PMC

Primary measures of performance accuracy. Absolute distance error is defined as the Euclidean distance between the participant’s estimate of location one (goal) and the actual location of cone 1.

The effect of return condition within participant groups. The effect of return condition on absolute distance error averaged per participant in each group. Return conditions: green = no environmental change; blue = removal of distal boundary cues; red = removal of surface detail; open circle = mean; black line = median.

Scatterplots of path integration performance and region of interest volumetry. The relationship absolute distance error and regions of interest including; EC (A), alEC (B), pmEC (C), whole hippocampus (D), PCC (E) and isthus of the cingulate cortex (F) was assessed. Least square lines are group specific: grey = across MCI+and MCI−; purple = all participants), *P < 0.005 (Bonferroni adjusted α).

Graph summarizing the between group differences in path integration performance. Absolute distance error (Euclidean distance) error in metres (A and B). (A) Group comparison between healthy controls and total MCI and (B) between MCI− and MCI+. Each marker represents the mean performance across trials of each individual: blue circles = healthy control subjects; black asterisks = MCI without biomarkers; red triangles = MCI+; green inverted triangles = MCI−; central grey line = mean; dark grey inner box = 95% CIs; light grey outer box = 1 standard deviation. *P < 0.05, ***P < 0.001.

Protocol for the segmentation of the whole entorhinal cortex and partial segmentation of its anteriolateral and posterioromedial subdivisions. Anteriolateral EC (alEC, green) is segmented two slices anterior to the emergence of the hippocampal head (slice 3). Posteromedial (pmEC, pink) is segmented from one slice anterior to, and one slice posterior from, the uncal apex (slice 10). This partial approach for EC subdivisions does not encompasses the full extent of the EC but rather reflects the anterior and posterior extremes of the EC that avoid segmenting the progressive boundary in the absence of consistent landmarks. Correspondingly, all intermediate slices between alEC and pmEC are segmented as EC (brown), total EC volume is produced by summing all three EC subdivision volumes. Arrow schematic indicates anatomical plane for 3D segmentation.

ROC plot. Accuracy of path integration task performance for classifying (A) total MCI from healthy control subjects and (B) MCI+ from MCI− patients. Path integration performance is represented by absolute distance error (solid red line). Classification of reference cognitive tests is represented by dashed lines for comparison. ACE-R (grey), Trail Making Test B (green), 4MT (yellow), Free and Cued Selective Reminding Test – delayed free recall (blue) and Rey Figure Recall – delayed recall (purple). Asterisk indicates optimal operating point for absolute distance error.

Path integration task. (A) Illustration of the path integration task. Each numbered inverted blue cone is a location marker. Only one cone was visible at a time; upon reaching a blue cone it disappeared and the next one in the sequence appeared. Red arrows indicate the guided sequence along two sides of the triangle. The yellow arrow, the last side of the triangle, signifies the assessed return path, performed in the absence of any cones. (B) Demonstration of VR equipment on a participant during the task, used with permission. (C) Example environment from the head mounted display with textural and boundary cues present, with cone 1 and the controller shown. Texture and boundary cues are present in all trials when navigating between cones. (D–F) Return conditions applied when attempting to return to the location of cone 1 only (yellow arrow, A) and included no change (D), removal of environment boundaries (E) and removal of surface detail (F).