ArXiv最新文献

Cherenkov imaging enables real-time visualization of megavoltage X-ray or electron beam delivery to the patient during Radiation Therapy (RT). Bio-morphological features, such as vasculature, seen in these images are patient-specific signatures that can be used for verification of positioning and motion management that are essential to precise RT treatment. However until now, no concerted analysis of this biological feature-based tracking was utilized because of the slow speed and accuracy of conventional image processing for feature segmentation. This study demonstrated the first deep learning framework for such an application, achieving video frame rate processing. To address the challenge of limited annotation of these features in Cherenkov images, a transfer learning strategy was applied. A fundus photography dataset including 20,529 patch retina images with ground-truth vessel annotation was used to pre-train a ResNet segmentation framework. Subsequently, a small Cherenkov dataset (1,483 images from 212 treatment fractions of 19 breast cancer patients) with known annotated vasculature masks was used to fine-tune the model for accurate segmentation prediction. This deep learning framework achieved consistent and rapid segmentation of Cherenkov-imaged bio-morphological features on another 19 patients, including subcutaneous veins, scars, and pigmented skin. Average segmentation by the model achieved Dice score of 0.85 and required less than 0.7 milliseconds processing time per instance. The model demonstrated outstanding consistency against input image variances and speed compared to conventional manual segmentation methods, laying the foundation for online segmentation in real-time monitoring in a prospective setting.

Intracorporeal needle-based therapeutic ultrasound (NBTU) offers a minimally invasive approach for the thermal ablation of malignant brain tumors, including both primary and metastatic cancers. NBTU utilizes a high-frequency alternating electric field to excite a piezoelectric transducer, generating acoustic waves that cause localized heating and tumor cell ablation, and it provides a more precise ablation by delivering lower acoustic power doses directly to targeted tumors while sparing surrounding healthy tissue. Building on our previous work, this study introduces a database for optimizing pre-operative surgical planning by simulating ablation effects in varied tissue environments and develops an extended simulation model incorporating various tumor types and sizes to evaluate thermal damage under trans-tissue conditions. A comprehensive database is created from these simulations, detailing critical parameters such as CEM43 isodose maps, temperature changes, thermal dose areas, and maximum ablation distances for four directional probes. This database serves as a valuable resource for future studies, aiding in complex trajectory planning and parameter optimization for NBTU procedures. Moreover, a novel probe selection method is proposed to enhance pre-surgical planning, providing a strategic approach to selecting probes that maximize therapeutic efficiency and minimize ablation time. By avoiding unnecessary thermal propagation and optimizing probe angles, this method has the potential to improve patient outcomes and streamline surgical procedures. Overall, the findings of this study contribute significantly to the field of NBTU, offering a robust framework for enhancing treatment precision and efficacy in clinical settings.

Functional magnetic resonance imaging (fMRI) is a commonly used technique to measure neural activation. Its application has been particularly important in identifying underlying neurodegenerative conditions such as Parkinson's, Alzheimer's, and Autism. Recent analysis of fMRI data models the brain as a graph and extracts features by graph neural networks (GNNs). However, the unique characteristics of fMRI data require a special design of GNN. Tailoring GNN to generate effective and domain-explainable features remains challenging. In this paper, we propose a contrastive dual-attention block and a differentiable graph pooling method called ContrastPool to better utilize GNN for brain networks, meeting fMRI-specific requirements. We apply our method to 5 resting-state fMRI brain network datasets of 3 diseases and demonstrate its superiority over state-of-the-art baselines. Our case study confirms that the patterns extracted by our method match the domain knowledge in neuroscience literature, and disclose direct and interesting insights. Our contributions underscore the potential of ContrastPool for advancing the understanding of brain networks and neurodegenerative conditions. The source code is available at https://github.com/AngusMonroe/ContrastPool.

