Synthetic data generation method improves risk prediction model for early tumor recurrence after surgery in patients with pancreatic cancer.

Pancreatic cancer is aggressive with high recurrence rates, necessitating accurate prediction models for effective treatment planning, particularly for neoadjuvant chemotherapy or upfront surgery. This study explores the use of variational autoencoder (VAE)-generated synthetic data to predict early tumor recurrence (within six months) in pancreatic cancer patients who underwent upfront surgery. Preoperative data of 158 patients between January 2021 and December 2022 was analyzed, and machine learning models-including Logistic Regression, Random Forest (RF), Gradient Boosting Machine (GBM), and Deep Neural Networks (DNN)-were trained on both original and synthetic datasets.

Prognosis of Breast Cancer in Women in Their 20s: Clinical and Radiological Insights.

We analyzed clinical and radiological characteristics and prognostic factors specific to young patients with breast cancer (YBC) aged <30 years. This retrospective study included 132 women aged <30 years who underwent breast surgery between 2008 and 2013. The clinical and radiological findings of the patients were examined and compared according to recurrence or death status at follow-up.

Local gyrification index and sulcal depth as imaging markers of cognitive decline in Alzheimer's disease.

Purpose: To investigate the correlation between cortical thickness (CT), sulcal depth (SD), local gyrification index (LGI), and cognitive scores in patients with Alzheimer's disease (AD).

Methods: A total of 200 patients with AD from 2014 to 2021 were included, confirmed by 18F-florbetaben-positron emission tomography, and having a Clinical Dementia Rating score of 0.5 or 1.

Decomposed Linear Dynamical Systems (dLDS) models reveal instantaneous, context-dependent dynamic connectivity in C. elegans.

Mounting evidence indicates that neural "tuning" can be highly variable within an individual across time and across individuals. Furthermore, modulatory effects can change the relationship between neurons as a function of behavioral or other conditions, meaning that the changes in activity (the derivative) may be as important as the activity itself. Current computational models cannot capture the nonstationarity and variability of neural coding, preventing the quantitative evaluation of these effects.

Multiscale comparative connectomics.

The connectome, a map of the structural and/or functional connections in the brain, provides a complex representation of the neurobiological phenotypes on which it supervenes. This information-rich data modality has the potential to transform our understanding of the relationship between patterns in brain connectivity and neurological processes, disorders, and diseases. However, existing computational techniques used to analyze connectomes are often insufficient for interrogating multi-subject connectomics datasets: many current methods are either solely designed to analyze single connectomes or leverage heuristic graph statistics that are unable to capture the complete topology of multiscale connections between brain regions.

Impact of an immediate short waiting period on ultrasound-based ablative margin assessment following radiofrequency ablation for hepatocellular carcinoma.

Purpose: This study aimed to evaluate whether an immediate short waiting period after radiofrequency ablation (RFA) can improve the accuracy of ultrasound (US)-based assessment of the ablation zone in patients with hepatocellular carcinoma (HCC).

Methods: A prospective cohort study was conducted involving 41 patients who underwent USguided RFA for HCC. Tumor margin conspicuity, electrode tip visibility, and operator confidence in assessing the ablative margin were recorded immediately following electrode deactivation and at 1-minute intervals for 5 minutes.

Gloss discrimination: Toward an image-based perceptual model.

Gloss is typically considered the perceptual counterpart of a surface's reflectance characteristics. Yet, asking how discriminable two surfaces are on the basis of surface properties is a poorly posed question, as scene factors other than reflectance can have substantial effects on how discriminable two glossy surfaces are to humans. This difficulty with predicting gloss discrimination has so far hobbled efforts to establish a perceptual standard for surface gloss.

Phenotype augmentation using generative AI for isocitrate dehydrogenase mutation prediction in glioma.

This study investigated the effects of feature augmentation, which uses generated images with specific imaging features, on the performance of isocitrate dehydrogenase (IDH) mutation prediction models in gliomas. A total of 598 patients were included from our institution (310 training, 152 internal test) and the Cancer Genome Atlas (136 external test). Score-based diffusion models were used to generate T2-weighted, FLAIR, and contrast-enhanced T1-weighted image triplets.

Brain-wide coordination of decision formation and commitment.

