Employing feature engineering strategies to improve the performance of machine learning algorithms on echocardiogram dataset.

Objectives: This study mainly uses machine learning (ML) to make predictions by inputting features during training and inference. The method of feature selection is an important factor affecting the accuracy of ML models, and the process includes data extraction, which is the collection of all data required for ML. It also needs to import the concept of feature engineering, namely, this study needs to label the raw data of the cardiac ultrasound dataset with one or more meaningful and informative labels so that the ML model can learn from it and predict more accurate target values. Therefore, this study will enhance the strategies of feature selection methods from the raw dataset, as well as the issue of data scrubbing. Methods: In this study, the ultrasound dataset was cleaned and critical features were selected through data standardization, normalization, and missing features imputation in the field of feature engineering. The aim of data scrubbing was to retain and select critical features of the echocardiogram dataset while making the prediction of the ML algorithm more accurate. Results: This paper mainly utilizes commonly used methods in feature engineering and finally selects four important feature values. With the ML algorithms available on the Azure platform, namely, Random Forest and CatBoost, a Voting Ensemble method is used as the training algorithm, and this study also uses visual tools to gain a clearer understanding of the raw data and to improve the accuracy of the predictive model. Conclusion: This paper emphasizes feature engineering, specifically on the cleaning and analysis of missing values in the raw dataset of echocardiography and the identification of important critical features in the raw dataset. The Azure platform is used to predict patients with a history of heart disease (individuals who have been under surveillance in the past three years and those who haven't). Through data scrubbing and preprocessing methods in feature engineering, the model can more accurately predict the future occurrence of heart disease in patients.

Is rehabilitation intervention during hospitalization enough for functional improvements in patients undergoing lumbar decompression surgery? A prospective randomized controlled study.

OBJECTIVE: Rehabilitation has been reported to improve pain and disability for patients after lumbar surgery. However, studies to investigate the rehabilitation intervention for lumbar decompression surgery during hospitalization are scarce. The aim of this study was to examine outcomes of perioperative rehabilitation intervention for patients who underwent lumbar decompression surgery (LDS). METHODS: Patients aged 18-65 years old who received their first LDS were randomized into the perioperative rehabilitation group (PG) or control group (CG). The PG received rehabilitation intervention during hospitalization for lumbar decompression surgery. Pain, functional capacity, Roland-Morris Disability Questionnaire (RMDQ), and Short-Form Health Survey (SF-12) were assessed on admission, at discharge, and at follow-ups one month, three months, and six months after surgery. Two-way repeated measures ANOVAs were used for statistical analysis. RESULTS: A total of 60 patients scheduled for decompression surgery for lumbar stenosis were enrolled into the study. After surgery, the PG showed significant pain relief and improvement of disability as well as quality of life, but there were no significant functional improvements compared with the CG. CONCLUSIONS: The findings of this study indicate that the rehabilitation intervention during hospitalization improves pain intensity as well as disability and quality of life, yet has limited effects on the functional performance over time up to six months post-surgery in patients who received LDS. The study suggest that rehabilitation interventions during hospitalization must include regular support for patient adherence to the intervention program and focus on task-oriented programs for lower extremities such as closed-chain exercises in functional postures.

Visualization of subunit interactions and ternary complexes of protein phosphatase 2A in mammalian cells.

Protein phosphatase 2A (PP2A) is a ubiquitous phospho-serine/threonine phosphatase that controls many diverse cellular functions. The predominant form of PP2A is a heterotrimeric holoenzyme consisting of a scaffolding A subunit, a variable regulatory B subunit, and a catalytic C subunit. The C subunit also associates with other interacting partners, such as α4, to form non-canonical PP2A complexes. We report visualization of PP2A complexes in mammalian cells. Bimolecular fluorescence complementation (BiFC) analysis of PP2A subunit interactions demonstrates that the B subunit plays a key role in directing the subcellular localization of PP2A, and confirms that the A subunit functions as a scaffold in recruiting the B and C subunits to form a heterotrimeric holoenzyme. BiFC analysis also reveals that α4 promotes formation of the AC core dimer. Furthermore, we demonstrate visualization of specific ABC holoenzymes in cells by combining BiFC and fluorescence resonance energy transfer (BiFC-FRET). Our studies not only provide direct imaging data to support previous biochemical observations on PP2A complexes, but also offer a promising approach for studying the spatiotemporal distribution of individual PP2A complexes in cells.

Comparison of intracerebral hemorrhage and subarachnoid hemorrhage in patients with autosomal-dominant polycystic kidney disease.

