Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “multimodal data”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 109 records · Page 6Linked to original sources

Medical management of early-stage breast cancer.

With improved screening and education, a greater proportion of breast cancer is detected at an early stage. Although the prognosis for many of these patients is excellent following definitive local therapy alone, some subsets of node-negative patients have a 30% chance of eventually developing metastatic disease that will be incurable with current therapy. Thus, an increasing proportion of early-stage patients are being offered some form of adjuvant therapy, with the expectation of improved relapse-free survival, and possibly improved overall survival. Efforts have been made to base the selection of patients for adjuvant therapy on specific prognostic factors. Meanwhile, the scope and complexity of putative prognostic factors continues to widen, and now includes such items as the presence of occult microscopic metastases, DNA ploidy and proliferative fraction, cytogenetic abnormalities, oncogene expression, growth factor receptors, and expression of hormonally regulated proteins. In addition, there is now a considerable range of options with regard to the composition, dose intensity, and sequence of multimodality therapy. Data regarding the classification, significance, and interpretation of prognostic factors is reviewed together with the development, current status, and recommendations regarding adjuvant therapy for patients with early-stage breast cancer. For 1991, the National Cancer Institute (NCI) has estimated that 175,000 new cases of breast cancer will be diagnosed in American women. It is also estimated that 44,500 women will die of breast cancer. Unfortunately, the age-adjusted death rate from breast cancer has shown no overall change from 1930 through 1987. However, effective screening techniques continue to identify an increasing percentage of early-stage tumors, which should exceed 50% of all new tumors in 1991. Ultimately, our understanding of environmental and genetic risk factors may identify new ways to reduce the impact of this disease. In the interim, development and application of effective systemic adjuvant chemotherapy and hormonal therapy has become increasingly important. There is no question that a greater proportion of patients with less extensive disease are now being offered some form of adjuvant therapy. Meanwhile, selection of patients for adjuvant therapy, and choice among specific adjuvant regimens, has remained controversial. Analysis of multiple prognostic factors is performed not only in the context of cooperative investigational trials, but more often in the offices of individual physicians caring for individual patients. Tumor biopsies can now be routinely sent to specialized laboratories for performance of complex assays with potential prognostic information, although interpretation of these results with reference to a specific patient is often uncertain.(ABSTRACT TRUNCATED AT 400 WORDS)

Antineoplastic Agents↗

Artificial intelligence agents and agentic artificial intelligence applied to precision medicine.

Precision medicine seeks to individualise care by integrating multimodal biomedical data, yet most deployed clinical artificial intelligence (AI) remains assistive, providing predictions without managing workflows or adapting autonomously. Agentic AI, built on large language models (LLMs), has emerged as a paradigm characterised by autonomy, goal-directed reasoning, memory, planning and tool use. This review synthesises evidence on agentic AI and LLMs applied to precision medicine, encompassing drug discovery, genomics, oncology, rare disease diagnostics and clinical pharmacology. This review also examines architectural components, recent validation milestones and emerging challenges, including hallucination, sociodemographic bias and evolving regulatory frameworks across the FDA, the EU AI Act and the WHO.

agentic AI↗

ChromoViz: multimodal visualization of gene expression data onto chromosomes using scalable vector graphics.

SUMMARY: ChromoViz is an R package for the visualization of microarray gene expression data, cross-species and cross-platform comparisons, as well as non-expression genomic data obtained from public databases onto chromosomes. Chromosomal visualization format is proposed for the clear decoupling of the data layer from the procedure layer and the combined visualization of genomic data from heterogeneous data sources. Visualization with Javascript-enabled scalable vector graphics enables interactive visualization and navigation of data objects on the Web. AVAILABILITY: http://www.snubi.org/software/ChromoViz/

Chromosome Mapping↗

Efforts towards a precision medicine approach in juvenile idiopathic arthritis.

