Artificial intelligence for healthcare and medical education: a systematic review.

BACKGROUND: Human society has entered the age of artificial intelligence, medical practice and medical education are undergoing profound changes. Artificial intelligence (AI) is now applied in many industries, particularly in healthcare and medical education, where it deeply intersects. The purpose of this paper is to overview the current situation and problems of "AI+medicine/medical" education and to provide our own perspective on the current predicament. METHODS: We searched PubMed, Embase, Cochrane and CNKI databases to assess the literature on AI+medical/medical education from 2017 to July 2022. The main inclusion criteria include literature describing the current situation or predicament of "AI+medical/medical education". RESULTS: Studies have shown that the current application of AI in medical education is focused on clinical specialty training and continuing education, with the main application areas being radiology, diagnostics, surgery, cardiology, and dentistry. The main role is to assist physicians to improve their efficiency and accuracy. In addition, the field of combining AI with medicine/medical education is steadily expanding, and the most urgent need is for policy makers, experts in the medical field, AI and education, and experts in other fields to come together to reach consensus on ethical issues and develop regulatory standards. Our study also found that most medical students are positive about adding AI-related courses to the existing medical curriculum. Finally, the quality of research on "AI+medical/medical education" is poor. CONCLUSION: In the context of the COVID-19 pandemic, our study provides an innovative systematic review of the latest "AI+medicine/medical curriculum". Since the AI+medicine curriculum is not yet regulated, we have made some suggestions.

Effects of white matter lesion grading on the cognitive function of patients with chronic alcohol dependence.

BACKGROUND: Alcohol dependence has become a major problem that poses a serious threat to public health. Long-term heavy alcohol consumption can lead to brain functional disorders. This study aimed to investigate the relationship of the severity of cerebral white matter lesions (WMLs), serum neurofilament light (NfL) and inflammatory factors, tumour necrosis factor alpha (TNF-α) and Interleukin-1ß (IL-1ß), with the cognitive function of patients with alcohol dependence. METHODS: A total of 118 patients were enrolled in this prospective study, and divided into alcohol-dependent and non-alcohol-dependent groups. The severity of WMLs was assessed using the Fazekas scale based on magnetic resonance imaging analysis. The expression levels of NfL, TNF-α and IL-1ß in the serum of the subjects were measured by enzyme-linked immunosorbent assay. The cognitive function and psychological status of the patients were assessed using the Minimum Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Hamilton Depression Rating Scale (HAMD) and Hamilton Anxiety Rating Scale (HAMA). The severity of WMLs and the expression levels of serum NfL, TNF-α and IL-1ß in alcohol-dependent patients were analysed for their influence on cognitive function. This clinical trial was approved by China Clinical Trials Registry, and the trial number is ChiCTR2200066057 (http://www.chictr.org.cn/searchproj.aspx). RESULTS: The score of Fazekas scale was higher, and the MMSE score and MoCA score were lower in the alcohol-dependent group than those in the non-alcohol-dependent group. Moreover, the Fazekas score of the alcohol-dependent group was negatively correlated with the MMSE and MoCA scores. The serum NfL, TNF-α and IL-1ß levels were higher in the alcohol-dependent group than in the non-alcohol-dependent group, and the serum NfL, TNF-α and IL-1ß levels in the alcohol-dependent group were negatively correlated with the MMSE and MoCA scores. CONCLUSION: Alcohol-dependent patients have more severe cerebral WMLs and significant cognitive impairment, particularly in visuospatial and executive functions, attention, calculation, abstraction, delayed recall and orientation. Serum NfL, TNF-α and IL-1ß may be used as biomarkers to assess alcohol related cognitive decline.

Study of gray matter atrophy pattern with subcortical ischemic vascular disease-vascular cognitive impairment no dementia based on structural magnetic resonance imaging.

