Whole Brain Histogram Analysis for Amyloid PET
A semi‑quantitative analysis method for amyloid PET that does not require normalization
Embedded Files
Introduction
Introduction
This webpage provides a detailed overview of the Whole-Brain Histogram Analysis (WBHA) method originally published in the Annals of Nuclear Medicine. Developed by Okuyama et al., WBHA represents a novel approach to brain PET image analysis. The methodology featured on this page introduces a newly advanced, MRI-based WBHA technique that builds directly upon the foundational work of Okuyama et al. Utilizing MRI allows for high-precision brain segmentation, which facilitates the calculation of standard quantitative measures of WBHA—specifically skewness and the mode-to-mean ratio (MMR)—alongside a novel quantitative metric: the Top 20% GW-ratio. When referencing this page for research presentations or academic papers, please cite both the original method by Okuyama et al. https://doi.org/10.1007/s12149-024-01956-y and our paper https://doi.org/10.1007/s12149-026-02218-9 .
What is WBHA?
What is WBHA?
Whole-brain histogram analysis (WBHA) was developed for the evaluation of amyloid PET scans. This technique extracts brain parenchyma from PET images and generates a histogram based on all voxel values. From this distribution, two indices are calculated: the mode-to-mean ratio (MMR), defined as the ratio of the mode to the mean, and skewness, which reflects the degree of deviation from a normal distribution. This method offers advantages because it eliminates the need for normalization, allows measurements in an individual’s native space, and does not require defining reference or target brain regions.
WBHA Process Quick Overview
WBHA Process Quick Overview
①Brain Extraction
Brain parenchyma is extracted using structural MRI
②Make Histogram
Voxel values are extracted from the brain parenchyma and used to generate a histogram.
③Calculation Skewness & MMR
Calculate the skewness and mode-to-mean ratio from the histogram.
- Skewness
- Mean
- Mode
- Mode-to-mean ratio (MMR)
④Make Top 20% Map and calucuate Top 20% GW-ratio
What is Top 20% Map and Top 20% GW-ratio？
What is Top 20% Map and Top 20% GW-ratio？
Top 20% Map
Top 20% Map
The Top 20% Map display method highlights regions with high tracer accumulation that occupy 20% of the total brain parenchymal volume, thereby supporting radiographic interpretation.
Top 20% GW-ratio
Top 20% GW-ratio
Top 20% maps were subsequently segmented into Gray matter (GM) and white matter (WM) components using tissue probability maps, applying a threshold of 0.5. Voxels within the separated Top 20% GM and WM maps were then counted. The Top 20% GW‑ratio was calculated using the following formula:
Analysis Example
Analysis Example
Negative example
Negative example
Positive example
Positive example
Diagnostic Performance of the WBHA Compared to the Centiloid
Diagnostic Performance of the WBHA Compared to the Centiloid
AUC analysis for predicting visual analysis results. This study was conducted using PADNI (https://padni.org/) (https://center6.umin.ac.jp/cgi-open-bin/ctr_e/ctr_his_list.cgi?recptno=R000041332)data.
Prediction Visual Analysis result
Prediction Visual Analysis result
When compared with the Centiloid scale, the WBHA parameter shows a lower AUC, and although the difference is only marginal, it remains statistically significant.
Prediction no Cognitive Impairment
Prediction no Cognitive Impairment
Compared to the Centiloid scale, the AUC of the GW‑ratio was higher but not statistically significant.
Correlation between the Centiloid scale and the WBHA
Correlation between the Centiloid scale and the WBHA
A strong correlation was observed between the Centiloid scale and MMR, as well as between the Centiloid scale and the GW‑ratio.
Caution: ¹⁸F‑Flutemetamol results. This conversion equation cannot be applied to other tracers.
Normal range of WBHA
Normal range of WBHA
From the cognitively unimpaired (G‑CDR = 0, MMSE ≥ 28) and visually Aβ‑negative group, the mean, SD, and 95% reference ranges for CL, skewness, MMR, and the Top 20% GW‑ratio were calculated as follows.
Caution: ¹⁸F‑Flutemetamol results. This normal range cannot be applied to other tracers.
Conclusion
Conclusion
In conclusion, we successfully validated WBHA for the Centiloid scale. The parameters correlate strongly with the Centiloid scale, and the diagnostic performance is comparable to established methods. Because it is independent of anatomical templates, WBHA provides a robust, personalized metric for amyloid PET quantification. WBHA is considered a useful alternative when the Centiloid scale cannot be measured accurately for any reason or when there are concerns about its reliability.
Page updated
Google Sites
Report abuse