During development and under normal physiological conditions, biological tissues are continuously subjected to substantial mechanical stresses. In response to large deformations cells in a tissue must undergo multicellular rearrangements in order to maintain integrity and robustness. However, how these events are connected in time and space remains unknown. Here, using computational and theoretical modeling, we studied the mechanical plasticity of epithelial monolayers under large deformations. Our results demonstrate that the jamming-unjamming (solid-fluid) transition in tissues can vary significantly depending on the degree of deformation, implying that tissues are highly unconventional materials. Using analytical modeling, we elucidate the origins of this behavior. We also demonstrate how a tissue accommodates large deformations through a collective series of rearrangements, which behave similarly to avalanches in non-living materials. We find that these 'tissue avalanches' are governed by stress redistribution and the spatial distribution of vulnerable spots. Finally, we propose a simple and experimentally accessible framework to predict avalanches and infer tissue mechanical stress based on static images.

The ability of a brain or a neural network to efficiently learn depends crucially on both the task structure and the learning rule. Previous works have analyzed the dynamical equations describing learning in the relatively simplified context of the perceptron under assumptions of a student-teacher framework or a linearized output. While these assumptions have facilitated theoretical understanding, they have precluded a detailed understanding of the roles of the nonlinearity and input-data distribution in determining the learning dynamics, limiting the applicability of the theories to real biological or artificial neural networks. Here, we use a stochastic-process approach to derive flow equations describing learning, applying this framework to the case of a nonlinear perceptron performing binary classification. We characterize the effects of the learning rule (supervised or reinforcement learning, SL/RL) and input-data distribution on the perceptron's learning curve and the forgetting curve as subsequent tasks are learned. In particular, we find that the input-data noise differently affects the learning speed under SL vs. RL, as well as determines how quickly learning of a task is overwritten by subsequent learning. Additionally, we verify our approach with real data using the MNIST dataset. This approach points a way toward analyzing learning dynamics for more-complex circuit architectures.

Stochasticity plays a central role in nearly every biological process, and the noise power spectral density (PSD) is a critical tool for understanding variability and information processing in living systems. In steady-state, many such processes can be described by stochastic linear time-invariant (LTI) systems driven by Gaussian white noise, whose PSD is a complex rational function of the frequency that can be concisely expressed in terms of their Jacobian, dispersion, and diffusion matrices, fully defining the statistical properties of the system's dynamics at steady-state. Here, we arrive at compact element-wise solutions of the rational function coefficients for the auto- and cross-spectrum that enable the explicit analytical computation of the PSD in dimensions n=2,3,4. We further present a recursive Leverrier-Faddeev-type algorithm for the exact computation of the rational function coefficients. Crucially, both solutions are free of matrix inverses. We illustrate our element-wise and recursive solutions by considering the stochastic dynamics of neural systems models, namely Fitzhugh-Nagumo (n=2), Hindmarsh-Rose (n=3), Wilson-Cowan (n=4), and the Stabilized Supralinear Network (n=22), as well as an evolutionary game-theoretic model with mutations (n=5, 31). We extend our approach to derive a recursive method for calculating the coefficients in the power series expansion of the integrated covariance matrix for interacting spiking neurons modeled as Hawkes processes on arbitrary directed graphs.

Transcranial focused ultrasound stimulation (TUS) holds promise for non-invasive neural modulation in treating neurological disorders. Most clinically relevant targets are deep within the brain (near or at its geometric center), surrounded by other sensitive regions that need to be spared clinical intervention. However, in TUS, increasing frequency with the goal of improving spatial resolution reduces the effective penetration depth. We show that by using a pair of 1 MHz, orthogonally arranged transducers we improve the spatial resolution afforded by each of the transducers individually, by nearly 40 fold, achieving a sub-cubic millimeter target volume of $0.24 mm^3$. We show that orthogonally placed transducers generate highly localized standing waves with Acoustic Radiation Force (ARF) arranged into periodic regions of compression and tension near the target. We further present an extended capability of the orthogonal setup, which is to impart selective pressures--either positive or negative, but not both--on the target. Lastly, we share our preliminary findings that strain can arise from both particle motion and ARF with the former reaching its maximum value at the focus, and the latter remaining null at the focus and reaching its maximum around the focus. As the field is investigating the mechanism of interaction in TUS by way of elucidating the mapping between ultrasound parameters and neural response, orthogonal transducers expand our toolbox by making it possible to conduct these investigations at much finer spatial resolutions, with localized and directed (compression vs. tension) ARF and the capability of applying selective pressures at the target.

