RESUMO
BACKGROUND: Diabetic kidney disease (DKD) stands as the predominant cause of chronic kidney disease and end-stage kidney disease. Its diverse range of manifestations complicates the treatment approach for patients. Although kidney biopsy is considered the gold standard for diagnosis, it lacks precision in predicting the progression of kidney dysfunction. Herein, we addressed whether the presence of glomerular crescents is linked to the outcomes in patients with biopsy-confirmed type 2 DKD. METHODS: We performed a retrospective evaluation, involving 327 patients diagnosed with biopsy-confirmed DKD in the context of type 2 diabetes, excluding cases with other glomerular diseases, from nine tertiary hospitals. Hazard ratios (HRs) were calculated using a Cox regression model to assess the risk of kidney disease progression, defined as either ≥ 50% decrease in estimated glomerular filtration rates or the development of end-stage kidney disease, based on the presence of glomerular crescents. RESULTS: Out of the 327 patients selected, ten patients had glomerular crescents observed in their biopsied tissues. Over the follow-up period (median of 19 months, with a maximum of 18 years), the crescent group exhibited a higher risk of kidney disease progression than the no crescent group, with an adjusted HR of 2.82 (1.32-6.06) (P = 0.008). The presence of heavy proteinuria was associated with an increased risk of developing glomerular crescents. CONCLUSION: The presence of glomerular crescents is indeed linked to the progression of type 2 DKD. Therefore, it is important to determine whether there is an additional immune-mediated glomerulonephritis requiring immunomodulation, and it may be prudent to monitor the histology and repeat a biopsy.
Assuntos
Diabetes Mellitus Tipo 2 , Nefropatias Diabéticas , Progressão da Doença , Glomérulos Renais , Humanos , Nefropatias Diabéticas/patologia , Estudos Retrospectivos , Masculino , Feminino , Pessoa de Meia-Idade , Diabetes Mellitus Tipo 2/complicações , Glomérulos Renais/patologia , Idoso , Taxa de Filtração Glomerular , Estudos de Coortes , Biópsia , Falência Renal Crônica , Fatores de RiscoRESUMO
BACKGROUND: Sepsis is an important cause of acute kidney injury in intensive care unit patients, accounting for 15% to 20% of renal replacement therapy prescriptions. The neutrophil-lymphocyte ratio (NLR), a marker of systemic inflammation and immune response, was previously associated with the mortality rate in multiple conditions. Herein, we aimed to examine how the NLR relates to the mortality rate in septic acute kidney injury patients requiring continuous renal replacement therapy (CRRT). METHODS: The NLRs of 6 and 18 were used for dividing NLRs into three groups and, thus, were set higher than those in previous studies accounting for steroid use in sepsis. Cox proportional hazard models were used to calculate hazard ratios of mortality outcomes before and after matching their propensity scores. RESULTS: A total of 798 septic acute kidney injury patients requiring CRRT were classified into three NLR groups (low, <6 [n = 277]; medium, ≥6 and <18 [n = 115], and high, ≥18 [n = 406], respectively). The in-hospital mortality rates per group were 83.4%, 74.8%, and 70.4%, respectively (p < 0.001). Per the univariable Cox survival analysis after propensity score matching, a high NLR was related to approximately 24% reduced mortality. The survival benefit of the high NLR group compared with the other two groups remained consistent across all subgroups, showing any p for interactions of >0.05. CONCLUSION: A high NLR is associated with better clinical outcomes, such as low mortality, in septic acute kidney injury patients undergoing CRRT.