Objective: This study explored the structural imaging changes in patients with subcortical ischemic vascular disease (SIVD)-vascular cognitive impairment no dementia (VCIND) and the correlation between the changes in gray matter volume and the field of cognitive impairment to provide new targets for early diagnosis and treatment. Methods: Our study included 15 patients with SIVD-normal cognitive impairment (SIVD-NCI), 63 with SIVD-VCIND, 26 with SIVD-vascular dementia (SIVD-VD), and 14 normal controls (NC). T1-weighted images of all participants were collected, and DPABI and SPM12 software were used to process the gray matter of the four groups based on voxels. Fisher's exact test, one-way ANOVA and Kruskal-Wallis H test were used to evaluate all clinical and demographic data and compare the characteristics of diencephalic gray matter atrophy in each group. Finally, the region of interest (ROI) of the SIVD-VCIND was extracted, and Pearson correlation analysis was performed between the ROI and the results of the neuropsychological scale. Results: Compared to the NC, changes in gray matter atrophy were observed in the bilateral orbitofrontal gyrus, right middle temporal gyrus, superior temporal gyrus, and precuneus in the SIVD-VCIND. Gray matter atrophy was observed in the left cerebellar region 6, cerebellar crural region 1, bilateral thalamus, right precuneus, and calcarine in the SIVD-VD. Compared with the SIVD-VCIND, gray matter atrophy changes were observed in the bilateral thalamus in the SIVD-VD (p < 0.05, family-wise error corrected). In the SIVD-VCIND, the total gray matter volume, bilateral medial orbital superior frontal gyrus, right superior temporal gyrus, middle temporal gyrus, and precuneus were positively correlated with Boston Naming Test score, whereas the total gray matter volume, right superior temporal gyrus, and middle temporal gyrus were positively correlated with overall cognition. Conclusion: Structural magnetic resonance imaging can detect extensive and subtle structural changes in the gray matter of patients with SIVD-VCIND and SIVD-VD, providing valuable evidences to explain the pathogenesis of subcortical vascular cognitive impairment and contributing to the early diagnosis of SIVD-VCIND and early warning of SIVD-VD.

MiR-107 overexpression attenuates neurotoxicity induced by 6-hydroxydopamine both in vitro and in vivo.

Alzheimer's disease (AD), the most common form of dementia, is a neurodegenerative disease characterized by neuronal atrophy in various brain regions. The expression of miR-107 is down-regulated in AD patients and target genes of miR-107 have been shown to directly involved in AD. In this study, we aimed to investigate the potential neuroprotective effects of miR-107. We first assessed brain activity in health controls and patients with AD. Then we examined miR-107 expression in SH-SY5Y and PC12 cells treated with 6-hydroxydopamine (6-OHDA), and investigated its function in cytotoxicity induced by 6-OHDA. We predicted a potential miR-107 target and assessed its role in miR-107 mediated effects and explored the intracellular signaling pathways downstream of miR-107. Finally, we assessed the function of miR-107 in the mouse model insulted by 6-OHDA. We found that 6-OHDA suppressed miR-107 expression and miR-107 played neuroprotective effects against 6-OHDA mediated cytotoxicity. We showed that miR-107 targeted programmed cell death 10 (PDCD10). MiR-107 suppressed PDCD10 expression and exogenous expression of PDCD10 inhibited miR-107 mediated neuroprotection. Additionally, we found that Notch signal pathway was downstream of miR-107/PDCD10. Finally, we found that 6-OHDA treatment suppressed miR-107 in mice and restoration of miR-107 alleviated motor disorder in the mouse model. Our study shows that miR-107 plays important neuroprotective roles against neurotoxicity both in vitro and in vivo by inhibiting PDCD10. Our findings confirm that miR-107 may be involved in AD pathogenesis and may be a therapeutic target for the treatment of AD-related impairments.

Automated Brain MRI Volumetry Differentiates Early Stages of Alzheimer's Disease From Normal Aging.

