Anomaly detection and localization in medical imaging remain critical challenges in healthcare. This paper introduces Spatial-MSMA (Multiscale Score Matching Analysis), a novel unsupervised method for anomaly localization in volumetric brain MRIs. Building upon the MSMA framework, our approach incorporates spatial information and conditional likelihoods to enhance anomaly detection capabilities. We employ a flexible normalizing flow model conditioned on patch positions and global image features to estimate patch-wise anomaly scores. The method is evaluated on a dataset of 1,650 T1- and T2-weighted brain MRIs from typically developing children, with simulated lesions added to the test set. Spatial-MSMA significantly outperforms existing methods, including reconstruction-based, generative-based, and interpretation-based approaches, in lesion detection and segmentation tasks. Our model achieves superior performance in both distance-based metrics (99th percentile Hausdorff Distance: 7.05 ±0.61, Mean Surface Distance: 2.10 ±0.43) and component-wise metrics (True Positive Rate: 0.83 ±0.01, Positive Predictive Value: 0.96 ±0.01). These results demonstrate Spatial-MSMA's potential for accurate and interpretable anomaly localization in medical imaging, with implications for improved diagnosis and treatment planning in clinical settings. Our code is available at~https://github.com/ahsanMah/sade/.