BACKGROUND/AIMS: Subarachnoid hemorrhage (SAH) and intracerebral hemorrhage (ICH) are two subtypes of hemorrhagic stroke that may cause severe complications in patients with autosomal-dominant polycystic kidney disease (ADPKD). The differences in clinical features between SAH and ICH associated with ADPKD are not known. METHODS: Among 647 ADPKD patients hospitalized between 1997 and 2007 in our hospital, 11 with ICH (1.7%) and 6 with SAH (0.9%) were identified. RESULTS: Patients with SAH were significantly younger than patients with ICH (39 +/- 6 vs. 57 +/- 15 years, p = 0.013). The systolic blood pressure on admission was significantly higher in patients with ICH (194 +/- 26 vs. 145 +/- 18 mm Hg, p = 0.001). Two patients (18.2%) with ICH died after a first episode, 6 had a second episode, and 2 had a third episode. Two patients (33.3%) with SAH died after a first episode but the survivors had no recurrence during follow-up. The 30-day survival curves comparing patients with ICH and SAH were not significantly different. Patients with a Glasgow Coma Score less than 9 on arrival had a significantly worse outcome. CONCLUSION: Clinical features differed between ICH and SAH associated with ADPKD. Nevertheless, blood pressure control and early recognition of hemorrhagic stroke are important in ADPKD patients.

Long-term online hemodiafiltration does not reduce the frequency and severity of acquired cystic kidney disease in hemodialysis patients.

BACKGROUND: Acquired cystic kidney disease (ACKD) is a frequent complication in chronic hemodialysis (HD) patients and a risk factor for renal cell carcinoma. Online hemodiafiltration (HDF) provides better clearance of middle molecular weight solutes, but its effect on ACKD has not been investigated. MATERIALS AND METHODS: This case-control study enrolled 86 patients (43 HDF patients and 43 HD patients) who were matched according to age, sex, and duration of renal replacement therapy. The mean duration of HDF was 63 (+/- 35) months. The frequency and severity of ACKD was evaluated by ultrasonography using a severity scoring system. RESULTS: We observed ACKD in 23 of the HD patients (53.5%) and 21 of the HDF patients (48.8%). This difference was not statistically significant (p = 0.829). The overall ACKD severity scores were similar in the two groups (p = 0.875). Patients on HDF had significantly lower serum levels of alkaline phosphatase and intact parathyroid hormone. Multiple logistic regression analysis indicated that duration of renal replacement therapy was the only risk factor for the presence of ACKD (p < 0.001). There was a significant correlation between duration of renal replacement therapy and ACKD severity score (r = 0.589, p < 0.001). CONCLUSIONS: Our results suggest that long-term online HDF does not reduce the frequency and severity of ACKD in dialysis patients. Duration of renal replacement therapy is the most important risk factor for ACKD. Factors that cannot be corrected by use of HDF may contribute to the formation of renal cysts.

Concurrent renal cell carcinoma and central nervous system lymphoma in a patient with autosomal dominant polycystic kidney disease.

OBJECTIVE: To report an unusual case of synchronous renal cell carcinoma and CNS lymphoma in a patient with autosomal dominant polycystic kidney disease (ADPKD). CASE PRESENTATION AND INTERVENTION: A 58-year-old woman presented with progressive right hemiparesis of 2 months' duration. A brain CT scan revealed multiple enhanced lesions in the basal ganglia and the right occipital lobe. CNS lymphoma was confirmed by a stereotactic biopsy. Polycystic kidneys and a right renal mass were found incidentally. It was decided to treat the patient with cranial radiotherapy and chemotherapy first. The patient achieved complete remission of CNS lymphoma after 3 months, but the renal mass remained unchanged. A needle biopsy of the renal mass revealed renal cell carcinoma and unilateral nephrectomy was performed successfully. The patient remained in complete remission at 6-year follow-up. CONCLUSION: The patient was treated successfully with a combination of chemotherapy, radiotherapy, and unilateral nephrectomy. This report highlights the need for clinicians to remain alert to the possibility of double malignancies while caring for ADPKD patients, especially when multiple unexplained manifestations exist.

Symmetric synaptic patterns between starburst amacrine cells and direction selective ganglion cells in the rabbit retina.

Inputs from starburst amacrine cells (SACs) to ON-OFF direction selective ganglion cells (DSGCs) in the rabbit retina are themselves directional. However, the synaptic asymmetry between SACs and DSGCs required for generating direction selectivity has been controversial. We investigated dendritic contacts and distribution of inhibitory synapses between SACs and their overlapped DSGCs. Double injection of SAC/DSGC pairs and quantitative analysis revealed no obvious asymmetry of dendritic contacts between SACs and DSGCs. Furthermore, examination of the inhibitory input pattern on the dendrites of DSGCs using antibodies against GABA(A) receptors also suggested an isotropic arrangement with the overlapping SACs in both the preferred and the null directions. Therefore, the presynaptic mechanism of direction selectivity upon DSGCs may not result from a simple asymmetric arrangement with overlapping SACs. Multiple layer interactions and sophisticated synaptic connections between SACs and DSGCs are necessary.