SEARCH · PubMed Health

Results for “multimodal integration”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 451 records · Page 25Linked to original sources

PubMed1999

Multimodal MR imaging: functional, diffusion tensor, and chemical shift imaging in a patient with localization-related epilepsy.

PURPOSE: To demonstrate the integration of complementary functional and structural data acquired with magnetic resonance imaging (MRI) in a patient with localization-related epilepsy. METHODS: We studied a patient with partial and secondarily generalized seizures and a hemiparesis due to a malformation of cortical development (MCD) in the right hemisphere by using EEG-triggered functional MRI (fMRI), diffusion tensor imaging (DTI), and chemical shift imaging (CSI). RESULTS: fMRI revealed significant changes in regional blood oxygenation associated with interictal epileptiform discharges within the MCD. DTI showed a heterogeneous microstructure of the MCD with reduced fractional anisotropy, a high mean diffusivity, and displacement of myelinated tracts. CSI demonstrated low N-acetyl aspartate (NAA) concentrations in parts of the MCD. CONCLUSIONS: The applied MR methods described functional, microstructural, and biochemical characteristics of the epileptogenic tissue that cannot be obtained with other noninvasive means and thus improve the understanding of the pathophysiology of epilepsy.

Adult↗

PubMed2025

Target and biomarker exploration portal for drug discovery.

MOTIVATION: The discovery of novel drug targets and precision biomarkers remains a major challenge in drug development, with traditional differential expression analysis often overlooking key regulatory proteins. Here, we present a novel, web-based bioinformatics tool, the Target and Biomarker Exploration Portal (TBEP), designed to accelerate the drug discovery process by integrating large-scale biomedical data with network analysis techniques. RESULTS: TBEP harnesses machine-learning approaches to mine and combine multimodal datasets, including human genetics, functional genomics, and protein-protein interaction networks, to decode causal disease mechanisms and uncover novel therapeutic targets and precision biomarkers for specific phenotypes. A unique feature of the tool is its ability to process large-scale data in real-time, facilitated by an efficient cloud-based architecture. Additionally, the tool incorporates an integrated large language model (LLM), which assists researchers in exploring and interpreting complex biological relationships within the generated networks and multi-omics data using natural language (English). By offering an intuitive, interactive interface, the LLM enhances the exploration of biological insights, making it easier for scientists to derive actionable conclusions. This powerful integration of network analysis, multi-omics data, and LLM provides a robust framework for accelerating the identification of novel drug targets. AVAILABILITY AND IMPLEMENTATION: The tool is publicly available at https://tbep.missouri.edu. The source code, documentation and installation instructions are available at GitHub repository: https://github.com/mizzoudbl/tbep.

Drug Discovery↗

PubMed2025

Q RadFusion: Hybrid Quantum Classical Radiogenomic Framework for Breast Cancer Diagnosis.

BACKGROUND AND PURPOSE: Breast cancer remains the most common cancer in women worldwide, with early and accurate diagnosis critical for patient survival. Radiogenomics integrates imaging phenotypes with genomic profiles, offering a pathway to precision diagnostics. However, existing classical machine learning models often struggle with the high dimensionality and heterogeneity of multimodal data, leading to issues in calibration and reproducibility. This study presents Q RadFusion, a hybrid quantum-classical framework designed to enhance breast cancer diagnosis by fusing mammography and genomics data. METHODS: Q RadFusion was implemented on two publicly available datasets: CBIS-DDSM (2,600 curated mammography cases, TCIA) and TCGA-BRCA (1,000 genomic profiles, GDC). Imaging preprocessing included bias-field correction, segmentation, and harmonization, while genomic data underwent normalization and imputation. Feature selection was performed using the Quantum Approximate Optimization Algorithm (QAOA), and features were mapped into a quantum Hilbert space using Variational Quantum Circuits (VQC). For multimodal fusion, ResNet encoded mammography features, and a Transformer encoded genomic features. Patient-level and site-held-out splits were used for evaluation. RESULTS: Q RadFusion achieved an AUC of 0.96 and accuracy of 94%, outperforming baselines including CNN-LSTM, ResNet + XGBoost, and multimodal Transformers. Ablation studies confirmed the contribution of quantum components, with optimal performance observed at circuit depth, qubits, and QAOA layers. The model also demonstrated improved calibration and ~ 80% fewer parameters compared to deep fusion networks. CONCLUSION: Q RadFusion demonstrates that hybrid quantum-classical radiogenomic integration can deliver accurate, reproducible, and clinically meaningful diagnostic support for breast cancer, with strong potential for future clinical translation.

