Research on the impairment of episodic memory in AD usually treats all memory contents equally. However, recent studies have revealed that certain images are more diagnostic than other images in classifying whether a person may have AD or its precursors (e.g., mild cognitive impairment, MCI), highlighting the potential insights we may gain when looking at differences across memory content. In this study, we explore whether a set of potentially diagnostic images, for which healthy controls have much higher hit rates than MCI participants, could reflect abnormalities in AD biomarker status and brain processing better than other images. 64 healthy controls and 48 MCI participants in the DELCODE dataset performed a visual scene memory task during fMRI, with CSF AD biomarker data collected (i.e., amyloid and tau biomarkers). We tested whether diagnostic versus non-diagnostic images showed a stronger relationship between behavior and CSF biomarker status, and showed differences in visual patterns during fMRI. We found that diagnostic images have larger behavior-biomarker correlations for total tau, phospho-tau, Aβ42/Aβ40, Aβ42/phospho-tau compared to non-diagnostic images, suggesting that memory for these specific images are more affected by AD pathology. The fMRI data revealed an interaction effect between group membership (healthy control / MCI) and image diagnosticity (diagnostic / non-diagnostic scene images), with MCI participants having higher activation in scene processing regions (PPA, MPA/RSC and OPA) for diagnostic images than non-diagnostic images. In contrast, healthy controls showed no differences in processing between diagnostic and non-diagnostic images. These results suggest that MCI individuals may engage in heightened compensatory activation for these diagnostic images. Our results suggest that diagnostic images are particularly effective at revealing brain pathologies, highlighting direct clinical applications and potential implications for our broader understanding of episodic memory.