Biomedical subjects

Fuhai Li

Publications and source records attributed to Fuhai Li.

12 recordsLinked to original sources

PubMed2026

BioMedGraphica: an all-in-one platform for joint textual biomedical prior knowledge and numeric graph generation.

MOTIVATION: Multiomics data analysis is essential for scientific discovery in precision medicine. However, translating analysis results of omics data analysis into novel scientific hypotheses remains a significant challenge. Human experts must manually review analysis results and generate new hypotheses based on extensive and interconnected biomedical prior knowledge, which is subjective and not scalable. While large language models can accelerate the discovery, their reasoning improves when grounded in structured, auditable, and comprehensive biomedical prior knowledge. However, biomedical knowledge is scattered across heterogeneous databases that use diverse and inconsistent nomenclature systems, making it difficult to integrate resources into a unified format for scalable analysis. This fragmentation limits the ability of artificial intelligence systems to fully leverage biomedical data for scientific discovery. RESULTS: We developed BioMedGraphica, a novel all-in-one platform that harmonizes fragmented biomedical resources by integrating 11 entity types and 30 relation types from 43 databases into a unified textual prior knowledge graph containing 2 306 921 entities and 27 232 091 relations. In addition, we present a novel textual-numeric graph (TNG) data structure concept, where textual information captures prior biological knowledge (e.g. transcription start sites, functions, mechanisms), numeric values represent quantitative biomedical features, and the integrated relations can help uncover mechanisms. By bridging prior knowledge with user-specific data, TNG is a novel and ideal data structure for developing novel graph analysis models. AVAILABILITY AND IMPLEMENTATION: The code is available at: https://github.com/FuhaiLiAiLab/BioMedGraphica and BioMedGraphica knowledge graph database can be downloaded from huggingface dataset: https://huggingface.co/datasets/FuhaiLiAiLab/BioMedGraphica.

Humans↗

PubMed2025

BioMedGraphica: An All-in-One Platform for Joint Textual Biomedical Prior Knowledge and Numeric Graph Generation.

Multi-omic data analysis is essential for scientific discovery in precision medicine. However, translating statistical results of omic data analysis into novel scientific hypothesis remains a significant challenge. Human experts must manually review analysis results and generate new hypothesis based on extensive and inter-connected biomedical prior knowledge, which is subjective and not scalable. While large language models (LLMs) can accelerate the discovery, their reasoning improves when grounded in structured, auditable and comprehensive biomedical prior knowledge. Biomedical knowledge, however, is scattered across heterogeneous databases that use diverse and inconsistent nomenclature systems, making it difficult to integrate resources into a unified format for scalable analysis. This fragmentation limits the ability of AI systems to fully leverage biomedical data for scientific discovery. To address these challenges, we developed BioMedGraphica , an all-in-one platform that harmonizes fragmented biomedical resources by integrating 11 entity types and 30 relation types from 43 databases into a unified knowledge graph containing 2,306,921 entities and 27,232,091 relations. In addition, to the best of our knowledge, this is the first work to propose a novel Textual-Numeric Graph (TNG) data-structure for multi-omics data analysis. In TNG, textual information captures prior biological knowledge (e.g., transcription start sites, functions, mechanisms), while numeric values represent quantitative biomedical features, and the integrated relations can help uncover mechanisms. By bridging prior knowledge with user-specific data, TNG is a novel and ideal data-structure for the development of graph foundation models, with the potential to improve prediction performance and interpretability, while also augmenting LLMs by supplying graph-structured mechanistic context to strengthen reasoning. The details for BioMedGraphica code can be accessed by github link: https://github.com/FuhaiLiAiLab/BioMedGraphica and BioMedGraphica knowledge graph data can be downloaded from huggingface dataset: https://huggingface.co/datasets/FuhaiLiAiLab/BioMedGraphica.

biomedical knowledge graph↗

PubMed2006

Long-term inhibition of Rho kinase with fasudil attenuates high flow induced pulmonary artery remodeling in rats.