As an enrichment strategy supplemented by the diagnostic framework of subjective cognitive decline (SCD), SCD plus identifies features that may increase the likelihood of including future-Alzheimer's disease (AD) patients. This study aimed to identify the shared and distinct atrophy patterns between patients specified by SCD plus and amnestic mild cognitive impairment (aMCI, a prodromal stage of AD) and to investigate the extent that automated brain magnetic resonance imaging (MRI) volumetry can differentiate patients with SCD from normal control (NC) participants and patients with aMCI. We acquired structural MRI brain scans from 44 patients with aMCI, 40 patients with SCD (who met the major criteria of SCD plus), and 48 NC participants. Automatic brain segmentation was performed to quantify the volumetric measures of cognitive-relevant areas. These volumetric measures were compared across the 3 groups with analysis of variance. In addition, we performed support vector machine analyses using volumetric measures of single regions or multiple regions to further evaluate the sensitivity of automated brain volumetry in differentiating a specific group from another. The atrophy patterns in patients with aMCI and SCD were similar. Using the regional volumetric measures, we achieved high performance in differentiating aMCI and SCD from NCs (average classification accuracy [ACC] > 90%). However, the performance was not ideal when differentiating aMCI from SCD (ACC < 63%). In conclusion, patients with SCD specified by SCD plus presented similar atrophy patterns as patients with aMCI, which was distinguishable from NC participants. Future studies should aim to associate the atrophy patterns of SCD with possible conversion to aMCI or AD in a longitudinal design.

Oxytocin inhibits ox-LDL-induced adhesion of monocytic THP-1 cells to human brain microvascular endothelial cells.

The attachment of monocytes to human brain microvascular endothelial cells (HBMVEs) caused by oxidized low-density lipoprotein (ox-LDL) is associated with an early event and the pathological progression of cerebrovascular diseases. Oxytocin (OT) is a human peptide hormone that is traditionally used as a medication to facilitate childbirth. However, little information is available regarding the physiological function of OT in brain endothelial dysfunction. In the present study, our results indicate that the oxytocin receptor (OTR) was expressed in human brain microvascular endothelial cells (HBMVEs) and was upregulated in response to ox-LDL in a concentration-dependent manner. Notably, OT significantly suppressed ox-LDL-induced attachment of THP-1 monocytes to HBMVEs. Furthermore, we found that OT reduced the expression of adhesion molecules, such as VCAM-1 and E-selectin. Interestingly, it was shown that OT could restore ox-LDL-induced reduction of KLF4 in HBMVEs. Importantly, knockdown of KLF4 abolished the inhibitory effects of OT on ox-LDL-induced expressions of VCAM-1 and E-selectin as well as the adhesion of human monocytic THP-1 cells to endothelial HBMVEs. Mechanistically, we found that the stimulatory effects of OT on KLF4 expression are mediated by the MEK5/MEF2A pathway.

Motility and stem cell properties induced by the epithelial-mesenchymal transition require destabilization of lipid rafts.

The Epithelial-Mesenchymal Transition (EMT) is a developmental program that provides cancer cells with the characteristics necessary for metastasis, including increased motility and stem cell properties. The cellular and molecular mechanisms underlying this process are not yet fully understood, hampering efforts to develop therapeutics. In recent years, it has become apparent that EMT is accompanied by wholesale changes in diverse signaling pathways that are initiated by proteins at the plasma membrane (PM). The PM contains thousands of lipid and protein species that are dynamically and spatially organized into lateral membrane domains, an example of which are lipid rafts. Since one of the major functions of rafts is modulation of signaling originating at the PM, we hypothesized that the signaling changes occurring during an EMT are associated with alterations in PM organization. To test this hypothesis, we used Giant Plasma Membrane Vesicles (GPMVs) to study the organization of intact plasma membranes isolated from live cells. We observed that induction of EMT significantly destabilized lipid raft domains. Further, this reduction in stability was crucial for the maintenance of the stem cell phenotype and EMT-induced remodeling of PM-orchestrated pathways. Exogenously increasing raft stability by feeding cells with ω-3 polyunsaturated fatty acid docosahexaenoic acid (DHA) repressed these phenotypes without altering EMT markers, and inhibited the metastatic capacity of breast cancer cells. Hence, modulating raft properties regulates cell phenotype, suggesting a novel approach for targeting the impact of EMT in cancer.