Juvenile idiopathic arthritis (JIA) is the commonest group of childhood arthritides. Despite the availability of advanced therapeutics, many children and young people (CYP) with JIA experience disease flares, and in some, chronic joint damage. Tailoring treatment based on unique biological profiles would benefit CYP with JIA given their variable clinical presentation and disease course. To date, biomarkers to predict treatment response are lacking. With advances in single cell technologies, we are now able to profile the genes and proteins of target tissues at unprecedented resolution to define the biological basis of disease and guide novel treatment approaches. The complex analyses and combination of biological and clinical outcome data from large datasets across disease phenotypes have become possible with the development of computational and machine learning methods. Here, we summarize the strategies to integrate data through multimodal based approaches to maximize precision medicine and research priorities for CYP with JIA.

Humans↗

Characterization of neurophysiologic alerts during anterior cervical spine surgery.

STUDY DESIGN: A retrospective review of neurophysiologic alerts during anterior cervical surgery. OBJECTIVES: To examine incidence and types of neurophysiologic alerts and their correlation with new postoperative neurologic deficits after anterior cervical discectomy or corpectomy procedures. SUMMARY OF BACKGROUND DATA: Although multimodality neurophysiologic monitoring has been shown to predict iatrogenic neurologic injuries in scoliosis surgeries, their role in degenerative or trauma-related anterior cervical spine surgery is still unclear. MATERIALS AND METHODS: We retrospectively reviewed 1,445 patients who underwent anterior cervical discectomy or corpectomy and arthrodesis with neurophysiologic monitoring that included transcranial electrical motor-evoked potentials (tceMEP), somatosensory-evoked potentials (SSEP), and spontaneous electromyography (EMG). Intraoperative alerts were analyzed for type, perceived cause, actions taken to reverse or minimize the possible spinal cord injury, and any new postoperative neurologic deficits. RESULTS: There were 267 (18.4%) procedures that had either minor (spontaneous, sustained EMG) or major (tceMEP/SSEP amplitude reduction) alerts. Patients who underwent corpectomies had 28% increased risk of having a major neurophysiologic alert compared with those who had discectomies. Diagnosis of cervical spondylotic myelopathy or trauma increased the risk of having a major neurophysiologic alert 30% and 76%, respectively, compared with cervical radiculopathy. Eight surgeries were aborted due to persistent tceMEP/SSEP amplitude loss, but none resulted in new postoperative neurologic deficits. Two patients had halo-vest applied due to early termination of surgery. One of these patients ultimately could not receive definitive surgical stabilization. DISCUSSION AND CONCLUSION: Diagnosis of cervical spondylotic myelopathy or trauma and cervical corpectomy procedures increase the risk for having major intraoperative alerts. In case of persistent tceMEP/SSEP amplitude loss, consider delaying potentially harmful interventions, such as premature termination of the procedure or methylprednisolone infusion, until a new neurologic deficit is verified with an awake-clinical examination.

Adult↗

Computer-assisted surgical treatment of orbitozygomatic fractures.

Orbitozygomatic fractures pertain to the most common injuries in craniofacial trauma patients. Accurate fracture reduction is of high importance for a successful outcome. This pilot study was performed to assess the potential benefit of surgical navigation to aid in orbitozygomatic fracture reduction. A non-comparative series of five consecutive patients with severely displaced orbitozygomatic fractures was treated using the guidance of computed tomography (CT)-based surgical navigation. Using a previously developed software platform, the fracture was reduced virtually by a three-dimensional shifting of the orbitozygomatic complex within the patient's preoperative multimodal CT data set. This treatment plan was transferred to a navigation system. Fracture reduction was performed according to the treatment plan using surgical navigation. Intraoperative control of fracture reduction by comparing the real with the virtual bone position using surgical navigation showed up as a helpful tool. Accurate treatment planning and immediate evaluation of craniofacial surgery outcome are the benefits of the new approach demonstrated. A major drawback of the presented approach is a high consumption of human and financial resources. A larger clinical series with long-term follow-up will be needed to determine reproducibility and cost-effectiveness. In addition to bone repositioning, a future application may include simulation of craniofacial osteotomies.

Feasibility Studies↗

JASMINE: A powerful representation learning method for enhanced analysis of incomplete multi-omics data.