RESUMO
Toxin- and drug-induced tubulointerstitial nephritis (TIN), characterized by interstitial infiltration of immune cells, frequently necessitates dialysis for patients due to irreversible fibrosis. However, agents modulating interstitial immune cells are lacking. Here, we addressed whether the housekeeping enzyme glutamyl-prolyl-transfer RNA synthetase 1 (EPRS1), responsible for attaching glutamic acid and proline to transfer RNA, modulates immune cell activity during TIN and whether its pharmacological inhibition abrogates fibrotic transformation. The immunological feature following TIN induction by means of an adenine-mixed diet was infiltration of EPRS1high T cells, particularly proliferating T and γδ T cells. The proliferation capacity of both CD4+ and CD8+ T cells, along with interleukin-17 production of γδ T cells, was higher in the kidneys of TIN-induced Eprs1+/+ mice than in the kidneys of TIN-induced Eprs1+/- mice. This discrepancy contributed to the fibrotic amelioration observed in kidneys of Eprs1+/- mice. TIN-induced fibrosis was also reduced in Rag1-/- mice adoptively transferred with Eprs1+/- T cells compared to the Rag1-/- mice transferred with Eprs1+/+ T cells. The use of an EPRS1-targeting small molecule inhibitor (bersiporocin) under clinical trials to evaluate its therapeutic potential against idiopathic pulmonary fibrosis alleviated immunofibrotic aggravation in TIN. EPRS1 expression was also observed in human kidney tissues and blood-derived T cells, and high expression was associated with worse patient outcomes. Thus, EPRS1 may emerge as a therapeutic target in toxin- and drug-induced TIN, modulating the proliferation and activity of infiltrated T cells.
Assuntos
Aminoacil-tRNA Sintetases , Nefrite Intersticial , Insuficiência Renal , Animais , Humanos , Camundongos , Aminoacil-tRNA Sintetases/metabolismo , Linfócitos T CD8-Positivos , Proliferação de Células , Fibrose , Proteínas de Homeodomínio , Nefrite Intersticial/induzido quimicamente , Nefrite Intersticial/genética , Nefrite Intersticial/tratamento farmacológicoRESUMO
Background: Intradialytic hypotension (IDH) is a critical complication related to worse outcomes in patients undergoing maintenance hemodialysis. Herein, we addressed the impact of IDH on mortality and other outcomes in patients with severe acute kidney injury (AKI) requiring intermittent hemodialysis. Methods: We retrospectively reviewed 1,009 patients who underwent intermittent hemodialysis due to severe AKI. IDH was defined as either dialysis discontinuation due to hemodynamic instability or a decrease in systolic blood pressure (BP) of ≥30 mmHg, with or without a nadir systolic BP of <90 mmHg during the first session. The primary outcome was all-cause mortality, and transfer to the intensive care unit (ICU) due to unstable status was additionally analyzed. Hazard ratios (HRs) of outcomes were calculated using a Cox regression model after adjusting for multiple variables. Risk factors for IDH were evaluated using a logistic regression model. Results: IDH occurred in 449 patients (44.5%) during the first hemodialysis session. Patients with IDH had a higher mortality rate than those without IDH (40% vs. 23%; HR, 1.30; 95% confidence interval [CI], 1.02-1.65). The rate of ICU transfer was higher in patients experiencing IDH than in those without IDH (17% vs. 11%; HR, 1.43; 95% CI, 1.02-2.02). Factors such as old age, high BP and pulse rate, active malignancy, cirrhosis, and hypoalbuminemia were associated with an increased risk of IDH episodes. Conclusion: The occurrence of IDH is associated with worse outcomes in patients with AKI requiring intermittent hemodialysis. Therefore, careful monitoring and early intervention of IDH may be necessary in this patient subset.
RESUMO
OBJECTIVES: Automatic detection of atrial fibrillation and flutter (AF/AFL) is a significant concern in preventing stroke and mitigating hemodynamic instability. Herein, we developed a Transformer-based deep learning model for AF/AFL segmentation in single-lead electrocardiograms (ECGs) by self-supervised learning with masked signal modeling (MSM). MATERIALS AND METHODS: We retrieved data from 11 open-source databases on PhysioNet; 7 of these databases included labeled ECGs, while the other 4 were without labels. Each database contained ECG recordings with durations of ≥30 s. A total of 24 intradialytic ECGs with paroxysmal AF/AFL during 4 h of hemodialysis sessions at Seoul National University Hospital were used for external validation. The model was pretrained by predicting masked areas of ECG signals and fine-tuned by predicting AF/AFL areas. Cross-database validation was used for evaluation, and the intersection over union (IOU) was used as a main performance metric in external database validation. RESULTS: In the 7 labeled databases, the areas marked as AF/AFL constituted 41.1% of the total ECG signals, ranging from 0.19% to 51.31%. In the evaluation per ECG segment, the model achieved IOU values of 0.9254 and 0.9477 for AF/AFL segmentation and other segmentation tasks, respectively. When applied to intradialytic ECGs with paroxysmal AF/AFL, the IOUs for the segmentation of AF/AFL and non-AF/AFL were 0.9896 and 0.9650, respectively. Model performance by different training procedure indicated that pretraining with MSM and the application of an appropriate masking ratio both contributed to the model performance. It also showed higher IOUs of AF/AFL labels than in previous studies when training and test databases were matched. CONCLUSION: The present model with self-supervised learning by MSM performs robustly in segmenting AF/AFL.
