Objective Heart failure(HF)complicated by acute kidney injury(AKI)significantly impacts patient outcomes,and it is crucial to make early predictions of short-term mortality.This study is focused on developing an interpretable machine learning model to enhance early prediction accuracy in such clinical scenarios.Methods This retrospective cohort study utilized data from the Medical Information Mart for Intensive Care Ⅳ(MIMIC-Ⅳ,version 2.0)database.Data from the first 24 hours after admission to the ICU were extracted and divided into a training set(70％)and a validation set(30％).We utilized the SHapley Additive exPlanation(SHAP)method to interpret the workings of an extreme gradient boosting(XGBoost)model and identify key prognostic factors.The XGBoost model's predictive ability was evaluated against three other machine learning models using the area under the curve(AUC)metric,and its interpretation was enhanced using the SHAP method.Results The study included 8 028 patients with HF complicated by AKI.The XGBoost model outperformed the other models,achieving an AUC of 0.93(95％confidence interval[CI]:0.78-0.94;accuracy=0.89),while neural network model showed the worst performance(AUC=0.79,95％CI:0.77-0.82;accuracy=0.82).Decision curve analysis showed the superior net benefit of the XGBoost model within the 9％to 60％threshold probabilities.SHAP analysis was performed to identify the top 20 predictors,with age(mean SHAP value 1.29)and Glasgow Coma Scale score(mean SHAP value 1.24)emerging as significant factors.Conclusions Our interpretable model offers an enhanced ability to predict mortality risk in HF patients with AKI in ICUs.This model can be used to assist in formulating effective treatment plans and optimizing resource allocation.

Objective Tunneled-cuffed catheters(TCCs)are frequently used for establishing hemodialysis access for maintenance hemodialysis in older patients with exhausted resources of peripheral vessels.Fibrin sheath formation around the catheter is one of the most common complications of long-term use of indwelling catheter,which may cause the malfunction of the catheter.In this study,we intend to compare the prognosis of two catheter replacement methods,in situ replacement and replacement through a fibrin sheath crevice,with both being assisted by balloon dilation,and to explore the optimal catheter replacement process.Methods A retrospective study was conducted with 52 patients who underwent a replacement of their TCCs.Among them,27 cases had their TCC replaced by the modified method of replacement through a fibrin sheath crevice and were referred to as the sheath crevice group,while 25 cases underwent in situ catheter replacement and were referred to as the in situ group.The primary outcome indicators included maximum blood flow in hemodialysis catheter and the urea clearance rate calculated by Kt/V values at the 1,3,and 6-month follow-ups.The secondary outcomes included dialyzer alarms being set off and catheter-related infections during follow-up.Results There was no significant difference between the general data of the two groups.There was no massive blood loss during the replacement procedure.Neither were there cardiac tamponade,catheter-associated infections,or other complications.Follow-ups were made 1,3,and 6 months after the replacement procedure.The sheath crevice group had higher catheter blood flow and Kt/V values at the 6-month follow-up than the in situ group did([241.85±9.62]mL/min vs.[234.40±11.21]mL/min,P=0.014 and 1.31±0.55 vs.1.27±0.49,P=0.005,respectively).During the follow-up process,access alarms were reported in 5 patients(three in the in situ group and two in the sheath crevice group)during dialysis.No catheter-associated infection occurred in either group.Conclusion The catheter replacement method of balloon dilation-assisted catheter insertion through a fibrin sheath crevice is safe and effective,resulting in better long-term catheter blood flow compared with that of in situ catheter replacement.

Conjunctival melanoma (CM) is a rare and fatal malignant eye tumor. In this study, we deciphered a novel anti-CM mechanism of a natural tetracyclic compound named as cucurbitacin B (CuB). We found that CuB remarkably inhibited the proliferation of CM cells including CM-AS16, CRMM1, CRMM2 and CM2005.1, without toxicity to normal cells. CuB can also induce CM cells G2/M cell cycle arrest. RNA-seq screening identified KIF20A, a key downstream effector of FOXM1 pathway, was abolished by CuB treatment. Further target identification by activity-based protein profiling chemoproteomic approach revealed that GRP78 is a potential target of CuB. Several lines of evidence demonstrated that CuB interacted with GRP78 and bound with a K _d value of 0.11 μmol/L. Furthermore, ATPase activity evaluation showed that CuB suppressed GRP78 both in human recombinant GRP78 protein and cellular lysates. Knockdown of the GRP78 gene significantly induced the downregulation of FOXM1 and related pathway proteins including KIF20A, underlying an interesting therapeutic perspective. Finally, CuB significantly inhibited tumor progression in NCG mice without causing obvious side effects in vivo. Taken together, our current work proved that GRP78-FOXM1-KIF20A as a promising pathway for CM therapy, and the traditional medicine CuB as a candidate drug to hinder this pathway.