Integrative analysis of multi-omics data provides a more comprehensive and nuanced view of a subject's biological state. However, high-dimensionality and ubiquitous modality missingness present significant analytical challenges. Existing methods for incomplete multi-omics data are scarce, do not fully leverage both modality-specific and shared information, and produce task-biased representations. We propose JASMINE, a self-supervised representation learning method for incomplete multi-omics data that preserves both modality-specific and joint information and enhances sample similarity structure. JASMINE produces embeddings that achieve superior performance across multiple tasks for two different incomplete multi-omics datasets while requiring only a single round of training per dataset.

missing data↗

Mammogram registration: a phantom-based evaluation of compressed breast thickness variation effects.

The temporal comparison of mammograms is complex; a wide variety of factors can cause changes in image appearance. Mammogram registration is proposed as a method to reduce the effects of these changes and potentially to emphasize genuine alterations in breast tissue. Evaluation of such registration techniques is difficult since ground truth regarding breast deformations is not available in clinical mammograms. In this paper, we propose a systematic approach to evaluate sensitivity of registration methods to various types of changes in mammograms using synthetic breast images with known deformations. As a first step, images of the same simulated breasts with various amounts of simulated physical compression have been used to evaluate a previously described nonrigid mammogram registration technique. Registration performance is measured by calculating the average displacement error over a set of evaluation points identified in mammogram pairs. Applying appropriate thickness compensation and using a preferred order of the registered images, we obtained an average displacement error of 1.6 mm for mammograms with compression differences of 1-3 cm. The proposed methodology is applicable to analysis of other sources of mammogram differences and can be extended to the registration of multimodality breast data.

Algorithms↗

Machine learning for detection and diagnosis of disease.

Machine learning offers a principled approach for developing sophisticated, automatic, and objective algorithms for analysis of high-dimensional and multimodal biomedical data. This review focuses on several advances in the state of the art that have shown promise in improving detection, diagnosis, and therapeutic monitoring of disease. Key in the advancement has been the development of a more in-depth understanding and theoretical analysis of critical issues related to algorithmic construction and learning theory. These include trade-offs for maximizing generalization performance, use of physically realistic constraints, and incorporation of prior knowledge and uncertainty. The review describes recent developments in machine learning, focusing on supervised and unsupervised linear methods and Bayesian inference, which have made significant impacts in the detection and diagnosis of disease in biomedicine. We describe the different methodologies and, for each, provide examples of their application to specific domains in biomedical diagnostics.

Algorithms↗

Slow rhythmic oscillations of blood pressure, intracranial pressure, microcirculation, and cerebral oxygenation. Dynamic interrelation and time course in humans.

BACKGROUND AND PURPOSE: Various biological signals show nonpulsatile, slow rhythmic oscillations. These include arterial blood pressure (aBP), blood flow velocity in cerebral arteries, intracranial pressure (ICP), cerebral microflow, and cerebral tissue PO2. Generation and interrelations between these rhythmic fluctuations remained unclear. The aim of this study was to analyze whether stable dynamic interrelations in the low-frequency range exist between these different variables, and if they do, to analyze their exact time delay. METHODS: In a clinical study, 16 comatose patients with either higher-grade subarachnoid hemorrhage or severe traumatic brain injury were examined. A multimodal digital data acquisition system was used to simultaneously monitor aBP, flow velocity in the middle cerebral artery (FVMCA), ICP, cerebral microflow, and oxygen saturation in the jugular bulb (SjO2). Cross-correlation as a means to analyze time delay and correlation between two periodic signals was applied to a time series of 30 minutes' duration divided into four segments of 2048 data points (approximately 436 seconds) each. This resulted in four cross-correlations for each 30-minute time series. If the four cross-correlations were consistent and reproducible, averaging of the original cross-correlations was performed, resulting in a representative time delay and correlation for the complete 30-minute interval. RESULTS: Reproducible cross-correlations and stable dynamic interrelations were found between aBP, FVMCA, ICP, and SjO2. The mean time delay between aBP and ICP was 6.89 +/- 1.90 seconds, with a negative correlation in 81%. A mean time delay of 1.50 +/- 1.29 seconds (median, 0.85 seconds) was found between FVMCA and ICP, with a positive correlation in 94%. The mean delay between ICP and SjO2 was 9.47 +/- 2.21 seconds, with a positive correlation in 77%. Mean values of aBP and ICP did not influence the time delay and dynamic interrelation between the different parameters. CONCLUSIONS: These results strongly support Rosner's theory that ICP B-waves are the autoregulatory response of spontaneous fluctuations of cerebral perfusion pressure. There is casuistic evidence that failure of autoregulation significantly modifies time delay and the correlation between aBP and ICP.