Immunoreceptor TIGIT inhibits the cytotoxicity of human cytokine-induced killer cells by interacting with CD155.

T cell Ig and ITIM domain (TIGIT) is a newly identified inhibitory receptor expressed on T and natural killer (NK) cells. Cytokine-induced killer (CIK) cells express CD3 and CD56 molecules, and share functional properties with both NK and T cells. However, it remains unknown whether TIGIT is expressed in CIK cells. Here, we show that TIGIT is expressed by CIK cells and interacts with CD155. By blocking TIGIT using an anti-TIGIT functional antibody, we demonstrate that CIK cells display increased proliferation; higher cytotoxic targeting of tumor cells expressing CD155; and higher expression of interferon-γ (IFN-γ), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α). Furthermore, increases in IFN-γ and cytotoxicity by blockade of TIGIT were reduced by blocking DNAX accessory molecule-1 (DNAM-1) signaling, implying that TIGIT exerts immunosuppressive effects by competing with DNAM-1 for the same ligand, CD155. Our results provide evidence that blockade of TIGIT may be a novel strategy to improve the cytotoxic activity of CIK cells.

Candidate Antimetastasis Drugs Suppress the Metastatic Capacity of Breast Cancer Cells by Reducing Membrane Fluidity.

Despite the high mortality from metastatic cancer, therapeutic targets to prevent metastasis are limited. Efforts to identify genetic aberrations that predispose tumors to metastasis have been mostly unsuccessful. To understand the nature of candidate targets for metastatic disease, we performed an in silico screen to identify drugs that can inhibit a gene expression signature associated with epithelial-mesenchymal transition (EMT). Compounds discovered through this method, including those previously identified, appeared to restrict metastatic capacity through a common mechanism, the ability to modulate the fluidity of cell membranes. Treatment of breast cancer cell lines with the putative antimetastasis agents reduced membrane fluidity, resulting in decreased cell motility, stem cell-like properties, and EMT in vitro, and the drugs also inhibited spontaneous metastasis in vivo When fluidity was unchanged, the antimetastasis compounds could no longer restrict metastasis, indicating a causal association between fluidity and metastasis. We further demonstrate that fluidity can be regulated by cellular cholesterol flux, as the cholesterol efflux channel ABCA1 potentiated metastatic behaviors in vitro and in vivo The requirement for fluidity was further supported by the finding in breast cancer patients that ABCA1 was overexpressed in 41% of metastatic tumors, reducing time to metastasis by 9 years. Collectively, our findings reveal increased membrane fluidity as a necessary cellular feature of metastatic potential that can be controlled by many currently available drugs, offering a viable therapeutic opportunity to prevent cancer metastasis. Cancer Res; 76(7); 2037-49. ©2016 AACR.

Tip110 protein binds to unphosphorylated RNA polymerase II and promotes its phosphorylation and HIV-1 long terminal repeat transcription.