Neural correlates of a subject's upcoming choice in decision making tasks are remarkably widespread throughout the brain, but how these brain-wide signals are coordinated remains unknown. Do brain regions reflect competing or shared decision computations? Are decision signals generated by distributed circuits or are they computed locally and then broadcast broadly? To address these fundamental questions, we took a brain-wide approach, simultaneously recording from thousands of neurons across up to twenty brain regions in rats performing a decision making task requiring gradual accumulation of auditory evidence. This allowed decoding, on single trials and for each of the simultaneously recorded regions, the time-varying state of the decision as it is being formed.

Effect of Deep Learning-Based Artificial Intelligence on Radiologists' Performance in Identifying Nigrosome 1 Abnormalities on Susceptibility Map-Weighted Imaging.

Objective: To evaluate the effect of deep learning (DL)-based artificial intelligence (AI) software on the diagnostic performance of radiologists with different experience levels in detecting nigrosome 1 (N1) abnormalities on susceptibility map-weighted imaging (SMwI).

Materials And Methods: This retrospective diagnostic case-control study analyzed 139 SMwI scans of 59 patients with Parkinson's disease (PD) and 80 healthy participants. Participants were imaged using 3T MRI, and AI-generated assessments for N1 abnormalities were obtained using an AI model (version 1.

Deep-Learning Model for Real-Time Prediction of Recurrence in Early-Stage Non-Small Cell Lung Cancer: A Multimodal Approach (RADAR CARE Study).

Purpose: The surveillance protocol for early-stage non-small cell lung cancer (NSCLC) is not contingent upon individualized risk factors for recurrence. This study aimed to use comprehensive data from clinical practice to develop a deep-learning model for practical longitudinal monitoring.

Methods: A multimodal deep-learning model with transformers was developed for real-time recurrence prediction using baseline clinical, pathological, and molecular data with longitudinal laboratory and radiologic data collected during surveillance.

SLAy-ing oversplitting errors in high-density electrophysiology spike sorting.

The growing channel count of silicon probes has substantially increased the number of neurons recorded in electrophysiology (ephys) experiments, rendering traditional manual spike sorting impractical. Instead, modern ephys recordings are processed with automated methods that use waveform template matching to isolate putative single neurons. While scalable, automated methods are subject to assumptions that often fail to account for biophysical changes in action potential waveforms, leading to systematic errors.

Library of realistic 4D digital beating heart models based on patient CT data.

Background: The prevalence of cardiovascular disease (CVD) has risen alongside new medical imaging technologies designed for its diagnosis and treatment. Computational phantoms play a crucial role in imaging research, supporting applications ranging from basic simulation studies to larger-scale virtual imaging trials (VITs).

Purpose: In this work, we develop a population of detailed, anatomically variable 4D beating heart models for medical imaging research.

Reduced Vestibular Function is Associated With Cortical Surface Shape Changes in the Frontal Cortex.

Aging-associated decline in peripheral vestibular function is linked to deficits in behaviors and cognitive abilities that are known to rely on the sensorimotor and frontal cortices, but the precise neural pathways are unknown. To fill this knowledge gap, this cross-sectional study investigates the relationship between age-related variation in vestibular function and surface shape alterations of the frontal and sensorimotor cortices, considering age, intracranial volume, and sex. Data from 117 older adults (aged 60+) from the Baltimore Longitudinal Study of Aging, who underwent end-organ-specific vestibular tests (cVEMP for the saccule, oVEMP for the utricle, and vHIT for the horizontal canal) and T1-weighted MRI scans on the same visit, were analyzed.

Impact of LI-RADS CT and MRI Ancillary Features on Diagnostic Performance: An Individual Participant Data Meta-Analysis.

Background A recent meta-analysis showed independent associations between most Liver Imaging Reporting and Data System (LI-RADS) ancillary features (AFs) and hepatocellular carcinoma (HCC), malignancy, and benignity. However, the impact of AFs on the diagnostic performance of LI-RADS remains unclear. Purpose To evaluate the impact of applying individual AFs on the diagnostic performance of CT and MRI LI-RADS using an individual participant data (IPD) meta-analysis.

Pattern-based volumetric CT quantification to predict radiation pneumonitis in patients with non-small-cell lung cancer who have diffuse parenchymal lung disease.

