We present a compact, learning-based representation that captures the full Monte Carlo sampling distribution of a rendered image. Our approach enables rendering at arbitrary samples per pixel (SPP) during inference without requiring expensive path tracing operations. This is achieved by fitting parametric distributions to per-pixel radiance values, which can be efficiently estimated, stored, and sampled. Our method proceeds in three stages. First, we map radiance samples into radial log space, which encourages Gaussian-like distributions while preserving angular relationships. Second, we fit each pixel’s distribution using 3D Gaussian Mixture Models (GMMs), trained online with minimal memory overhead, making the approach compatible with standard path tracers. For inference, we introduce an optimized sampling scheme whose complexity is independent of the target SPP, enabling fast synthesis of high-SPP images. Additionally, we demonstrate that the learned representations can be heavily compressed using quantization and codebook techniques with negligible quality loss. Experiments show that GMMs strike an effective balance between expressiveness and sparsity. Compared to alternative models, our method better captures pixel-wise Monte Carlo distributions. Lastly, we illustrate the versatility of our representation with applications such as firefly rejection and ray-distribution-driven denoising.

Machine learning models have significantly enriched the toolbox in the field of neuroimaging analysis. Among them, Support Vector Machines (SVM) have been one of the most popular models for supervised learning, but their use primarily relies on linear SVM models due to their explainability. Kernel SVM models are capable classifiers but more opaque. Recent advances in eXplainable AI (XAI) have developed several feature importance methods to address the explainability problem. However, noise variables can affect these explanations, making irrelevant variables regarded as important variables. This problem also appears in explaining linear models, which the linear pattern can address. This paper proposes a fast method to explain RBF kernel SVM globally by adopting the notion of linear pattern in kernel space. Our method can generate global explanations with low computational cost and is less affected by noise variables. We successfully evaluate our method on simulated and real MEG/EEG datasets.

Machine learning methods have shown great performance in many areas, including neuroimaging data analysis. However, model performance is only one objective in neuroimaging analysis. Gaining insight from the data is also critical in this field, such as identifying regions where detected signals are relevant to cognitive and diagnostic tasks. To fulfil this need, enabling the explainability of a model’s decision-making process is critical. Predictions of complex machine learning models are notoriously difficult to explain. This limits the use of complex models like kernel support vector machines (SVM) in neuroimaging analysis. Recently, several permutation-based methods have been developed to explain these complex models. However, the explanation results are affected by class-irrelevant features like suppressor variables and high background noise variables. This problem may also happen when explaining linear models. One possible reason is that the permutation process will produce unrealistic data instances when features are not independent, e.g. correlated. These unrealistic data instances will influence the explanation results. In neuroimaging analysis, the activation pattern, the estimated weight of the assumed generative model corresponding to the current classifier, is used to deal with this problem for linear models. This method does not rely on a permutation process but rather on the available data information. In this paper, we propose a novel method of Explanation through Activation Pattern (EAP) to explain the SVM models with different types of kernels for neuroimaging data analysis. Our method can generate a global feature importance score by estimating the activation pattern of kernel SVM models. We evaluate our method against three popular methods on both simulation datasets and a publicly available EEG/MEG dataset on visual tasks. The experimental results demonstrate that the proposed EAP method can provide explanations with low computational cost and is less affected by class-irrelevant features than the other three methods. In the experiment using the MEG/EEG dataset of visual tasks, the proposed EAP method provides agreement results with the brain’s electrical activity patterns reported in the literature on the visual tasks EEG/MEG data and is significantly faster than the other explanation methods.

In neuroimaging, the difference between predicted brain age and chronological age, known as brain age delta, has shown its potential as a biomarker related to various pathological phenotypes. There is a frequently observed bias when estimating brain age delta using regression models. This bias manifests as an overestimation of brain age for young participants and an underestimation of brain age for older participants. Therefore, the brain age delta is negatively correlated with chronological age, which can be problematic when evaluating relationships between brain age delta and other age-associated variables. This paper proposes a novel bias correction method for regression models by introducing a skewed loss function to replace the normal symmetric loss function. The regression model then behaves differently depending on whether it makes overestimations or underestimations. Our approach works with any type of MR image and no specific preprocessing is required, as long as the image is sensitive to age-related changes. The proposed approach has been validated using three classic deep learning models, namely ResNet, VGG, and GoogleNet on publicly available neuroimaging aging datasets. It shows flexibility across different model architectures and different choices of hyperparameters. The corrected brain age delta from our approach then has no linear relationship with chronological age and achieves higher predictive accuracy than a commonly-used two-stage approach.

An approach to the direct measurement of perception of video quality change using electroencephalography (EEG) is presented. Subjects viewed 8-s video clips while their brain activity was registered using EEG. The video signal was either uncompressed at full length or changed from uncompressed to a lower quality level at a random time point. The distortions were introduced by a hybrid video codec. Subjects had to indicate whether they had perceived a quality change. In response to a quality change, a positive voltage change in EEG (the so-called P3 component) was observed at latency of about 400-600 ms for all subjects. The voltage change positively correlated with the magnitude of the video quality change, substantiating the P3 component as a graded neural index of the perception of video quality change within the presented paradigm. By applying machine learning techniques, we could classify on a single-trial basis whether a subject perceived a quality change. Interestingly, some video clips wherein changes were missed (i.e., not reported) by the subject were classified as quality changes, suggesting that the brain detected a change, although the subject did not press a button. In conclusion, abrupt changes of video quality give rise to specific components in the EEG that can be detected on a single-trial basis. Potentially, a neurotechnological approach to video assessment could lead to a more objective quantification of quality change detection, overcoming the limitations of subjective approaches (such as subjective bias and the requirement of an overt response). Furthermore, it allows for real-time applications wherein the brain response to a video clip is monitored while it is being viewed.

Covert attention shifts to the visual periphery induce modulations of α-bandpower over occipital cortex. By demonstrating robust classification of covert attention shifts to four different target locations, a recent magnetoencephalography (MEG) study set the first step for its use as a new input modality to brain-computer interfaces (BCIs) [1]. Here, we set the next step by investigating its feasibility using electroencephalography (EEG). Eight healthy participants had to shift covert visual attention to one of six different target locations while strictly fixating the center of the screen. To enhance the spatial resolution, we used a current source density (CSD) estimate instead of raw voltage maps. Covert attention shifts to the different target locations yielded distinctive topographical distributions of posterior alpha activity.