Transcription plays an important role in both HIV-1 gene expression and replication and mandates complicated but coordinated interactions between the host and virus. Our previous studies have shown that an HIV-1 Tat-interacting protein of 110 kDa, Tip110, binds to and enhances Tat function in Tat-mediated HIV-1 gene transcription and replication (Liu, Y., Li, J., Kim, B. O., Pace, B. S., and He, J. J. (2002) HIV-1 Tat protein-mediated transactivation of the HIV-1 long terminal repeat promoter is potentiated by a novel nuclear Tat-interacting protein of 110 kDa, Tip110. J. Biol. Chem. 277, 23854-23863). However, the underlying molecular mechanisms by which this takes place were not understood. In this study, we demonstrated that Tip110 bound to unphosphorylated RNA polymerase II (RNAPII) in a direct and specific manner. In addition, we detected Tip110 at the HIV-1 long terminal repeat (LTR) promoter and found that Tip110 expression was associated with increased phosphorylation of serine 2 of the heptapeptide repeats within the RNAPII C-terminal domain and increased recruitment of positive transcription elongation factor b to the LTR promoter. Consistent with these findings, we showed that Tip110 interaction with Tat directly enhanced transcription elongation of the LTR promoter. Taken together, these findings have provided additional and mechanistic evidence to support Tip110 function in HIV-1 transcription.

SIMPL enhancement of tumor necrosis factor-α dependent p65-MED1 complex formation is required for mammalian hematopoietic stem and progenitor cell function.

Significant insight into the signaling pathways leading to activation of the Rel transcription factor family, collectively termed NF-κB, has been gained. Less well understood is how subsets of NF-κB-dependent genes are regulated in a signal specific manner. The SIMPL protein (signaling molecule that interacts with mouse pelle-like kinase) is required for full Tumor Necrosis Factor-α (TNFα) induced NF-κB activity. We show that SIMPL is required for steady-state hematopoiesis and the expression of a subset of TNFα induced genes whose products regulate hematopoietic cell activity. To gain insight into the mechanism through which SIMPL modulates gene expression we focused on the Tnf gene, an immune response regulator required for steady-state hematopoiesis. In response to TNFα SIMPL localizes to the Tnf gene promoter where it modulates the initiation of Tnf gene transcription. SIMPL binding partners identified by mass spectrometry include proteins involved in transcription and the interaction between SIMPL and MED1 was characterized in more detail. In response to TNFα, SIMPL is found in p65-MED1 complexes where SIMPL enhances p65/MED1/SIMPL complex formation. Together our results indicate that SIMPL functions as a TNFα-dependent p65 co-activator by facilitating the recruitment of MED1 to p65 containing transcriptional complexes to control the expression of a subset of TNFα-induced genes.

FIT interacts with AtbHLH38 and AtbHLH39 in regulating iron uptake gene expression for iron homeostasis in Arabidopsis.

Iron is an essential element for plant growth and development. Iron homeostasis in plants is tightly regulated at both transcriptional and posttranscriptional level. Several bHLH transcription factors involved in iron homeostasis have been identified recently. However, their regulatory mechanisms remain unknown. In this work, we demonstrate that the transcription factor FIT interacted with AtbHLH38 and AtbHLH39 and directly conferred the expression regulation of iron uptake genes for iron homeostasis in Arabidopsis. Yeast two-hybrid analysis and transient expression in Arabidopsis protoplasts showed that AtbHLH38 or AtbHLH39 interacted with FIT, a central transcription factor involved in iron homeostasis in Arabidopsis. Expression of FIT/AtbHLH38 or FIT/AtbHLH39 in yeast cells activated GUS expression driven by ferric chelate reductase (FRO2) and ferrous transporter (IRT1) promoters. Overexpression of FIT with either AtbHLH38 or AtbHLH39 in plants converted the expression of the iron uptake genes FRO2 and IRT1 from induced to constitutive. Further analysis revealed that FRO2 and IRT1 were not regulated at the posttranscriptional level in these plants because IRT1 protein accumulation and high ferric chelate reductase activity were detected in the overexpression plants under both iron deficiency and iron sufficiency. The double overexpression plants accumulated more iron in their shoots than wild type or the plants overexpressing either AtbHLH38, AtbHLH39 or FIT. Our data support that ferric-chelate reductase FRO2 and ferrous-transporter IRT1 are the targets of the three transcription factors and the transcription of FRO2 and IRT1 is directly regulated by a complex of FIT/AtbHLH38 or FIT/AtbHLH39.