Understanding Dmek Graft Unfolding: Shifting The Focus From Ladders To Chutes

To analyze DMEK graft unfolding using a geometry-based hierarchy and identify destabilizing transitions (“chutes”) that increase graft complexity following specific intraoperative maneuvers.

Retrospective video-based analysis of recorded DMEK procedures.

Sixty discrete unfolding events were extracted from recorded DMEK procedures. Graft conformations were classified a priori into hierarchical groups (A–C) with defined subgroups (A1–C3), ordered by increasing geometric complexity. For each event, the transition between consecutive graft configurations was classified within this hierarchy. Transitions that increased geometric complexity (“chutes”) were categorized as Type I (no subgroup change), Type II (change within the same group), Type III (escalation across one group, e.g., A→B), or Type IV (escalation across two groups, e.g., A→C). For each chute, a primary cause category was assigned based on the maneuver preceding the transition.

Twenty “chutes” were identified. Type I chutes accounted for 50%, while escalation events accounted for 45% (Type III–IV), including major escalations (Type IV: 10%). Most chutes originated from low-to-intermediate complexity states (A/B: 95%) rather than from the highest complexity group. The most frequent cause category was uncontrolled BSS jet (50%), which was primarily observed in escalation chutes, whereas an excessively deep anterior chamber (30%) was most often observed in persistent configurations (Type I). Tapping misapplication accounted for 20%.

DMEK unfolding can be understood as a geometry-based hierarchy in which chutes represent maneuver-related errors that increase graft complexity and recovery difficulty. Nearly half of observed chutes involved escalation across groups, and most occurred while the graft remained in A/B states, highlighting vulnerability early in the unfolding pathway. Uncontrolled BSS jet was the most frequent driver of escalation, whereas an excessively deep anterior chamber was more often associated with persistent configurations. This “chutes” framework links graft morphology to recognizable and potentially avoidable error patterns and may help guide intraoperative decision-making and surgical training.