Assuntos
Fibrilação Atrial , Flutter Atrial , Acidente Vascular Cerebral , Humanos , Fibrilação Atrial/diagnóstico , Flutter Atrial/diagnóstico , Eletrocardiografia , Aprendizado de Máquina SupervisionadoRESUMO
Both intradialytic hypotension (IDH) and hypertension (IDHTN) are associated with poor outcomes in hemodialysis patients, but a model predicting dual outcomes in real-time has never been developed. Herein, we developed an explainable deep learning model with a sequence-to-sequence-based attention network to predict both of these events simultaneously. We retrieved 302,774 hemodialysis sessions from the electronic health records of 11,110 patients, and these sessions were split into training (70%), validation (10%), and test (20%) datasets through patient randomization. The outcomes were defined when nadir systolic blood pressure (BP) < 90 mmHg (termed IDH-1), a decrease in systolic BP ≥ 20 mmHg and/or a decrease in mean arterial pressure ≥ 10 mmHg (termed IDH-2), or an increase in systolic BP ≥ 10 mmHg (i.e., IDHTN) occurred within 1 h. We developed a temporal fusion transformer (TFT)-based model and compared its performance in the test dataset, including receiver operating characteristic curve (AUROC) and area under the precision-recall curves (AUPRC), with those of other machine learning models, such as recurrent neural network, light gradient boosting machine, random forest, and logistic regression. Among all models, the TFT-based model achieved the highest AUROCs of 0.953 (0.952-0.954), 0.892 (0.891-0.893), and 0.889 (0.888-0.890) in predicting IDH-1, IDH-2, and IDHTN, respectively. The AUPRCs in the TFT-based model for these outcomes were higher than the other models. The factors that contributed the most to the prediction were age and previous session, which were time-invariant variables, as well as systolic BP and elapsed time, which were time-varying variables. The present TFT-based model predicts both IDH and IDHTN in real time and offers explainable variable importance.
Assuntos
Aprendizado Profundo , Hipertensão , Hipotensão , Falência Renal Crônica , Humanos , Hipotensão/etiologia , Hipertensão/epidemiologia , Hipertensão/etiologia , Diálise Renal/efeitos adversos , Pressão Sanguínea , Falência Renal Crônica/etiologiaRESUMO
Diabetic nephropathy (DN) in the context of type 2 diabetes is the leading cause of end-stage renal disease (ESRD) in the United States. DN is graded based on glomerular morphology and has a spatially heterogeneous presentation in kidney biopsies that complicates pathologists' predictions of disease progression. Artificial intelligence and deep learning methods for pathology have shown promise for quantitative pathological evaluation and clinical trajectory estimation; but, they often fail to capture large-scale spatial anatomy and relationships found in whole slide images (WSIs). In this study, we present a transformer-based, multi-stage ESRD prediction framework built upon nonlinear dimensionality reduction, relative Euclidean pixel distance embeddings between every pair of observable glomeruli, and a corresponding spatial self-attention mechanism for a robust contextual representation. We developed a deep transformer network for encoding WSI and predicting future ESRD using a dataset of 56 kidney biopsy WSIs from DN patients at Seoul National University Hospital. Using a leave-one-out cross-validation scheme, our modified transformer framework outperformed RNNs, XGBoost, and logistic regression baseline models, and resulted in an area under the receiver operating characteristic curve (AUC) of 0.97 (95% CI: 0.90-1.00) for predicting two-year ESRD, compared with an AUC of 0.86 (95% CI: 0.66-0.99) without our relative distance embedding, and an AUC of 0.76 (95% CI: 0.59-0.92) without a denoising autoencoder module. While the variability and generalizability induced by smaller sample sizes are challenging, our distance-based embedding approach and overfitting mitigation techniques yielded results that suggest opportunities for future spatially aware WSI research using limited pathology datasets.
