Hybrid Deep Learning (hDL)-Based Brain-Computer Interface (BCI) Systems: A Systematic Review.

BACKGROUND: Brain-Computer Interface (BCI) is becoming more reliable, thanks to the advantages of Artificial Intelligence (AI). Recently, hybrid Deep Learning (hDL), which combines different DL algorithms, has gained momentum over the past five years. In this work, we proposed a review on hDL-based BCI starting from the seminal studies in 2015. OBJECTIVES: We have reviewed 47 papers that apply hDL to the BCI system published between 2015 and 2020 extracting trends and highlighting relevant aspects to the topic. METHODS: We have queried four scientific search engines (Google Scholar, PubMed, IEEE Xplore and Elsevier Science Direct) and different data items were extracted from each paper such as the database used, kind of application, online/offline training, tasks used for the BCI, pre-processing methodology adopted, type of normalization used, which kind of features were extracted, type of DL architecture used, number of layers implemented and which optimization approach were used as well. All these items were then investigated one by one to uncover trends. RESULTS: Our investigation reveals that Electroencephalography (EEG) has been the most used technique. Interestingly, despite the lower Signal-to-Noise Ratio (SNR) of the EEG data that makes pre-processing of that data mandatory, we have found that the pre-processing has only been used in 21.28% of the cases by showing that hDL seems to be able to overcome this intrinsic drawback of the EEG data. Temporal-features seem to be the most effective with 93.94% accuracy, while spatial-temporal features are the most used with 33.33% of the cases investigated. The most used architecture has been Convolutional Neural Network-Recurrent Neural Network CNN-RNN with 47% of the cases. Moreover, half of the studies have used a low number of layers to achieve a good compromise between the complexity of the network and computational efficiency. SIGNIFICANCE: To give useful information to the scientific community, we make our summary table of hDL-based BCI papers available and invite the community to published work to contribute to it directly. We have indicated a list of open challenges, emphasizing the need to use neuroimaging techniques other than EEG, such as functional Near-Infrared Spectroscopy (fNIRS), deeper investigate the advantages and disadvantages of using pre-processing and the relationship with the accuracy obtained. To implement new combinations of architectures, such as RNN-based and Deep Belief Network DBN-based, it is necessary to better explore the frequency and temporal-frequency features of the data at hand.

The significance of the furosemide test for predicting ascites control by diuretics in cirrhotics: a comparison with volume expansion and octreotide infusion.

To verify prospectively the usefulness of the furosemide-induced natriuresis test in predicting ascites control by medical treatment, 15 stable cirrhotics (9 male) with ascites were studied. Sodium excretion was measured after this test and after volume expansion with saline associated with intravenous infusion of octreotide; 6 months later, response to medical treatment was rated as good (N=9) or poor (N=6). Patients with poor ascites control had lower sodium excretion with the furosemide-induced natriuresis test (median, 88 vs 201 mmol; P < 0.01). Poor control was observed in four of four patients with sodium excretion < or =125 mmol, and good control in six of six patients with sodium excretion >175 mmol (P < 0.002). Volume expansion was followed by limited natriuresis (median, 20 mmol), in inverse relationship with plasma active renin concentration (P < 0.001). In conclusion, long-term ascites control is well predicted by the furosemide-induced natriuresis test.

Weight gain after liver transplantation and the insertion/deletion polymorphism of the angiotensin-converting enzyme gene.

BACKGROUND: Subjects who carry the D allele of the angiotensin-converting enzyme (ACE) gene have higher plasma and tissue angiotensin II levels, possibly concurrent with the development of obesity. In transplant recipients, treatment with calcineurin antagonists would magnify these effects. The present study verifies whether the allelic variants of ACE are a factor involved in excess weight gain after liver transplantation. METHODS: A consecutive series of 108 liver transplant recipients (73 males) were studied. Recipient ACE genotypes, determined by a polymerase chain reaction-based method, were related to body mass changes 1 year after transplant. RESULTS: Body mass index (BMI) increased from the pretransplant value of 25.1+/-3.3 kg/m2 to 25.9+/-3.5 kg/m2 (P<0.005). The difference was mainly attributable to recipients carrying 1 D allele or more (N=88) in whom the BMI increased from 25.3+/-3.1 kg/m2 to 26.3+/-3.3 kg/m2 (P<0.005). A BMI of 25 kg/m or greater was measured in 30 of 45 deletion/deletion homozygotes and 25 of 43 insertion/deletion heterozygotes; in contrast, 14 of 20 insertion/insertion homozygotes had a normal body mass (P<0.01). Among patients with normal body mass pretransplant (N=56), none of 13 insertion/insertion homozygotes reached a BMI value 25 kg/m or greater posttransplant (P<0.005). At multivariate analysis, pretransplant body mass and carriage of 1 D allele or more were independent predictors of body mass gain greater than 2 kg/m. CONCLUSIONS: Carriage of the D allele of the ACE gene is a strong, independent risk factor for excess weight gain after liver transplantation.