Accumulating evidences have demonstrated that RhoA/Rho-kinase pathway plays a pivotal role in various cellular functions. The aim of this study was to explore whether RhoA/Rho-kinase pathway is involved in the pathogenesis of high flow induced pulmonary hypertension and whether long-term inhibition of RhoA/Rho-kinase pathway with fasudil could attenuate high flow induced pulmonary artery remodeling in rats. Wistar rats in the shunt groups and treated groups were underwent left common carotid artery-external jugular vein shunt operation, rats in control groups were sham-operated animals. Rats in treated groups received fasudil treatment, the others received same dose of saline. At week 4, 8 of the study, rats were underwent haemodynamics measurements, pulmonary artery morphologic assessments, detection of pulmonary artery smooth muscle cells (SMCs) proliferation and apoptosis. RhoA and Rho-kinase activity in pulmonary arteries were also analyzed. Compared with the control groups, exposure to high blood flow induced a significant elevation of right ventricle systolic pressure at week 8, significant increase of the mean percentage of media wall thickness (%MT) in moderate size pulmonary arteries both at weeks 4 and 8, marked elevation of right ventricle (RV) to left ventricle plus septum (LV+SP) weight ratio at week 8, significant increase of PCNA-positive SMCs percentage at week 4 and significant decrease of TUNEL-positive SMCs percentage both at weeks 4 and 8. High pulmonary blood flow also induced 3.19+/-0.28-fold increase of RhoA and 3.63+/-0.52-fold increase of Rho-kinase over the control group at week 4, 1.57+/-0.35-fold increase of RhoA and 2.36+/-0.39-fold increase of Rho-kinase over the control group at week 8. Compared with the shunt groups, fasudil treatment significantly suppressed Rho-kinase activity at both weeks 4 and 8, improved pulmonary hypertension at week 8, attenuated right ventricular hypertrophy at week 8, and enhanced pulmonary vascular remodeling both at weeks 4 and 8, which were associated with suppressed pulmonary artery smooth muscle cells proliferation at week 4 and apoptosis both at weeks 4 and 8. These results indicated that RhoA/Rho-kinase mediated pathway participated in the process of high flow induced pulmonary artery remodeling; inhibition of Rho kinase with fasudil could attenuate pulmonary artery remodeling.

1-(5-Isoquinolinesulfonyl)-2-Methylpiperazine↗

PubMed2004

Microsphere-induced embolic stroke: an MRI study.

Despite the many studies of the middle cerebral artery occlusion (MCAO) model, efficient therapy for stroke is still lacking, emphasizing the need for further development and characterization of experimental stroke models. In the present study, the rather unexplored multifocal microsphere-induced stroke model in rats was characterized by multiparametric MRI. We induced microembolic infarction in a group of Sprague-Dawley rats by injecting a dose of about 1000 50-microm polyethylene microspheres intracranially from the external carotid artery. Diffusion-, perfusion-, and T(2)-weighted MRI were used to evaluate the infarct development during and following the first 3 hr after microsphere injection (N = 20). The animals were also imaged at 12-hr (N = 8), 24-hr (N = 17), and 48-hr (N = 5) time points. After the final imaging time point, the brains were removed and sectioned into 2-mm-thick slices, and infarct volumes were measured by 2,3,4-triphenyltetrazolium chloride (TTC) staining. From calculated apparent diffusion coefficient (ADC) maps, a volume of reduced ADC appeared 0.5-1.0 hr postinjection, and by the 3-hr time point the volume of ADC reduction had increased to a size of 5% +/- 1% (mean +/- SEM) of the brain hemisphere. The lesion volume increased significantly (P < 0.01) to 16% +/- 2% of the hemisphere volume at the 12-hr time point, while at 24 hr the lesion (15% +/- 2% of the hemisphere) was also significantly larger (P < 0.001) than at 3 hr. The perfusion deficit resulting from the microsphere injection was immediate, going from a cerebral blood flow index (CBF(i)) of 74% +/- 3% at the time of microsphere injection to 68% +/- 2% of the contralateral mean at 3 hr (P < 0.05), to 55% +/- 4% of the contralateral values at 12 hr (P < 0.05), and to 57% +/- 2% of the contralateral mean at 24 hr (P < 0.001). The lesion development in the microsphere-induced stroke model was found to be slower than in the MCAO model, and continued up to the 24-48-hr time point.

Animals↗

PubMed2004

Investigation of techniques to quantify in vivo lesion volume based on comparison of water apparent diffusion coefficient (ADC) maps with histology in focal cerebral ischemia of rats.