RESUMO
Background: The heterogeneous phenotype of diabetic nephropathy (DN) from type 2 diabetes complicates appropriate treatment approaches and outcome prediction. Kidney histology helps diagnose DN and predict its outcomes, and an artificial intelligence (AI)-based approach will maximize clinical utility of histopathological evaluation. Herein, we addressed whether AI-based integration of urine proteomics and image features improves DN classification and its outcome prediction, altogether augmenting and advancing pathology practice. Methods: We studied whole slide images (WSIs) of periodic acid-Schiff-stained kidney biopsies from 56 DN patients with associated urinary proteomics data. We identified urinary proteins differentially expressed in patients who developed end-stage kidney disease (ESKD) within two years of biopsy. Extending our previously published human-AI-loop pipeline, six renal sub-compartments were computationally segmented from each WSI. Hand-engineered image features for glomeruli and tubules, and urinary protein measurements, were used as inputs to deep-learning frameworks to predict ESKD outcome. Differential expression was correlated with digital image features using the Spearman rank sum coefficient. Results: A total of 45 urinary proteins were differentially detected in progressors, which was most predictive of ESKD (AUC=0.95), while tubular and glomerular features were less predictive (AUC=0.71 and AUC=0.63, respectively). Accordingly, a correlation map between canonical cell-type proteins, such as epidermal growth factor and secreted phosphoprotein 1, and AI-based image features was obtained, which supports previous pathobiological results. Conclusions: Computational method-based integration of urinary and image biomarkers may improve the pathophysiological understanding of DN progression as well as carry clinical implications in histopathological evaluation.
RESUMO
BACKGROUND: Despite efforts to treat critically ill patients who require continuous renal replacement therapy (CRRT) due to acute kidney injury (AKI), their mortality risk remains high. This condition may be attributable to complications of CRRT, such as arrhythmias. Here, we addressed the occurrence of ventricular tachycardia (VT) during CRRT and its relationship with patient outcomes. METHODS: This study retrospectively enrolled 2,397 patients who started CRRT due to AKI from 2010 to 2020 at Seoul National University Hospital in Korea. The occurrence of VT was evaluated from the initiation of CRRT until weaning from CRRT. The odds ratios (ORs) of mortality outcomes were measured using logistic regression models after adjustment for multiple variables. RESULTS: VT occurred in 150 patients (6.3%) after starting CRRT. Among them, 95 cases were defined as sustained VT (i.e., lasting ≥30 seconds), and the other 55 cases were defined as non-sustained VT (i.e., lasting <30 seconds). The occurrence of sustained VT was associated with a higher mortality rate than a nonoccurrence (OR, 2.04 and 95% confidence interval [CI], 1.23-3.39 for the 30- day mortality; OR, 4.06 and 95% CI, 2.04-8.08 for the 90-day mortality). The mortality risk did not differ between patients with non-sustained VT and nonoccurrence. A history of myocardial infarction, vasopressor use, and certain trends of blood laboratory findings (such as acidosis and hyperkalemia) were associated with the subsequent risk of sustained VT. CONCLUSION: Sustained VT occurrence after starting CRRT is associated with increased patient mortality. The monitoring of electrolytes and acid-base status during CRRT is essential because of its relationship with the risk of VT.
