph type: Cluster Heatmap
• Add new columns for quick tracking quality control
• Modifications to column names to improve consistency and accuracy

Cluster Heatmap

This update brings a new graph to the Visualize page: Cluster Heatmap. The cluster heatmap allows for in-depth analysis of all feature columns at both group and individual levels.

Group-Level Cluster Heatmaps

This mode visualizes mean Z-scores across all feature columns, complete with significance markers (e.g., *, **, ***). It incorporates ANOVA-based statistical testing with Bonferroni-corrected p-values, highlighting significant features across different groups. Hierarchical clustering orders features, making group comparisons intuitive and clear.

Individual-Level Cluster Heatmaps

This mode presents Z-scores for all feature columns at the recording level. Recordings are sorted using hierarchical clustering, independent of group identity, to uncover latent cluster structures within the dataset. Once clustering is applied, the group identity is recovered and displayed, allowing users to assess whether these underlying structures align with predefined groupings. This approach is valuable for identifying hidden patterns and verifying if natural data clusters correspond with existing group labels.

Tracking Quality Control

Average tracking likelihood for each of the four paws

Averages the tracking likelihood reported by the DLC model for each paw, scaled between 0-1.

Paws tracking quality control flag

• 0: good tracking quality
• 1: potential imprecise paw luminance readouts for the given recording

Nomenclature Changes

Body parts in the skeleton tracking model

• Hip is now sacrum
• Sternumhead is now sternal notch
• Sternumtail is now xyphoid
• Left/right Shoulder is now Left/right elbow
• Left/right hip is now left/right femur
• Left/right ankle is now left/right heel

Standardized acronyms for paws

To ensure consistency, acronyms for paw labels are now unified across the software, summary readouts, and user manual.

• LHP: left hind paw
• RHP: right hind paw
• LFP: left front (fore) paw
• RFP: right front (fore) paw

FTIR (luminance)-based readouts

• Paw print size remains unchanged and measures the pixel area of the contact surface of each paw per time frame.
• Paw luminescence is now paw luminance, representing the sum of all pixel intensity signals a given paw induces in each time frame.
• Paw luminance is redefined as paw pressure index, calculated as the paw luminance divided by the paw print size. This metric indicates the “pressure” applied by a paw in each frame.