Breast Cancer↗

PubMed2003

Combined microsurgical and endovascular management of complex intracranial aneurysms.

OBJECTIVE: The disciplines of microneurosurgery and cranial base surgery have reached maturity, and technical advances in the surgical management of aneurysms are limited. Although most aneurysms can be clipped microsurgically or coiled endovascularly, a subset of patients may require a combined approach. A consecutive series of patients with aneurysms in one surgeon's cerebrovascular practice was reviewed retrospectively to analyze strategies for integrating microsurgical and endovascular techniques in the management of complex aneurysms. METHODS: Between 1997 and 2001, 596 aneurysms in 491 patients were treated microsurgically by the senior author (MTL) at the University of California, San Francisco, and 77 of these patients (96 aneurysms) were managed with a multimodality approach comprising a total of eight different combinations: selective revascularization and aneurysm occlusion (n = 23), endovascular and surgical trapping (n = 1), clipping of the aneurysm after attempted or incomplete coiling (n = 22), coiling after attempted or incomplete clipping (n = 5), clipping of recurrent aneurysm after coiling (n = 6), coiling of recurrent aneurysm after clipping (n = 1), clipping and coiling of multiple remote aneurysms (n = 13), and coiling after previous surgery (n = 6). RESULTS: A total of 96 aneurysms were treated with combined therapy, of which 43% were large or giant in size and 34% had fusiform or dolichoectatic morphology. Complete angiographic obliteration was achieved in 91 aneurysms (95%). Overall, 66 patients (86%) had good outcomes (Glasgow Outcome Scale score of 4 or 5; mean follow-up, 9 mo). The treatment mortality rate was 9.1% (seven patients), and permanent treatment-associated neurological morbidity rate was 5.2% (four patients). CONCLUSION: Evolving endovascular technologies need to be integrated into the microsurgical management of aneurysms. Multimodality approaches are best used with complex aneurysms in which conventional therapy with a single modality has failed. Revascularization remains a unique surgical contribution to the overall management of aneurysms with which current endovascular techniques cannot be used. Multimodality management should be considered an elegant addition to the therapeutic armamentarium that, through simplification and increased safety, improves the treatment of complex aneurysms beyond what is achievable by performing clipping or coiling alone.

Adolescent↗

PubMed2004

Interpretation of monte carlo simulations using parameter space plots.

Multimodal distribution functions that result from Monte Carlo simulations can be interpreted by superimposing joint probability density functions onto the contour space of the simulated calculations. The method is demonstrated by analysis of the pathway of a radioactive groundwater contaminant using an analytical solution to the transport equation. Simulated concentrations at a fixed time and distance produce multimodal histograms, which are understood with reference to the parameter space for the two random variables-velocity and dispersivity. Numerical integration under the joint density function up to the contour of the analytical solution gives the probability of contaminant exceeding a target concentration. This technique is potentially more efficient than Monte Carlo simulation for low probability events. Visualization of parameter space is restricted to two random variables. Nevertheless, analyzing the two most pertinent random variables in a simulation might still offer insights into the multimodal nature of output histograms.

Journal Article↗

PubMed2025

CrossAttOmics: multiomics data integration with cross-attention.

MOTIVATION: Advances in high throughput technologies enabled large access to various types of omics. Each omics provides a partial view of the underlying biological process. Integrating multiple omics layers would help have a more accurate diagnosis. However, the complexity of omics data requires approaches that can capture complex relationships. One way to accomplish this is by exploiting the known regulatory links between the different omics, which could help in constructing a better multimodal representation. RESULTS: In this article, we propose CrossAttOmics, a new deep-learning architecture based on the cross-attention mechanism for multiomics integration. Each modality is projected in a lower dimensional space with its specific encoder. Interactions between modalities with known regulatory links are computed in the feature representation space with cross-attention. The results of different experiments carried out in this article show that our model can accurately predict the types of cancer by exploiting the interactions between multiple modalities. CrossAttOmics outperforms other methods when there are few paired training examples. Our approach can be combined with attribution methods like LRP to identify which interactions are the most important. AVAILABILITY AND IMPLEMENTATION: The code is available at https://github.com/Sanofi-Public/CrossAttOmics and https://doi.org/10.5281/zenodo.15065928. TCGA data can be downloaded from the Genomic Data Commons Data Portal. CCLE data can be downloaded from the depmap portal.