Stroke lesion-volume estimates derived from calculated water apparent diffusion coefficient (ADC) maps provide a quantitative surrogate end-point for investigating the efficacy of drug treatment or studying the temporal evolution of cerebral ischemia. Methodology is described for estimating ischemic lesion volumes in a rat model of permanent middle cerebral artery occlusion (MCAO) based on absolute and percent-reduction threshold values of the water ADC at 3 h post-MCAO. Volume estimates derived from average ADC (ADC(av)) maps were compared with those derived from post-mortem histological sections. Optimum ADC thresholds were established as those that provided the best correlation and one-to-one correspondence between ADC- and histologically derived lesion-volume estimates. At 3 h post-MCAO, an absolute-ADC(av) threshold of 47 x 10(-5) mm(2)/s (corresponding to a 33% reduction in ADC(av) based on a contralateral hemisphere comparison) provided the most accurate estimate of percent hemispheric lesion volume (%HLV). Experimental and data analysis issues for improving and validating the usefulness of DWI as a surrogate endpoint for the quantification of ischemic lesion volume are discussed.

Acute Disease↗

PubMed2004

The role of diffusion- and perfusion-weighted magnetic resonance imaging in drug development for ischemic stroke: from laboratory to clinics.

Ischemic stroke is a major cause of mortality and morbidity in industrialized countries and is almost always caused by occlusion of a cerebral artery by a clot. As the reversibly injured brain tissue evolves into irreversible infarction within a short period of time after onset of ischemia, it is extremely important and urgent to reverse the serious consequences of brain ischemia in the hyperacute phase when the ischemic brain tissue is still salvageable. Numerous thrombolytic and potentially neuroprotective agents have been studied in stroke patients with little success as the only approved therapy is thrombolysis with recombinant tissue plasminogen activator (r-tPA) within 3 h of stroke onset in highly selected patients (approximately 5 to 10 % of all acute stroke patients). One major obstacle in the development of effective therapies for ischemic stroke has been the lack of versatile imaging techniques. New magnetic resonance imaging (MRI) modalities, specially diffusion- and perfusion-weighted MRI (DWI and PWI, respectively) have been used in experimental studies with great success for over a decade and now are gradually entering clinical use. DWI and PWI can detect brain ischemia in the early phase in its full extent thus ensuring a definite diagnosis, allowing for follow-up of the ischemic lesion size over time with good spatial and temporal resolution, demonstrating perfusion deficit and reperfusion and the existence and the extent of penumbra while only requiring a few minutes of imaging time. DWI and PWI do not just give us the correct diagnosis of ischemic stroke, but allow us to acquire in vivo lesion size before therapeutic regimen is started and monitor the therapeutic efficacy thereafter, thus overcoming the potential pretreatment bias. We used DWI and PWI to evaluate novel therapeutic approaches for ischemic stroke in numerous experimental studies and lately in humans. With DWI and PWI, we are able to determine the in vivo efficacy (or lack of efficacy) of new therapeutic regiments (both neuroprotective and thrombolytic agents, or combination therapies) in a rapid, safe, and reliable way and in a relatively small number of well-selected, well-defined, and homogeneous patients. This approach may, therefore, significantly accelerate the development of new remedies for stroke patients.

Animals↗

PubMed2003

Neuroprotective effects of a new synthetic peptide, CMX-9236, in in vitro and in vivo models of cerebral ischemia.

NGF (nerve growth factor) and BDNF (brain-derived neurotrophic factor) are protein molecules (MW 26 and 13.6 kDa, respectively) that are neuroprotective in the middle cerebral artery occlusion (MCAO) rat stroke model. Their mechanism of action involves the activation of transcription factor AP-1 that turns on neuronal growth genes. In our ongoing studies we are designing short peptides that mimic some of the properties of full-length neurotrophic factors. We have synthesized a neuroprotective 14-amino acid peptide (CMX-9236) with an N-terminal docosahexaenoic acid (DHA). DHA enhances entry through the blood-brain barrier. Using primary rat brain cortical cultures and a fluorescent assay we found that CMX-9236 can counteract the excitotoxic effects of glutamate or kainate, reversing the intracellular accumulation of Ca(2+) to normal levels. Administration (i.v.) of CMX-9236 post initiation of ischemia reduced the lesion volumes from 178+/-50 to 117+/-55 mm(3) in the temporary rat MCAO model (90 min), and from 216+/-58 to 127+/-57 mm(3) in the permanent (24 h) model for stroke, corresponding to 34+/-28% (P=0.01) and 41+/-19% (P=0.038) reductions of the infarct volumes. Neurological behavior scores showed 57 and 47% improvements for treated temporary and permanent models, respectively. Dose-response studies indicated a 60-fold activation of AP-1 transcription factor in cells treated with 100 ng/ml of the peptide. These studies illustrate that a small peptide can function as a neuroprotective agent and an activator of a beneficial signal transduction pathway.