RESUMO
Diabetic nephropathy (DN) in the context of type 2 diabetes is the leading cause of end-stage renal disease (ESRD) in the United States. DN is graded based on glomerular morphology and has a spatially heterogeneous presentation in kidney biopsies that complicates pathologists' predictions of disease progression. Artificial intelligence and deep learning methods for pathology have shown promise for quantitative pathological evaluation and clinical trajectory estimation; but, they often fail to capture large-scale spatial anatomy and relationships found in whole slide images (WSIs). In this study, we present a transformer-based, multi-stage ESRD prediction framework built upon nonlinear dimensionality reduction, relative Euclidean pixel distance embeddings between every pair of observable glomeruli, and a corresponding spatial self-attention mechanism for a robust contextual representation. We developed a deep transformer network for encoding WSI and predicting future ESRD using a dataset of 56 kidney biopsy WSIs from DN patients at Seoul National University Hospital. Using a leave-one-out cross-validation scheme, our modified transformer framework outperformed RNNs, XGBoost, and logistic regression baseline models, and resulted in an area under the receiver operating characteristic curve (AUC) of 0.97 (95% CI: 0.90-1.00) for predicting two-year ESRD, compared with an AUC of 0.86 (95% CI: 0.66-0.99) without our relative distance embedding, and an AUC of 0.76 (95% CI: 0.59-0.92) without a denoising autoencoder module. While the variability and generalizability induced by smaller sample sizes are challenging, our distance-based embedding approach and overfitting mitigation techniques yielded results that sugest opportunities for future spatially aware WSI research using limited pathology datasets.
RESUMO
BACKGROUND: Several studies suggest improved outcomes for patients with kidney disease who consult a nephrologist. However, it remains undetermined whether a consultation with a nephrologist is related to a survival benefit after starting continuous renal replacement therapy (CRRT) due to acute kidney injury (AKI). METHODS: Data from 2,397 patients who started CRRT due to severe AKI at Seoul National University Hospital, Korea between 2010 and 2020 were retrospectively collected. The patients were divided into two groups according to whether they underwent a nephrology consultation regarding the initiation and maintenance of CRRT. The Cox proportional hazards model was used to calculate the hazard ratio (HR) of mortality during admission to the intensive care unit after adjusting for multiple variables. RESULTS: A total of 2,153 patients (89.8%) were referred to nephrologists when starting CRRT. The patients who underwent a nephrology consultation had a lower mortality rate than those who did not have a consultation (HR = 0.47 [0.40-0.56]; P < 0.001). Subsequently, patients who had nephrology consultations were divided into two groups (i.e., early and late) according to the timing of the consultation. Both patients with early and late consultation had lower mortality rates than patients without consultations, with HRs of 0.45 (0.37-0.54) and 0.51 (0.42-0.61), respectively. CONCLUSIONS: Consultation with a nephrologist may contribute to a survival benefit after starting CRRT for AKI.
Assuntos
Injúria Renal Aguda , Terapia de Substituição Renal Contínua , Humanos , Nefrologistas , Estudos Retrospectivos , Terapia de Substituição Renal , Injúria Renal Aguda/terapia , Encaminhamento e ConsultaRESUMO
BACKGROUND: Hyperlactatemia occurs frequently in critically ill patients, and this pathologic condition leads to worse outcomes in several disease subsets. Herein, we addressed whether hyperlactatemia is associated with the risk of mortality in patients undergoing continuous renal replacement therapy (CRRT) due to acute kidney injury. METHODS: A total of 1,661 patients who underwent CRRT for severe acute kidney injury were retrospectively reviewed between 2010 and 2020. The patients were categorized according to their serum lactate levels, such as high (≥ 7.6 mmol/l), moderate (2.1-7.5 mmol/l) and low (≤ 2 mmol/l), at the time of CRRT initiation. The hazard ratios (HRs) for the risk of in-hospital mortality were calculated with adjustment of multiple variables. The increase in the area under the receiver operating characteristic curve (AUROC) for the mortality risk was evaluated after adding serum lactate levels to the Sequential Organ Failure Assessment (SOFA) and the Acute Physiology and Chronic Health Evaluation (APACHE) II score-based models. RESULTS: A total of 802 (48.3%) and 542 (32.6%) patients had moderate and high lactate levels, respectively. The moderate and high lactate groups had a higher risk of mortality than the low lactate group, with HRs of 1.64 (1.22-2.20) and 4.18 (2.99-5.85), respectively. The lactate-enhanced models had higher AUROCs than the models without lactates (0.764 vs. 0.702 for SOFA score; 0.737 vs. 0.678 for APACHE II score). CONCLUSIONS: Hyperlactatemia is associated with mortality outcomes in patients undergoing CRRT for acute kidney injury. Serum lactate levels may need to be monitored in this patient subset.