Background: Diffuse parenchymal lung disease (DPLD) is a well-known risk factor for radiation pneumonitis (RP) after radiation therapy (RT) for lung cancer. However, it is hard to evaluate the exact extent of DPLD and to predict RP. This study sought to quantify the extent of DPLD and to determine which pattern(s) of DPLD lead to RP using texture analysis of pre-treatment computed tomography (CT) scans.

Diagnostic Performance of CT/MRI LI-RADS Version 2018 Major Feature Combinations: Individual Participant Data Meta-Analysis.

Background The CT/MRI Liver Imaging Reporting and Data System (LI-RADS) diagnostic algorithm classifies liver observations in patients with high-risk hepatocellular carcinoma (HCC) using imaging features. However, data regarding the diagnostic performance of specific LI-RADS major feature combinations is limited. Purpose To conduct a systematic review and individual participant data (IPD) meta-analysis to establish the positive predictive values (PPVs) of LI-RADS major feature combinations using CT/MRI LI-RADS version 2018 in patients at risk for HCC.

Comparison of Gadoxetic Acid-Enhanced Liver Magnetic Resonance Imaging and Contrast-Enhanced Computed Tomography for the Noninvasive Diagnosis of Hepatocellular Carcinoma.

Introduction: Magnetic resonance imaging (MRI) has been shown to outperform computed tomography (CT) in diagnosing hepatocellular carcinoma (HCC), although inconsistencies exist across studies. We compared the performance of CT and gadoxetic acid-enhanced MRI in diagnosing HCC according to various guidelines, and to assess the incremental value of a second-line examination.

Methods: This retrospective multicenter study included patients at risk of developing HCC with focal liver lesions (FLLs) ≥10 mm.

Leveraging deep learning-based kernel conversion for more precise airway quantification on CT.

Objectives: To evaluate the variability of fully automated airway quantitative CT (QCT) measures caused by different kernels and the effect of kernel conversion.

Materials And Methods: This retrospective study included 96 patients who underwent non-enhanced chest CT at two centers. CT scans were reconstructed using four kernels (medium soft, medium sharp, sharp, very sharp) from three vendors.

Landsat Surface Product Validation Instrumentation: The BigMAC Exercise.

Users of remotely sensed Earth optical imagery are increasingly demanding a surface reflectance or surface temperature product instead of the top-of-atmosphere products that have been produced historically. Validating the accuracy of surface products remains a difficult task since it involves assessment across a range of atmospheric profiles, as well as many different land surface types. Thus, the standard approaches from the satellite calibration community do not apply, and new technologies need to be developed.

MD-SA2: optimizing Segment Anything 2 for multimodal, depth-aware brain tumor segmentation in sub-Saharan populations.

Purpose: Machine learning algorithms are emerging as valuable aides for radiologists in medical image segmentation due to their accuracy and speed. However, existing approaches, including both conventional machine learning and Segment Anything (SA)-based models, face challenges with the complex, multimodal, and varied quality of magnetic resonance imaging (MRI) scan images used for brain tumor segmentation. To address these challenges, we propose MD-SA2, adapting Segment Anything 2 (SA2) to medical image segmentation and introducing a lightweight U-Net "aggregator" model.

Application of LI-RADS CT/MRI Radiation Treatment Response Assessment Version 2024: a study after transarterial radioembolization for hepatocellular carcinoma.

Purpose: To compare the performance of the LI-RADS CT/MRI Radiation Treatment Response Assessment (TRA) version 2024 (v2024) after transarterial radioembolization (TARE) for hepatocellular carcinoma (HCC) with that of the LI-RADS CT/MRI TRA version 2017 (v2017).

Methods: This retrospective study included patients with HCC treated with TARE followed by hepatic surgery between November 2012 and April 2023 at two tertiary referral centers. Each treated lesion was assigned an LI-RADS treatment response (LR-TR) category based on a consensus reading of three radiologists using both v2024 and v2017.

Impact of portal/superior mesenteric vein abutment angle on prognosis in pancreatic cancer: a single-center retrospective cohort study.

Purpose: Pancreatic cancer has a poor prognosis; however, the implementation of neoadjuvant treatment enables borderline resectable cases to undergo curative resection and improves the overall survival rate. Attempts have been made to expand the eligibility criteria for neoadjuvant treatment, even in resectable cases. Some studies have suggested a correlation between vein abutment and poor prognosis or that the abutment angle may affect prognosis.