Animals↗

PubMed2003

The macrosphere model: evaluation of a new stroke model for permanent middle cerebral artery occlusion in rats.

BACKGROUND AND PURPOSE: The suture middle cerebral artery occlusion (MCAO) model is widely used for the simulation of focal cerebral ischemia in rats. This technique causes hypothalamic injury resulting in hyperthermia, which can worsen outcome and obscure neuroprotective effects. Herein, we introduce a new MCAO model that avoids these disadvantages. METHODS: Permanent MCAO was performed by intraarterial embolization using six TiO(2) macrospheres (0.3-0.4 mm in diameter) or by the suture occlusion technique. Body temperature was monitored, functional and histologic outcome was assessed after 24 h. Additional 16 rats were subjected to macrosphere or suture MCAO. Lesion progression was evaluated using magnetic resonance imaging (MRI). RESULTS: The animals subjected to suture MCAO developed hyperthermia (>39 degrees C), while the temperature remained normal in the macrosphere MCAO group. Infarct size, functional outcome and model failure rate were not significantly different between the groups. Lesion size on MRI increased within the first 90 min and remained unchanged thereafter in both groups. CONCLUSIONS: The macrosphere MCAO model provides reproducible focal cerebral ischemia, similar to the established suture technique, but avoids hypothalamic damage and hyperthermia. This model, therefore, may be more appropriate for the preclinical evaluation of neuroprotective therapies and can also be used for stroke studies under difficult conditions, e.g., in awake animals or inside the MRI scanner.

Animals↗

PubMed2003

Use of diffusion- and perfusion-weighted magnetic resonance imaging in drug development for ischemic stroke.

Diffusion- and perfusion-weighted magnetic resonance imaging (DWI and PWI, respectively) are novel imaging modalities that can detect brain ischemia early in its full extent, can be performed in minutes, can be repeated easily, and allow for follow-up of the ischemic lesion size over time with good spatial and temporal resolution. We have used DWI and PWI in evaluating novel therapeutic approaches for ischemic stroke in numerous studies in the rat and lately in humans. It is now clear that DWI and PWI offer a good combination for safe and reliable evaluation of novel drugs on the size and tissue characteristics of brain ischemia. After inducing focal brain ischemia in the rat, one can first detect the presence and extent of ischemia by DWI and hypoperfusion by PWI, calculate the volume of ischemic brain tissue, and then follow the development of the ischemic lesion over time for several hours during treatment, thus detecting in vivo effects of the novel drug on brain ischemia. Successful reperfusion (either mechanically or as a result of thrombolytic therapy) can also be detected easily. DWI and PWI when performed before starting treatment can also exclude the pretreatment bias, a potential reason for false-positive studies in which proper imaging studies are not employed. Thus we can determine the in vivo efficacy (or lack of efficacy) of new therapeutic regimens (both neuroprotective and thrombolytic) rapidly, safely, and reliably by using a small sample size only, and adapt the same strategy to clinical trials.

Animals↗

PubMed2002

Separating changes in the intra- and extracellular water apparent diffusion coefficient following focal cerebral ischemia in the rat brain.

Selective intracellular (IC) and extracellular (EC) brain water apparent diffusion coefficient (ADC) values were measured in normal and ischemic rat brain. Selective T(1)-relaxation enhancement of the EC water, using intracerebroventricular (ICV) infusion of an NMR contrast reagent (CR), was used to separate the IC and EC signal contributions. In the CR-infused, normal brain (n = 4), T(1) = 235 +/- 10 ms and T(2) = 46 +/- 2 ms for IC water (85%) and T(1) = 48 +/- 8 ms and T(2) = 6 +/- 2 ms for EC water (15%). Volume-localized ADC(z) (z-gradient axis) values were 0.90 +/- 0.02 (EC+IC), 0.81 +/- 0.05 (IC), 0.51 +/- 0.02 (EC+IC), and 0.53 +/- 0.07 (IC), for normal, CR-infused, ischemic, and ischemic/CR-infused groups, respectively (ADC values are x10(-3) mm(2)/s; n = 5 for each group). Imaging ADC(z) values were 0.81 +/- 0.03 (EC+IC), 0.75 +/- 0.05 (IC), 0.51 +/- 0.04 (EC+IC), and 0.52 +/- 0.05 (IC), respectively, for the same groups. Imaging ADC(av) (average diffusivity) values for the same groups were 0.70 +/- 0.05 (EC+IC), 0.69 +/- 0.06 (IC), 0.45 +/- 0.06 (EC+IC), and 0.44 +/- 0.06 (IC), respectively. These results suggest that the IC water ADC determines the overall water ADC value in normal and ischemic rat brain.