Intracorporeal needle-based therapeutic ultrasound (NBTU) is a minimally invasive option for intervening in malignant brain tumors, commonly used in thermal ablation procedures. This technique is suitable for both primary and metastatic cancers, utilizing a high-frequency alternating electric field (up to 10 MHz) to excite a piezoelectric transducer. The resulting rapid deformation of the transducer produces an acoustic wave that propagates through tissue, leading to localized high-temperature heating at the target tumor site and inducing rapid cell death. To optimize the design of NBTU transducers for thermal dose delivery during treatment, numerical modeling of the acoustic pressure field generated by the deforming piezoelectric transducer is frequently employed. The bioheat transfer process generated by the input pressure field is used to track the thermal propagation of the applicator over time. Magnetic resonance thermal imaging (MRTI) can be used to experimentally validate these models. Validation results using MRTI demonstrated the feasibility of this model, showing a consistent thermal propagation pattern. However, a thermal damage isodose map is more advantageous for evaluating therapeutic efficacy. To achieve a more accurate simulation based on the actual brain tissue environment, a new finite element method (FEM) simulation with enhanced damage evaluation capabilities was conducted. The results showed that the highest temperature and ablated volume differed between experimental and simulation results by 2.1884°C (3.71%) and 0.0631 cm³ (5.74%), respectively. The lowest Pearson correlation coefficient (PCC) for peak temperature was 0.7117, and the lowest Dice coefficient for the ablated area was 0.7021, indicating a good agreement in accuracy between simulation and experiment.

Characterizing judgments of similarity within a perceptual or semantic domain, and making inferences about the underlying structure of this domain from these judgments, has an increasingly important role in cognitive and systems neuroscience. We present a new framework for this purpose that makes limited assumptions about how perceptual distances are converted into similarity judgments. The approach starts from a dataset of empirical judgments of relative similarities: the fraction of times that a subject chooses one of two comparison stimuli to be more similar to a reference stimulus. These empirical judgments provide Bayesian estimates of underling choice probabilities. From these estimates, we derive indices that characterize the set of judgments in three ways: compatibility with a symmetric dis-similarity, compatibility with an ultrametric space, and compatibility with an additive tree. Each of the indices is derived from rank-order relationships among the choice probabilities that, as we show, are necessary and sufficient for local consistency with the three respective characteristics. We illustrate this approach with simulations and example psychophysical datasets of dis-similarity judgments in several visual domains and provide code that implements the analyses at https://github.com/jvlab/simrank.

Deep learning models have shown promise in histopathology image analysis, but their opaque decision-making process poses challenges in high-risk medical scenarios. Here we introduce HIPPO, an explainable AI method that interrogates attention-based multiple instance learning (ABMIL) models in computational pathology by generating counterfactual examples through tissue patch modifications in whole slide images. Applying HIPPO to ABMIL models trained to detect breast cancer metastasis reveals that they may overlook small tumors and can be misled by non-tumor tissue, while attention maps-widely used for interpretation-often highlight regions that do not directly influence predictions. By interpreting ABMIL models trained on a prognostic prediction task, HIPPO identified tissue areas with stronger prognostic effects than high-attention regions, which sometimes showed counterintuitive influences on risk scores. These findings demonstrate HIPPO's capacity for comprehensive model evaluation, bias detection, and quantitative hypothesis testing. HIPPO greatly expands the capabilities of explainable AI tools to assess the trustworthy and reliable development, deployment, and regulation of weakly-supervised models in computational pathology.