Pegylated versus standard interferon-alpha in antiviral regimens for post-transplant recurrent hepatitis C: Comparison of tolerability and efficacy.

BACKGROUND: In the treatment of hepatitis C virus (HCV) infection, regimens including pegylated interferon-alpha are superior to those including standard interferon; the present retrospective study was performed to verify whether the same is applicable to biopsy-proven recurrent hepatitis C (genotype 1b) after liver transplantation (OLT). METHODS: Twenty-four patients (16 male) were studied. Twelve had received interferon-alpha(2b) (IFN), 9 MU weekly and 12 received pegylated interferon-alpha(2b) (PEG-IFN), 0.5 microg/kg weekly. All had received oral ribavirin 600-800 mg/day. Treatment duration was intended for 12 months. A repeat liver biopsy, with evaluation of the Ishak grading and staging scores, was obtained at 1 year. RESULTS: Only 12/24 patients (50%) completed a full year of therapy; 17 (71%) experienced side-effects requiring a 50% dosage reduction or discontinuation of the IFN, PEG-IFN and/or ribavirin. This was observed in 6/12 patients (50%) treated with IFN in comparison to 11/12 patients (92%) treated with PEG-IFN (P < 0.05). The difference was mainly accounted for by anemia and leukopenia that were reported in 4/12 IFN patients (33%) versus 9/12 PEG-IFN patients (75%; P < 0.05), respectively. End-of-treatment viral response (ETVR) and histological response were always associated and occurred in 4/24 patients (17%), two in each treatment arm. Patients with ETVR were younger, had always completed 1 year of therapy, had had recurrent hepatitis later after transplantation and presented a higher baseline grading score. CONCLUSIONS: In the OLT setting, the potential benefits of antiviral treatments including PEG-IFN may be limited by the poor tolerability of the adopted drugs.

Carriage of the apolipoprotein E-epsilon4 allele and histologic outcome of recurrent hepatitis C after antiviral treatment.

Carriage of the epsilon4 allelic variant of the apolipoprotein E (ApoE) gene might affect the outcome of hepatitis C virus (HCV) infection. The liver transplantation setting offers the opportunity to verify the role of the donor's vs recipient's ApoE polymorphism. Twenty-four patients (16 men) with recurrent hepatitis C, all infected by HCV-1b and treated with interferon and ribavirin, were genotyped for ApoE variants. Liver biopsies were done at baseline and 12 months later After treatment, staging scores improved in 10 of 24 patients. Staging improvement was associated with recipient sex, completion of the full antiviral schedule, and recipient's epsilon4 carriage. The beneficial effect of epsilon4 carriage toward the progression of fibrosis was due entirely to the contribution given by male patients and was independent of the viral response. Recipients', but not donors', carriage of at least 1 epsilon4 allele might be associated with a better histologic outcome in recurrent HCV infection.

Favourable outcome of adefovir-dipivoxil treatment in acute de novo hepatitis B after liver transplantation.

Adefovir-dipivoxil has been shown to be effective against lamivudine-resistant mutants in immunocompetent patients and in a small number of liver transplant recipients with recurrent hepatitis B virus (HBV) infection. The therapeutic role of adefovir-dipivoxil in acute de novo HBV infection after transplantation is uncertain. We describe a case of acute de novo HBV infection that occurred after liver transplantation and that was treated with lamivudine followed (when viral escape mutants emerged) by adefovir-dipivoxil rescue. Treatment outcome was excellent, with complete viral clearance and development of a protective titer of antibodies to anti-hepatitis B surface antigen. Because the donor was vaccinated against HBV, it is conceivable that clearance of HBV infection in the recipient might have been favored by adoptive transfer of immunity to HBV. The immune status of the donor might be a factor to consider when determining the treatment options for de novo hepatitis B.