19 To our knowledge, the current study is the first to employ high-dimensional pattern recognition techniques to assess ABA patterns and to determine similarities and differences with clinical AD patterns in a large population-based cohort spanning almost the entire adulthood age range.
MRI pattern classification A high-dimensional pattern classification method was previously proposed to calculate SPARE-AD,16, 33 an index derived from the imaging data of cognitively normal older adults and clinical AD patients,18 to quantify atrophy patterns associated with AD. The pattern classifier was constructed to maximally differentiate between these two groups using a support vector machine.
Here is the full scientific article if you wish to download it.
ABA showed a spatial pattern that deviated notably from previously reported spatial patterns of AD-related gray matter atrophy,6, 14 as well as from AD-related atrophy patterns we identified by comparing individuals in age-adjusted high and low SPARE-AD groups in the SHIP data.
Blue: regions displaying significant regional atrophy patterns between resilient and ABA individuals; Red: regions displaying significantly AD-related patterns of atrophy in high vs low SPARE-AD individuals; Green: overlap of the blue … While, within the blue and green regions the density distributions of RAVENS values show a clear shift between ABA and RA individuals-indicating significant volume reduction-the two groups did not have such a shift for the density distributions within the red region, indicating that the correlation between SPARE-BA and SPARE-AD is largely driven by the spatial overlap of respective patterns.
Contribution of high-dimensional pattern classification techniques An important contribution of our study is the use of advanced methods of high dimensional pattern classification for BA assessment, which allowed us to investigate in detail the spatial patterns of atrophy, and to derive individualized indices that were further correlated with epidemiologic and clinical factors.
Recently, Janowitz et al.36 analyzed prediction patterns for hippocampal volumes in SHIP. Interestingly the associated risk factors with hippocampal volume were similar to those associated with the aging patterns in the current study but different from the prediction patterns of SPARE-AD. This finding indicates that the hippocampus alone is unlikely to adequately reflect the complexity of neurodegeneration in AD, as previously demonstrated in the study by Fan et al.15 Overlap between ABA and AD spatial patterns of atrophy Figure 4 shows that the regions of the ABA-related spatial patterns of atrophy overlapped only partially with the clinical AD-related patterns.
In summary, the current study is the first, to our knowledge, to employ high-dimensional pattern recognition techniques to assess BA patterns in a cohort of this size and show that it has a unique spatial pattern of brain atrophy that differs from the one found in AD. Go to: Acknowledgments SHIP is part of the Community Medicine Research net of the University of Greifswald, Germany, which is funded by the Federal Ministry of Education and Research, the Ministry of Cultural Affairs and the Social Ministry of the Federal State of Mecklenburg-West Pomerania.