Humans↗

PubMed2000

The development of intersensory temporal perception: an epigenetic systems/limitations view.

Several theories have stressed the importance of intersensory integration for development but have not identified specific underlying integration mechanisms. The author reviews and synthesizes current knowledge about the development of intersensory temporal perception and offers a theoretical model based on epigenetic systems theory, proposing that responsiveness to 4 basic features of multimodal temporal experience--temporal synchrony, duration, temporal rate, and rhythm--emerges in a sequential, hierarchical fashion. The model postulates that initial developmental limitations make intersensory synchrony the basis for the integration of intersensory temporal relations and that the emergence of responsiveness to the other, increasingly more complex, temporal relations occurs in a hierarchical, sequential fashion by building on the previously acquired intersensory temporal processing skills.

Adult↗

PubMed1999

Replication of passive whole-body linear displacements from inertial cues. Facts and mechanisms.

Using path integration, normal subjects should be able to compute the distance of a traveled path even from the sole inertial sensory input. Blindfolded subjects were submitted to a passive linear forward displacement along 2 to 10 m. Their task was to replicate the traveled distance, still blindfolded, by driving the vehicle they were seated upon using a joystick that controlled linear speed. Subjects replicated both the length and the velocity profile of the passive travel, suggesting that a dynamic record of experienced motion is stored in memory. Even when the replication of passive motion dynamics was made impossible, the subjects could still replicate the displacement. The results are explained by a dynamic feedback model that performs a running comparison between the perceived instantaneous displacement of the ongoing motion and the displacement derived from a spatiotemporal record of perceived passive motion. A multimodal acceleration-related sensory input is transformed into a displacement-related perception through double time-integration.

Adult↗

PubMed2024

Bridging the Gap From Proteomics Technology to Clinical Application: Highlights From the 68th Benzon Foundation Symposium.

The 68th Benzon Foundation Symposium brought together leading experts to explore the integration of mass spectrometry-based proteomics and artificial intelligence to revolutionize personalized medicine. This report highlights key discussions on recent technological advances in mass spectrometry-based proteomics, including improvements in sensitivity, throughput, and data analysis. Particular emphasis was placed on plasma proteomics and its potential for biomarker discovery across various diseases. The symposium addressed critical challenges in translating proteomic discoveries to clinical practice, including standardization, regulatory considerations, and the need for robust "business cases" to motivate adoption. Promising applications were presented in areas such as cancer diagnostics, neurodegenerative diseases, and cardiovascular health. The integration of proteomics with other omics technologies and imaging methods was explored, showcasing the power of multimodal approaches in understanding complex biological systems. Artificial intelligence emerged as a crucial tool for the acquisition of large-scale proteomic datasets, extracting meaningful insights, and enhancing clinical decision-making. By fostering dialog between academic researchers, industry leaders in proteomics technology, and clinicians, the symposium illuminated potential pathways for proteomics to transform personalized medicine, advancing the cause of more precise diagnostics and targeted therapies.

Proteomics↗

PubMed2002

Real-time computing without stable states: a new framework for neural computation based on perturbations.

A key challenge for neural modeling is to explain how a continuous stream of multimodal input from a rapidly changing environment can be processed by stereotypical recurrent circuits of integrate-and-fire neurons in real time. We propose a new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks. It does not require a task-dependent construction of neural circuits. Instead, it is based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry. It is shown that the inherent transient dynamics of the high-dimensional dynamical system formed by a sufficiently large and heterogeneous neural circuit may serve as universal analog fading memory. Readout neurons can learn to extract in real time from the current state of such recurrent neural circuit information about current and past inputs that may be needed for diverse tasks. Stable internal states are not required for giving a stable output, since transient internal states can be transformed by readout neurons into stable target outputs due to the high dimensionality of the dynamical system. Our approach is based on a rigorous computational model, the liquid state machine, that, unlike Turing machines, does not require sequential transitions between well-defined discrete internal states. It is supported, as the Turing machine is, by rigorous mathematical results that predict universal computational power under idealized conditions, but for the biologically more realistic scenario of real-time processing of time-varying inputs. Our approach provides new perspectives for the interpretation of neural coding, the design of experiments and data analysis in neurophysiology, and the solution of problems in robotics and neurotechnology.

