Biology subjects

Jaywant, A.

Publications and source records attributed to Jaywant, A..

5 recordsLinked to original sources

bioRxiv2026

CALM-VLM: CALIBRATION AND SELECTIVE PREDICTION IN VISION-LANGUAGE MODELS FOR RELIABLE BRAIN MRI CLASSIFICATION

AO_SCPLOWBSTRACTC_SCPLOWRecent advances in vision-language models (VLMs) have demonstrated strong multimodal capabilities for medical image analysis. However, their confidence in diagnostic predictions is often unclear, limiting adoption in clinical settings. We introduce CALM-VLM (CALibration Mechanism for Vision-Language Models), which integrates confidence calibration and selective prediction into a generative 3D VLM. To create CALM-VLM, we fine-tuned the Med3DVLM architecture for Alzheimers disease (AD) and stroke classification as initial test cases. To improve reliability, we incorporated temperature scaling on the VLMs generative outputs. The calibrated model then selectively abstained from predictions when uncertain; this also improved its diagnostic accuracy. Experiments across multi-site MRI datasets, from 10 countries worldwide, show that CALM-VLM improved confidence relative to uncalibrated VLMs. Coverage-adjusted test receiver-operator characteristic curve-area under the curve (ROC-AUC) increased by 5% to 13% for both diagnostic tasks across independent test sets. Our calibrated VLM achieved a test ROC-AUC of 0.951 for AD classification and 0.905 for stroke classification. These findings highlight the importance of calibrated, uncertainty-aware VLMs for trustworthy neuroimaging AI.

neuroscience↗

bioRxiv2025

Predicting cognition using estimated structural and functional connectivity networks and artificial intelligence in multiple sclerosis

BackgroundOur prior work demonstrated that estimated structural and functional connectomes (eSC and eFC) generated using multiple sclerosis (MS) lesion masks and artificial intelligence (AI) models can predict disability as effectively as SC and FC derived from diffusion and functional MRI in MS. The goal of this study was to assess the ability of eSC and eFC in predicting baseline and 4-year follow-up cognition in MS patients. MethodsOne hundred seventy-one MS patients (age: 42.67{+/-}10.41, 74% females) were included. The Symbol Digit Modalities Test (SDMT), California Verbal Learning Test (CVLT), and Brief Visuospatial Memory Test (BVMT) were used to assess cognition. The Network Modification tool was performed to estimate SC, which was then used as an input to Krakencoder, an encoder-decoder model, to estimate FC. Ridge regression was performed to predict cognition using regional eSC and eFC, along with demographics and clinical information as well as conventional MRI metrics. Baseline cognition was added to the models that were used to predict the follow-up cognition. Spearmans correlation (r) was used to assess the prediction accuracy. ResultsThe highest accuracy was obtained when predicting follow-up SDMT using regional eSC or eFC (median r=0.58 for eSC and r=0.56 for eFC). Decreased eSC and eFC in the cerebellum and increased eFC in the default mode network were associated with lower follow-up SDMT scores. Baseline SDMT, clinical subtype, and age were the most important non-connectome metrics in predicting follow-up SDMT. ConclusionsOur findings demonstrate that eSC and eFC derived from clinically acquired MRI and AI models can effectively predict cognition. The use of lesion-based estimates of connectome disruptions may potentially improve cognition-related individualized treatment planning.

neuroscience↗

bioRxiv2021

Shared functional connections within and between cortical networks predict individual cognitive abilities in males and females

A thorough understanding of sex-independent and sex-specific neurobiological features that underlie cognitive abilities in healthy individuals is essential for the study of neurological illnesses in which males and females differentially experience and exhibit cognitive impairment. Here, we evaluate sex-independent and sex-specific relationships between functional connectivity and individual cognitive abilities in 392 healthy young adults (196 males) from the Human Connectome Project. First, we establish that sex-independent models comparably predict crystallised abilities in males and females, but more accurately predict fluid abilities in males. Second, we demonstrate sex-specific models comparably predict crystallised abilities within and between sexes, and generally fail to predict fluid abilities in either sex. Third, we reveal that largely overlapping connections between visual, dorsal attention, ventral attention, and temporal parietal networks are associated with better performance on crystallised and fluid cognitive tests in males and females, while connections within visual, somatomotor, and temporal parietal networks are associated with poorer performance. Together, our findings suggest that shared neurobiological features of the functional connectome underlie crystallised and fluid abilities across the sexes.

neuroscience↗

bioRxiv2020

Integrating multimodal connectivity improves prediction of individual cognitive abilities

SO_SCPLOWUMMARYC_SCPLOWHow white matter pathway integrity and neural co-activation patterns in the brain relate to complex cognitive functions remains a mystery in neuroscience. Here, we integrate neuroimaging, connectomics, and machine learning approaches to explore how multimodal brain connectivity relates to cognition. Specifically, we evaluate whether integrating functional and structural connectivity improves prediction of individual crystallised and fluid abilities in 415 unrelated healthy young adults from the Human Connectome Project. Our primary results are two-fold. First, we demonstrate that integrating functional and structural information - at both a model input or output level - significantly outperforms functional or structural connectivity alone to predict individual verbal/language skills and fluid reasoning/executive function. Second, we show that distinct pairwise functional and structural connections are important for these predictions. In a secondary analysis, we find that structural connectivity derived from deterministic tractography is significantly better than structural connectivity derived from probabilistic tractography to predict individual cognitive abilities.

neuroscience↗

bioRxiv2020

White matter hyperintensity-associated structural disconnection, resting state functional connectivity, and cognitive control in older adults

ObjectiveWhite matter hyperintensities (WMH) are linked to deficits in cognitive functioning, including cognitive control and memory; however, the structural and functional mechanisms are largely unknown. We investigated the relationship between estimated regional disruptions to white matter fiber tracts from WMH, resting state functional connectivity (RSFC), and cognitive functions in older adults. DesignCross-sectional study. SettingCommunity. ParticipantsFifty-eight cognitively-healthy older adults. MeasurementsTasks of cognitive control and memory, structural MRI, and resting state fMRI. We estimated the disruption to white matter fiber tracts from WMH and its impact on gray matter regions in the cortical and subcortical frontoparietal network, default mode network, and ventral attention network by overlaying each subjects WMH mask on a normative tractogram dataset. We calculated RSFC between nodes in those same networks. ResultsThe interaction of estimated regional WMH burden and RSFC in cortico-striatal regions of the default mode network and frontoparietal network was associated with memory retrieval. Models predicting working memory, cognitive inhibition, and set-shifting were not significant. ConclusionsFindings highlight the role of circuit-level alterations at the structural and functional levels in resting state networks that are related to WMH and impact memory retrieval in older adults.

neuroscience↗