Blood Pressure↗

Monitoring of brain tissue oxygenation following severe subarachnoid hemorrhage.

The purpose of this prospective observational study was to investigate the relation between the frequency of critical neuromonitoring parameters (brain tissue pO2, (PtiO2) < or = 10 mmHg, intracranial pressure (ICP) > 20 mmHg, cerebral perfusion pressure (CPP) < or = 70 mmHg) and outcome after severe aneurysmal subarachnoid hemorrhage (SAH). In a prospective study on 42 patients monitoring of ICP, CPP, and PtiO2 (in the area at risk for vasospasm) was performed. All patients were primarily classified as Hunt and Hess grade 4 or with secondary deterioration to this grade. Relative proportions of PtiO2 < or = 10 mmHg (n = 42), ICP > 20 mmHg (n = 25) and CPP < or = 70 mmHg (n = 23) were derived from multimodal neuromonitoring data sets for different time intervals, i.e. 1. the total monitoring time; 2. the total monitoring time without the last two monitoring days; 3. the second last monitoring day; and 4. the last monitoring day. Patients were divided into nonsurvivors (GOS = 1) and survivors (GOS = 3-5). For the total monitoring time, significant differences in the relative proportion of critical values were found for all neuromonitoring parameters (p < 0.05). The detailed analysis of consecutive time intervals revealed significantly increased proportions of critical values in nonsurvivors for all neuromonitoring parameters during the last day only. Additionally, ICP > 20 mmHg was significantly more frequent during the second last day (p < 0.01). For other time periods no differences were observed. We conclude, that critical neuromonitoring values are not early predictors of nonsurvival in patients suffering from severe SAH.

Adult↗

Transcranial and extracranial ultrasound assessment of cerebral hemodynamics in vascular and Alzheimer's dementia.

BACKGROUND: Increasing life expectancy of the population leads to a higher incidence of dementia. Exact differentiation between the most common types, vascular dementia (VD) and Alzheimer's dementia (AD), is crucial to the development and application of new treatment strategies. Both conditions are thought to differ greatly by their extent of microvascular affection. Transcranial and extracranial ultrasound permits analysis of cerebral hemodynamics and should help to differentiate between VD and AD. We compare multimodal ultrasound data between VD, AD and controls, and give an overview of the literature on this topic. METHODS: Twenty VD and 20 AD patients were studied and compared with 12 age-matched controls. Transcranial color-coded ultrasound was performed to assess blood flow velocity (V(mean)) and pulsatility indices (PI) of the middle cerebral artery (MCA). Extracranial duplex and Doppler ultrasound techniques were used to assess the blood volume flow (BVF) in the anterior circulation (both internal carotid arteries [ICA]) and posterior circulation (both vertebral arteries [VA]), the global cerebral blood flow (CBF = BVF(ICA) + BVF(VA)), the global cerebral circulation time (CCT = time delay of echo-contrast bolus arrival between ICA and internal jugular vein) and global cerebral blood volume (CBV = CCT x CBF). RESULTS: MCA V(mean) in VD (36 +/- 8 cm/s) and AD (43 +/- 13 cm/s) were significantly lower than in controls (59 +/- 13 cm/s) but did not differ significantly between VD and AD groups. PI (1.1 +/- 0.2; 1 +/- 0.2; 0.9 +/- 0.2) only differed significantly between VD group and controls. CBF and CCT in VD (570 +/- 61 ml/min; 8.8 +/- 2.6 s) and AD (578 +/- 77 ml/min; 8.2 +/- 1.4 s) were similar but differed significantly from controls (733 +/- 54 ml/min; 6.4 +/- 0.8 s). BVF in the anterior and posterior circulation of VD group (412 +/- 62 and 158 +/- 38 ml/min) and AD group (428 +/- 62 and 150 +/- 41 ml/min) were significantly lower than in controls (537 +/- 48 and 199 +/- 26 ml/min) but did not differ significantly between the patient groups. DISCUSSION: Transcranial and extracranial ultrasound does not help to distinguish between VD and AD. However, our results add insight into the pathophysiology of dementia, arguing in favor of a common 'vascular' pathway in both conditions.

