This paper revises the activation likelihood estimation (ALE) algorithm for coordinate-based meta-analysis of neuroimaging data. The original ALE algorithm uses a permutation test to estimate the null-distribution of spatial associations between experiments, which is time-consuming and may underestimate the right tail of the distribution. The revised algorithm replaces this with a faster and more precise analytical method for estimating the null-distribution. Additionally, the previous false discovery rate (FDR) correction is supplemented with new approaches for correcting family-wise error rate and cluster-level significance. The revised algorithm is evaluated on an example dataset of face perception and compared with other methods for multiple-comparison correction. The revised ALE algorithm computes the null-distribution using histogram integration rather than permutation testing. This approach is more efficient and provides a more accurate estimation of the null-distribution. The algorithm also includes a new method for correcting family-wise error rate and cluster-level significance. The revised algorithm is compared with other methods for multiple-comparison correction, including FDR, family-wise error rate (FWE), and cluster-level inference. The results show that the revised ALE algorithm provides more accurate and efficient inference compared to the original algorithm. The revised ALE algorithm is applied to a dataset of face perception, where 19 papers reporting 20 individual experiments (305 subjects) and 183 activation foci were analyzed. The results show that the revised algorithm provides more accurate and efficient inference compared to the original algorithm. The results are compared with other methods for multiple-comparison correction, including FDR, FWE, and cluster-level inference. The results show that the revised ALE algorithm provides more accurate and efficient inference compared to the original algorithm. The revised algorithm is also compared with other methods for multiple-comparison correction, including FDR, FWE, and cluster-level inference. The results show that the revised ALE algorithm provides more accurate and efficient inference compared to the original algorithm.This paper revises the activation likelihood estimation (ALE) algorithm for coordinate-based meta-analysis of neuroimaging data. The original ALE algorithm uses a permutation test to estimate the null-distribution of spatial associations between experiments, which is time-consuming and may underestimate the right tail of the distribution. The revised algorithm replaces this with a faster and more precise analytical method for estimating the null-distribution. Additionally, the previous false discovery rate (FDR) correction is supplemented with new approaches for correcting family-wise error rate and cluster-level significance. The revised algorithm is evaluated on an example dataset of face perception and compared with other methods for multiple-comparison correction. The revised ALE algorithm computes the null-distribution using histogram integration rather than permutation testing. This approach is more efficient and provides a more accurate estimation of the null-distribution. The algorithm also includes a new method for correcting family-wise error rate and cluster-level significance. The revised algorithm is compared with other methods for multiple-comparison correction, including FDR, family-wise error rate (FWE), and cluster-level inference. The results show that the revised ALE algorithm provides more accurate and efficient inference compared to the original algorithm. The revised ALE algorithm is applied to a dataset of face perception, where 19 papers reporting 20 individual experiments (305 subjects) and 183 activation foci were analyzed. The results show that the revised algorithm provides more accurate and efficient inference compared to the original algorithm. The results are compared with other methods for multiple-comparison correction, including FDR, FWE, and cluster-level inference. The results show that the revised ALE algorithm provides more accurate and efficient inference compared to the original algorithm. The revised algorithm is also compared with other methods for multiple-comparison correction, including FDR, FWE, and cluster-level inference. The results show that the revised ALE algorithm provides more accurate and efficient inference compared to the original algorithm.