Action Potentials↗

PubMed2002

An assistive home care environment for people with special needs.

We have developed a system for supporting people with special needs. It allows remote monitoring of electrocardiographic and other signals together with multimode environmental control. The multimodal approach allows users with disabilities to interact more with the home environment. Telemedicine devices were integrated into the system to provide a link to health services. The system provided support for the independent living of people with special needs. The pilot site was a rehabilitation service attached to the National Paraplegic Hospital of Toledo, Spain. During the six-week trial period, the system did not fail, nor was any unauthorized access reported. Twelve people with special needs and three staff evaluated aspects of the system on a scale from 0 to 9, on which higher scores indicated a positive assessment. The mean score for efficiency was 7.8, for satisfaction 8, for helpfulness 7.4, for controllability 7.2 and for learnability 8.

Persons with Disabilities↗

PubMed2000

An overview of computer-integrated surgery and therapy.

Computer-integrated surgery and therapy (CIST): Methods and systems to help the surgeon or the physician use multimodality data (mainly medical images) in a rational and quantitative way, in order to plan but also to perform medical interventions through the use of passive, semi-active, or active guiding systems.

Diagnostic Imaging↗

PubMed1990

Clinical applications of evoked potentials in neurological disorders.

The functional integrity of the main sensorial systems and of the central motor pathways can be evaluated through multimodal evoked potentials and non-invasive transcranial stimulation of the motor cortex. In this review the authors summarize the main clinical indications of the afore said methods specifically in the field of the neurological disorders.

Evoked Potentials↗

PubMed1984

[Clinical and electroencephalographic features of the action of the psychic stimulant sydnocarb in various forms of epilepsy].

The qualitative and quantitative values of the baseline EEG differed significantly in patients with primarily generalized epilepsy and temporal epilepsy. The administration of anticonvulsants (phenobarbital, finlepsin) leads to a considerable clinical effect and a decrease in the values of the integrated EEG in the functional states studied in conscious subjects. Additional inclusion of sydnocarb into the multimodality treatment results in the intensification of the therapeutic response in patients with petit mal epilepsy which is accompanied by EEG desynchronization, whereas in temporal epilepsy with psychomotor paroxysms, sydnocarb addition produces an opposite effect.

Anticonvulsants↗

PubMed2000

How animals move: an integrative view.

Recent advances in integrative studies of locomotion have revealed several general principles. Energy storage and exchange mechanisms discovered in walking and running bipeds apply to multilegged locomotion and even to flying and swimming. Nonpropulsive lateral forces can be sizable, but they may benefit stability, maneuverability, or other criteria that become apparent in natural environments. Locomotor control systems combine rapid mechanical preflexes with multimodal sensory feedback and feedforward commands. Muscles have a surprising variety of functions in locomotion, serving as motors, brakes, springs, and struts. Integrative approaches reveal not only how each component within a locomotor system operates but how they function as a collective whole.

Animals↗

PubMedDate not supplied

Multimodal basis for egocentric spatial localization and orientation.