Aged↗

The Radiation Therapy Oncology Group: a progress report.

The Radiation Therapy Oncology Group (RTOG), founded in 1971, comprises 35 university hospitals which are participating in 21 active protocols, four registries and eight pilot studies, including Phase I/II studies. Case accession has grown to over 1000 in 1977 with 19,801 patients being entered in the RTOG initial registration. Goals of the RTOG include: advancing knowledge of the role of radiation therapy in disease management; improving control of primary and regional disease; reducing morbidity and complications from treatment; and collaborating with other oncology disciplines to obtain data on multimodality therapy.

Clinical Trials as Topic↗

Web-based submission, archive, and review of radiotherapy data for clinical quality assurance: a new paradigm.

PURPOSE: To report on the implementation of a web-based system (the Resource Center for Emerging Technologies [RCET] System) that provides immediate access to the patient radiotherapy planning and delivery data for clinical quality assurance (QA) by the experts. MATERIALS AND METHODS: An infrastructure of comprehensive tools required for preparation, submission, auto-archiving, web-based review, and retrieval of diagnostic images, treatment planning images, and radiation therapy objects has been developed. These tools represent approximately 1.1 million lines of computer code development in seven languages (V, C++, Visual Basic, Java, ASP, HTML, and SQL) and consist of a secure auto-anonymizing upload and auto-archiving patient database, a web-based secure object archiving network system, a web-based rapid review tool, a web-based upload/download tool, and a personal computer client data application for data object preparation, visualization, and submission, named NetSys. The RCET system enables users to share radiotherapy data in a secure environment. This paradigm of electronic data exchange makes remote peer review very efficient and convenient. RESULTS: The RCET system can help the radiation therapy community ensure consistent evaluation of its therapies. It will encourage proactive QA. An example of proactive clinical QA would be to provide atlases of target and critical structure definitions, to serve as class solutions, as well as dose prescription, specification, and reporting examples for guidance to the radiation oncologists in the community. The web-based clinical quality assurance is ideally suited for emerging technologies in radiation therapy that generate complex and voluminous multimodality imaging and planning data. CONCLUSIONS: The RCET system enables users to share multimodality imaging data, radiation therapy planning, and delivery data on demand. Our design paradigm will allow rapid peer review of radiotherapy data through a simple personal computer-based web browser.

Humans↗

Web architecture for the remote browsing and analysis of distributed medical images and data.

To provide easy retrieval, integration and evaluation of multimodal medical images and data in a web browser environment, distributed application technologies and Java programming were used to develop a client-server architecture based on software agents. The server side manages secure connections and queries to heterogeneous remote databases and file systems containing patient personal and clinical data. The client side is a Java applet running in a web browser and providing a friendly medical user interface to perform queries on patient and medical test data and integrate and visualize properly the various query results. A set of tools based on Java Advanced Imaging API enables to process and analyze the retrieved bioimages, and quantify their features in different regions of interest. The platform-independence Java technology makes the developed prototype easy to be managed in a centralized form and provided in each site where an intranet or internet connection can be located. Giving the healthcare providers effective tools for browsing, querying, visualizing and evaluating comprehensively medical images and records in all locations where they can need them - e.g. emergency, operating theaters, ward, or even outpatient clinics- the implemented prototype represents an important aid in providing more efficient diagnoses and medical treatments.