Assuntos
Injúria Renal Aguda , Terapia de Substituição Renal Contínua , Hiperlactatemia , Humanos , Terapia de Substituição Renal Contínua/efeitos adversos , Estudos Retrospectivos , Hiperlactatemia/complicações , APACHE , Ácido Láctico , Terapia de Substituição Renal , Estado Terminal/terapia , PrognósticoRESUMO
BACKGROUND: The transition of dialysis modalities from continuous renal replacement therapy (CRRT) to intermittent hemodialysis (iHD) is frequently conducted during the recovery phase of critically ill patients with acute kidney injury. Herein, we addressed the occurrence of intradialytic hypotension (IDH) after this transition, and its association with the mortality risk. METHODS: A total of 541 patients with acute kidney injury who attempted to transition from CRRT to iHD at Seoul National University Hospital, Korea from 2010 to 2020 were retrospectively collected. IDH was defined as a discontinuation of dialysis because of hemodynamic instability plus a nadir systolic blood pressure <90 mm Hg or a decrease in systolic blood pressure ≥30 mm Hg during the first session of iHD. Odds ratios (ORs) of outcomes, such as in-hospital mortality and weaning from RRT, were measured using a logistic regression model after adjusting for multiple variables. RESULTS: IDH occurred in 197 patients (36%), and their mortality rate (44%) was higher than that of those without IDH (19%; OR, 2.64; 95% confidence interval [CI], 1.70-4.08). For patients exhibiting IDH, the iHD sessions delayed successful weaning from RRT (OR, 0.62; 95% CI, 0.43-0.90) compared with sessions on those without IDH. Factors such as low blood pressure, high pulse rate, low urine output, use of mechanical ventilations and vasopressors, and hypoalbuminemia were associated with IDH risk. CONCLUSIONS: IDH occurrence following the transition from CRRT to iHD is associated with high mortality and delayed weaning from RRT.
RESUMO
BACKGROUND: Acidosis frequently occurs in severe acute kidney injury (AKI), and continuous renal replacement therapy (CRRT) can control this pathologic condition. Nevertheless, acidosis may be aggravated; thus, monitoring is essential after starting CRRT. Herein, we addressed the longitudinal trajectory of acidosis on CRRT and its relationship with worse outcomes. METHODS: The latent growth mixture model was applied to classify the trajectories of pH during the first 24 hours and those of C-reactive protein (CRP) after 24 hours on CRRT due to AKI (n = 1815). Cox proportional hazard models were used to calculate hazard ratios of all-cause mortality after adjusting multiple variables or matching their propensity scores. RESULTS: The patients could be classified into 5 clusters, including the normally maintained groups (1st cluster, pH = 7.4; and 2nd cluster, pH = 7.3), recovering group (3rd cluster with pH values from 7.2 to 7.3), aggravating group (4th cluster with pH values from 7.3 to 7.2), and ill-being group (5th cluster, pH < 7.2). The pH clusters had different trends of C-reactive protein (CRP) after 24 hours; the 1st and 2nd pH clusters had lower levels, but the 3rd to 5th pH clusters had an increasing trend of CRP. The 1st pH cluster had the best survival rates, and the 3rd to 5th pH clusters had the worst survival rates. This survival difference was significant despite adjusting for other variables or matching propensity scores. CONCLUSIONS: Initial trajectories of acidosis determine subsequent worse outcomes, such as mortality and inflammation, in patients undergoing CRRT due to AKI.