The perceptual and sensorimotor mechanisms that guide our abilities at localizing and orienting in space integrate sensory information from vision and from a "body-referenced mechanism" that itself makes use of extraretinal signals regarding eye position relative to the head and head orientation relative to the body and to gravity. The experiments and theoretical treatment center on two perceptual dimensions: the visual perception of elevation and of orientation within the frontoparallel plane. Several experiments measuring localization in the horizontal plane are also treated. The experiments involve measurements of the physical elevation of visually perceived eye level (VPEL, a norm for perceived elevation), measurements of the physical orientation within the frontoparallel plane corresponding to visually perceived vertical (VPV), and measurements of the direction within a horizontal plane perceived as straight ahead (VPSA). VPEL and VPV are each significantly and systematically influenced by both the pitch and the roll of visual fields, and it is these influences that provide the basis for experimentally isolating the contributions of vision from those of the body-referenced mechanism. The VPEL discrimination is nearly invariant with variation in head and eye orientation. The possibility that influences from vision and from the body-referenced mechanism combine linearly is well supported. The visual influences on VPEL and VPV are controlled by the action of individual lines, and the same pitched-from-vertical lines (from pitched planes) or oblique lines within erect planes influence both discriminations. The Great Circle Model (GCM) accounts for the influences of individual lines, and contains rules for the influence of combinations of lines on both VPEL and VPV. GCM is interpreted by a 3-dimensional vector treatment in "egocentric orientation space."

Eye↗

PubMed1995

ICOHR: intelligent computer based oral health record.

The majority of work on computer use in the dental field has focused on non-clinical practice management information needs. Very few computer-based dental information systems provide management support of the clinical care process, particularly with respect to quality management. Traditional quality assurance methods rely on the paper record and provide only retrospective analysis. Today, proactive quality management initiatives are on the rise. Computer-based dental information systems are being integrated into the care environment, actively providing decision support as patient care is being delivered. These new systems emphasize assessment and improvement of patient care at the time of treatment, thus building internal quality management into the caregiving process. The integration of real time quality management and patient care will be expedited by the introduction of an information system architecture that emulates the gathering and storage of clinical care data currently provided by the paper record. As a proposed solution to the problems associated with existing dental record systems, the computer-based patient record has emerged as a possible alternative to the paper dental record. The Institute of Medicine (IOM) recently conducted a study on improving the efficiency and accuracy of patient record keeping. As a result of this study, the IOM advocates the development and implementation of computer-based patient records as the standard for all patient care records. This project represents the ongoing efforts of The University of Iowa College of Dentistry's collaboration with the University of Uppsala Data Center, Uppsala, Sweden, on a computer-based patient dental record model. ICOHR (Intelligent Computer Based Oral Health Record) is an information system which brings together five important parts of the patient's dental record: medical and dental history; oral status; treatment planning; progress notes; and a Patient Care Database, generated from their clinical care information (the database is also stored in the ICOHR). ICOHR is designed to be integrated into a traditional practice management system. The components of the ICOHR system support the use of various types of clinical care quality management tools, including medical alerts, clinical care guidelines, care modifiers, and diagnostic decision support. Data input is multimodal, so the user may use both voice recognition and direct input with a digitizer board to enter information into the database. ICOHR is designed to be integrated into the clinical environment in an ergonomic fashion in order to facilitate the unobtrusive and efficient acquisition of patient information. ICOHR is currently under clinical evaluation in both private practice and institutional environments. The private practice is a large general dentistry practice with over twenty sites scattered throughout a large metropolitan area. The institutional settings are a College of Dentistry and a Hospital Dentistry program. The evaluations have started in two of the sites and the other site will be phased in during the next six months. Our demonstration of the system will include both prepared presentations of the system's various functions and provide an opportunity for hands-on use of the system for interested attendees.

Artificial Intelligence↗

PubMed1987

Electrophysiological monitoring in the intensive care unit.

Spontaneous electrical activity (EEG) and event-related computer averaged brain electrical potentials (EPs) are becoming an integral part of CNS function monitoring in neurological intensive care patients. EEG monitoring using the compressed spectral array (CSA) computer analysis offers continuous information about cerebral electrical activity, permitting an accurate definition of the severity of brain injury, forecasting patient's course, and early detection of secondary intracranial changes. It offers more precise indications for treatment and prognosis. Combining EEG with multimodality EPs permits a finer diagnosis of the location and extent of CNS damage and contributes information, not otherwise obtainable, on the integrity of CNS pathways. Changes is electrophysiological patterns are often related to changes in other physiological parameters or events affecting the patient. Consequently, it appears advantageous to monitor simultaneously other CNS and body functions and record, process and display the data obtained in the patient care area. To make this comprehensive monitoring system reliable and useful, a wise application of advanced computer technology and a high degree of understanding of intracranial dynamics are required.

Brain↗