Ambulatory Care Facilities↗

Foundation model based multimodal transformer framework for survival analysis in HER2 stratified breast cancer.

Objective. To improve survival prediction for HER2-positive breast cancer by integrating histopathological, molecular, and clinical data using a multimodal transformer framework.Approach. We propose a multimodal transformer framework for breast cancer survival prediction using HER2 stratified (SurvMBC), a foundation model-enhanced architecture that fuses three data modalities: whole-slide images, clinical narratives, and molecular features. Tumor microenvironment features are extracted using a pathology language and image pre-training (PLIP), clinical narratives are processed with BioBERT, and miRNA expression plus DNA methylation data are embedded using Gen2Vec. These representations are integrated through a cross-modal transformer with attention mechanisms for survival prediction.Main results. The model was evaluated on 1,095 HER2-positive breast cancer patients from The Cancer Genome Atlas. SurvMBC achieved a concordance index (C-index) of 0.857 (95% CI: 0.834, 0.880), a low integrated Brier score, and a strong inverse negative binomial log-likelihood. Risk stratification based on model outputs significantly separated high- and low-risk groups (log-rankp< 0.01) and showed strong associations with tumor stage, grade, and hormone receptor status (allp< 0.05).Significance. SurvMBC demonstrates the effectiveness of multimodal fusion in addressing tumor heterogeneity and improving prognostic accuracy. The attention-based integration enables context-aware learning of survival-relevant features across modalities, supporting individualized risk stratification and risk-adaptive treatment planning for HER2 stratified breast cancer patients.

Breast Neoplasms↗

A computer program for fitting multimodal probability density functions.

A FORTRAN IV program is described, which may be run interactively with tutorial assistance or in batch and which allows a user to selectively fit any of seven probability density functions (p.d.f.'s) or a combination of the p.d.f.'s to a unimodal or multimodal histogram of empirical data. A "best-fit", uni- or multimodal p.d.f., which may be obtained by a method of nonlinear least squares or a generated p.d.f. may be displayed on a Tektronix 4010 terminal as a continuous curve against the background of a bar, square wave, symbol or point-plot histogram. The following, supportive statistical information is also displayed: (1) Kolmogorov-Smirnov probability of goodness of fit, (2) mean square error, (3) correlation coefficient, and (4) parameter estimates. The resident driver program and six overlayable segments have been implemented on a Digital Equipment Corporation LAB-11 minicomputer (PDP-11/20).

Computers↗

Multimodal cerebral monitoring in comatose head-injured patients.

Monitoring of comatose patients in the neurosurgical intensive care unit (NICU) is constantly extended by the development of new methods for monitoring of cerebral function, metabolism and oxygenation. To simplify the interpretation of the rising number of parameters, and to avoid data overflow, a multimodal cerebral monitoring (MCM) system has been developed for the acquisition, display, on-line analysis and recording of physiological parameters from multiple bedside data sources. This article describes the technical details and the design of this computerized data acquisition system for variable applications in clinical patient monitoring and research. A Windows (Microsoft Corporation, Redmont, Washington) platform was equipped with an analog/digital converter board. Software for multimodal cerebral monitoring was developed using LabVIEW for Windows (National Instruments, Austin, Texas), a graphical programming system. Two software modules were created: One for the automatic acquisition of data, display of time dependent trend graphs, processing of on-line histograms, special functions for research, and storage of data in compatible format. The other module serves as an off-line monitor to display recorded data in various modalities. The MCM system has been used in 30 comatose patients with severe head injury. Mean time of MCM is 5.3 days (+/- 2.8 days), resulting in a total running time of the system of about 3800 hrs. Hardware and software proved to run stable and safe. The MCM system has become a valuable tool for monitoring of comatose patients. The simultaneous display of trend graphs of various monitoring parameters and the online processing of histograms improved the survey of the patient's condition in the ICU. Recorded data were analysed offline and contribute to a consecutively increasing data bank.

Brain Injuries↗