Assuntos
Acidose , Injúria Renal Aguda , Terapia de Substituição Renal Contínua , Humanos , Terapia de Substituição Renal , Proteína C-Reativa , Estudos Retrospectivos , Injúria Renal Aguda/terapia , Estado Terminal/terapiaRESUMO
R-peak detection is an essential step in analyzing electrocardiograms (ECGs). Previous deep learning models reported their performance primarily in a single database, and some models did not perform at the highest levels when applied to a database different from the testing database. To achieve high performances in cross-database validations, we developed a novel deep learning model for R-peak detection using stationary wavelet transform (SWT) and separable convolution. Three databases (i.e., the MIT-BIH Arrhythmia [MIT-BIH], the Institute of Cardiological Technics [INCART], and the QT) were used in both the training and testing models, and the MIT-BIH ST Change (MIT-BIH-ST), European ST-T, TELE and MIT-BIH Noise Stress Test (MIT-BIH-NST) databases were further used for testing. The detail coefficient of level 4 decomposition by SWT and the first derivative from filtered ECGs were used for model inputs, and the interval of 150 ms centered at marked peaks was used for labels. Separable convolution with atrous spatial pyramidal pooling was selected as the model's architecture, and noise-augmented waveforms of 5.69 s duration (2048 size in 360 Hz) were used in training. The model performance was evaluated using cross-database validation. The F1 scores of the peak detection model were 0.9994, 0.9985, and 0.9999 in the MIT-BIH, INCART, and QT databases, respectively. When the above three databases were pooled, the F1 scores were 0.9993 for fivefold cross-validation and 0.9991 for cross-database validation. The model performance remained high for MIT-BIH-ST, European ST-T, and TELE, with F1 scores of 0.9995, 0.9988, and 0.9790, respectively. The model performance when trained by severe noise augmentation increased for the MIT-BIH-NST database (F1 scores from 0.9504 to 0.9759) and decreased for the MIT-BIH database (F1 scores from 0.9994 to 0.9991). The present SWT and separable convolution-based model for R-peak detection yields a high performance even for cross-database validations.
Assuntos
Algoritmos , Eletrocardiografia , Humanos , Análise de Ondaletas , Arritmias Cardíacas/diagnóstico , Bases de Dados FactuaisRESUMO
BACKGROUND: Appropriate monitoring of intradialytic biosignals is essential to minimize adverse outcomes because intradialytic hypotension and arrhythmia are associated with cardiovascular risk in hemodialysis patients. However, a continuous monitoring system for intradialytic biosignals has not yet been developed. METHODS: This study investigated a cloud system that hosted a prospective, open-source registry to monitor and collect intradialytic biosignals, which was named the CONTINUAL (Continuous mOnitoriNg viTal sIgN dUring hemodiALysis) registry. This registry was based on real-time multimodal data acquisition, such as blood pressure, heart rate, electrocardiogram, and photoplethysmogram results. RESULTS: We analyzed session information from this system for the initial 8 months, including data for some cases with hemodynamic complications such as intradialytic hypotension and arrhythmia. CONCLUSION: This biosignal registry provides valuable data that can be applied to conduct epidemiological surveys on hemodynamic complications during hemodialysis and develop artificial intelligence models that predict biosignal changes which can improve patient outcomes.
RESUMO
Severe glomerular injury ultimately leads to tubulointerstitial fibrosis that determines patient outcome, but the immunological molecules connecting these processes remain undetermined. The present study addressed whether V-domain Ig suppressor of T cell activation (VISTA), constitutively expressed in kidney macrophages, plays a protective role in tubulointerstitial fibrotic transformation after acute antibody-mediated glomerulonephritis. After acute glomerular injury using nephrotoxic serum, tubules in the VISTA-deficient (Vsir-/-) kidney suffered more damage than those in WT kidneys. When interstitial immune cells were examined, the contact frequency of macrophages with infiltrated T cells increased and the immunometabolic features of T cells changed to showing high oxidative phosphorylation and fatty acid metabolism and overproduction of IFN-γ. The Vsir-/- parenchymal tissue cells responded to this altered milieu of interstitial immune cells as more IL-9 was produced, which augmented tubulointerstitial fibrosis. Blocking antibodies against IFN-γ and IL-9 protected the above pathological process in VISTA-depleted conditions. In human samples with acute glomerular injury (e.g., antineutrophil cytoplasmic autoantibody vasculitis), high VISTA expression in tubulointerstitial immune cells was associated with low tubulointerstitial fibrosis and good prognosis. Therefore, VISTA is a sentinel protein expressed in kidney macrophages that prevents tubulointerstitial fibrosis via the IFN-γ/IL-9 axis after acute antibody-mediated glomerular injury.