Animals↗

PubMed2002

Dynamic changes in cerebral blood flow and angiogenesis after transient focal cerebral ischemia in rats. Evaluation with serial magnetic resonance imaging.

BACKGROUND AND PURPOSE: Angiogenesis occurs after cerebral ischemia, but the relationship between angiogenesis and cerebral hemodynamic change is unknown. The aim of the present study was to investigate the relationship between ischemia-induced angiogenesis and hemodynamics in a well-defined 3-vessel occlusion model of the rat by using diffusion- (DWI), perfusion-, and T2-weighted MRI (T2WI). METHODS: Rats were subjected to 60 minutes of transient middle cerebral artery occlusion or sham operation. DWI and T2WI were used to characterize the extent of the ischemic lesion from 4.5 hours to 14 days after reperfusion. A flow-sensitive alternating inversion recovery method and dynamic susceptibility contrast MRI were used to evaluate the temporal changes in relative cerebral blood flow (CBF) and cerebral blood volume (CBV), respectively. Rats were randomly selected and killed at each time point for investigation of vascular density and for hematoxylin-eosin staining. RESULTS: Ischemic lesions developed in the ipsilateral cortex, as demonstrated by DWI and T2WI. CBF was significantly increased in the ipsilateral cortex, especially in the cortical outer layer from day 1 to day 14, and peaked on day 7 (P<0.05), while CBV was significantly increased on day 7 (P<0.01). The vascular density on the ipsilateral brain surface was gradually increased from day 1 to day 5, peaked on day 7, and then decreased on day 14. Histology study showed pannecrosis in the cortex from day 1 to day 5 and partial liquefaction of the necrotic tissues on days 7 and 14. CONCLUSIONS: A delayed increase in both CBF and CBV is documented in the ipsilateral cortex after transient focal brain ischemia, and such an increase may be associated with angiogenesis.

Animals↗

PubMed2002

Acute postischemic renormalization of the apparent diffusion coefficient of water is not associated with reversal of astrocytic swelling and neuronal shrinkage in rats.

BACKGROUND AND PURPOSE: Initially decreased apparent diffusion coefficient (ADC) values are reversible if reperfusion is rapidly performed after focal brain ischemia. We sought to determine if reperfusion-induced renormalization of initially abnormal values indicates reversal of cellular, morphologic changes that occur during acute ischemia. METHODS: Eighteen rats underwent 30 minutes of middle cerebral artery occlusion (MCAO) without reperfusion (group A, n = 6), with 1.5 hours of reperfusion (group B, n = 6), or with 12 hours of reperfusion (group C, n = 6). Diffusion- and perfusion-weighted MR images were obtained at the end of MCAO and 1.5 and 12 hours after reperfusion. Immediately after the final MR study, the brains were fixed by cardiac perfusion with 4% paraformaldehyde. Neuronal injury was evaluated on hematoxylin-eosin-stained slices, and astrocytic size was determined by the area of glial fibrillary acidic protein (GFAP) plus S-100 expression. RESULTS: In group A in which ADC values decreased significantly, 47 +/-12% of the neurons were slightly shrunken; astrocytes were moderately swollen, and the area expressing GFAP plus S-100 was larger than that in the contralateral hemisphere (117 microm(2) +/- 6 vs 89 microm(2) +/- 2; P <.001). In group B in which ADC had renormalized, most neurons were moderately shrunken, and the frequency of such neurons was greater in group B (92% +/- 2) than in group A (P <.001); astrocytes were markedly swollen, and the area was larger than that in the contralateral hemisphere (123 microm(2) +/- 8 vs 85 microm(2) +/- 4, P <.001). In group C in which a secondary ADC decrease occurred, most neurons (94% +/- 3) were severely shrunken, and some had eosinophilic cytoplasm; astrocytes were disintegrated, and the area of GFAP plus S-100 expression was reduced (78 microm(2) +/- 4 vs 90 microm(2) +/- 5, P <.001). CONCLUSION: Reperfusion-induced acute renormalization of ADC values is not associated with the reversal of neuronal shrinkage and astrocytic swelling that occur during ischemia. Conversely, the morphologic changes of astrocytes and neurons progressively worsen over time, although ADC values show a biphasic change.

Animals↗