Assuntos
Injúria Renal Aguda/metabolismo , Antígenos B7/metabolismo , Interferon gama/metabolismo , Interleucina-9/metabolismo , Glomérulos Renais/metabolismo , Proteínas de Membrana/metabolismo , Transdução de Sinais , Injúria Renal Aguda/genética , Animais , Antígenos B7/genética , Fibrose , Humanos , Interferon gama/genética , Interleucina-9/genética , Glomérulos Renais/lesões , Proteínas de Membrana/genética , Camundongos , Camundongos KnockoutRESUMO
Histological image data and molecular profiles provide context into renal condition. Often, a biopsy is drawn to diagnose or monitor a suspected kidney problem. However, molecular profiles can go beyond a pathologist's ability to see and diagnose. Using AI, we computationally incorporated urinary proteomic profiles with microstructural morphology from renal biopsy to investigate new and existing molecular links to image phenotypes. We studied whole slide images of periodic acid-Schiff stained renal biopsies from 56 DN patients matched with 2,038 proteins measured from each patient's urine. Using Seurat, we identified differentially expressed proteins in patients that developed end-stage renal disease within 2 years of biopsy. Glomeruli, globally sclerotic glomeruli, and tubules were segmented from WSI using our previously published HAIL pipeline. For each glomerulus, 315 handcrafted digital image features were measured, and for tubules, 207 features. We trained fully connected networks to predict urinary protein measurements that were differentially expressed between patients who did/ did not progress to ESRD within 2 years of biopsy. The input to this network was either glomerular or tubular histomorphological features in biopsy. Trained network weights were used as a proxy to rank which morphological features correlated most highly with specific urinary proteins. We identified significant image feature-protein pairs by ranking network weights by magnitude. We also looked at which features on average were most significant in predicting proteins. For both glomeruli and tubules, RGB color values and variance in PAS+ areas (specifically basement membrane for tubules) were, on average, more predictive of molecular profiles than other features. There is a strong connection between molecular profile and image phenotype, which can be elucidated through computational methods. These discovered links can provide insight to disease pathways, and discover new factors contributing to incidence and progression.
RESUMO
[This corrects the article DOI: 10.2196/27177.].
RESUMO
BACKGROUND: Precise prediction of contrast media-induced acute kidney injury (CIAKI) is an important issue because of its relationship with poor outcomes. OBJECTIVE: Herein, we examined whether a deep learning algorithm could predict the risk of intravenous CIAKI better than other machine learning and logistic regression models in patients undergoing computed tomography (CT). METHODS: A total of 14,185 patients who were administered intravenous contrast media for CT at the preventive and monitoring facility in Seoul National University Hospital were reviewed. CIAKI was defined as an increase in serum creatinine of ≥0.3 mg/dL within 2 days or ≥50% within 7 days. Using both time-varying and time-invariant features, machine learning models, such as the recurrent neural network (RNN), light gradient boosting machine (LGM), extreme gradient boosting machine (XGB), random forest (RF), decision tree (DT), support vector machine (SVM), κ-nearest neighbors, and logistic regression, were developed using a training set, and their performance was compared using the area under the receiver operating characteristic curve (AUROC) in a test set. RESULTS: CIAKI developed in 261 cases (1.8%). The RNN model had the highest AUROC of 0.755 (0.708-0.802) for predicting CIAKI, which was superior to that obtained from other machine learning models. Although CIAKI was defined as an increase in serum creatinine of ≥0.5 mg/dL or ≥25% within 3 days, the highest performance was achieved in the RNN model with an AUROC of 0.716 (95% confidence interval [CI] 0.664-0.768). In feature ranking analysis, the albumin level was the most highly contributing factor to RNN performance, followed by time-varying kidney function. CONCLUSIONS: Application of a deep learning algorithm improves the predictability of intravenous CIAKI after CT, representing a basis for future clinical alarming and preventive systems.