Implementation of a Hybrid Intelligence System Enabling the Effectiveness Assessment of Interaction Channels Use in HMI.

The article presents a novel idea of Interaction Quality Sensor (IQS), introduced in the complete solution of Hybrid INTelligence (HINT) architecture for intelligent control systems. The proposed system is designed to use and prioritize multiple information channels (speech, images, videos) in order to optimize the information flow efficiency of interaction in HMI systems. The proposed architecture is implemented and validated in a real-world application of training unskilled workers-new employees (with lower competencies and/or a language barrier). With the help of the HINT system, the man-machine communication information channels are deliberately chosen based on IQS readouts to enable an untrained, inexperienced, foreign employee candidate to become a good worker, while not requiring the presence of either an interpreter or an expert during training. The proposed implementation is in line with the labor market trend, which displays significant fluctuations. The HINT system is designed to activate human resources and support organizations/enterprises in the effective assimilation of employees to the tasks performed on the production assembly line. The market need of solving this noticeable problem was caused by a large migration of employees within (and between) enterprises. The research results presented in the work show significant benefits of the methods used, while supporting multilingualism and optimizing the preselection of information channels.

Digital Stereotypes in HMI-The Influence of Feature Quantity Distribution in Deep Learning Models Training.

This paper proposes a concept of Digital Stereotypes, observed during research on quantitative overrepresentation of one class over others, and its impact on the results of the training of Deep Learning models. The real-life observed data classes are rarely of the same size, and the intuition of presenting multiple examples of one class and then showing a few counterexamples may be very misleading in multimodal classification. Deep Learning models, when taught with overrepresentation, may produce incorrect inferring results, similar to stereotypes. The generic idea of stereotypes seems to be helpful for categorisation from the training point of view, but it has a negative influence on the inferring result. Authors evaluate a large dataset in various scenarios: overrepresentation of one or two classes, underrepresentation of some classes, and same-size (trimmed) classes. The presented research can be applied to any multiclassification applications, but it may be especially important in AI, where the classification, uncertainty and building new knowledge overlap. This paper presents specific 'decreases in accuracy' observed within multiclassification of unleveled datasets. The 'decreases in accuracy', named by the authors 'stereotypes', can also bring an inspiring insight into other fields and applications, not only multimodal sentiment analysis.

Elevated preoperative levels of CA 19-9 and CA 125 predicts overall survival time in the pancreatic adenocarcinoma. Single institution series.

<b>Introduction: </b>Pancreatic cancer is a devastating disease, being the seventh cause of cancer-related deaths worldwide. Its aggressiveness is due to its specific biology and the late diagnosis of cancer. Therefore, the prognosis for patients suffering from this cancer is dismal, with 5-year overall survival rate of around 6-10%. Up to date, only a complete surgical resection of the cancerous entity warrants a significant improvement in patients' survival. Nevertheless, the pancreatic cancer's biology is still not fully elucidated, so that the accuracy of prognosis for certain patients is highly uncertain. Consequently, the importance of both clinical and basic research aiming to reveal the crucial molecular factors affecting long-term prognosis should be highlighted. There is a growing number of evidence that biomarkers of PC not only reflect the presence of tumor itself but also present a "hint" regarding its physiology. Thus the aim of this study was to assess the levels of commonly measured biomarkers and their influence on patients' overall survival. <br><b>Materials and methods: </b>The retrospective analysis of data on 129 patients admitted to our Department due to the diagnosis of pancreatic cancer was carried out. On the day of admission all the patients had their levels of CA_19-9, CA₁₂₅, CEA and CA_15-3 measured. The overall survival (OS) was defined as time elapsing from the day of admission to the day of death. The Kaplan- Meier curves were built for all potential factors, Cox regression model was applied to carry out a multivariate analysis. <br><b>Results: </b>We retrospectively analyzed 129 patients with a mean age of 62 years. As many as 95 of them had an unresectable lesion and 34 underwent curative operation. In total, the analyzed patient group was characterized by a median survival of 7 months and 12 days. Cumulative 1-year, 2-year and 4-year survival rates were 35%, 16% and 15%, respectively. In univariate analysis, factors such as age >= 60, inoperable lesion, CA_19-9 >= 200, CA₁₂₅ >= 20 and Neutrophile to Lymphocyte Ratio (NLR) >= 5 were associated with a lower median OS. In multivariate analysis, three factors, CA_19-9 >= 200, CA₁₂₅ >= 20 and age >= 60, were found to be statistically significant. Indeed, patients possessing all of them noted much poorer outcomes regarding OS factors: 89 days versus 235 days for the other patients (log rank test P = 0.02). <br><b>Conclusions: </b>Our study fortifies the evidence that preoperative levels of CA_19-9 and CA₁₂₅ have a direct influence on the longterm OS. Interestingly, in our patient group, the correlation of biomarkers with OS was higher than that of resectability. However, our study has some limitations regarding, for instance, the lack of data on chemotherapy, comorbidities etc. In the view of recent molecular studies on mucin involvement in PC development, it provides a strong clinical evidence to prove their importance.

Neutrophil-lymphocyte ratio and creatinine reduction ratio predict good early graft function among adult cadaveric donor renal transplant recipients. Single institution series.

Background Delayed graft function (DGF) is a common complication following kidney transplantation and is associated with ischemia-reperfusion injury (IRI). Lymphocytes contribute to the pathogenesis of IRI and ischemia-reperfusion related delayed graft function Materials and Methods 135 Caucasian patients received a kidney graft from deceased heart-beating organ donors. We divided patients into 2 groups- patients with the eGFR>=30 on the 21st day post-transplantation (n=36) and patients with the eGFR<30 on the 21st day post-transplantation (n=99) to assess kidney graft function. We measured the serum creatinine levels on 1st and 2nd post-transplant day and preoperative levels of monocytes, lymphocytes, platelets and neutrophils and their ratios. Results We have found statistically significant differences between the eGFR<30 and the eGFR>=30 groups in the average lnLymphocytes (0,36 +/-0,6 vs -0,016 +/-0,74 respectively p=0,004) lnNLR ( 1,27 +/-0,92 vs. 1,73+/-1,08 p=0,016) lnLMR (1,01 +/-0,57 vs. 0,73 +/-0,64 p=0,02), lnPLR (4,97 +/-0,55 vs. 5,26 +/- 0,67 p=0,023) and CCR2% (-20,20 +/- 21,55 vs. -4,29 +/- 29,62 p=0,004 . On univariate analysis, factors of lnLymphocytes >=0,22 (OR=0,331 95%CI 0,151-0,728 p=0,006), lnLMR>=1,4 (OR=0,255 95%CI 0,072-0,903 p=0,034) were associated with worse graft function while lnNLR>=1,05 (OR=2,653 95%CI 1,158-6,078 p=0,021), lnPLR>=5,15 (OR=2,536 95%CI 1,155-5,566 p=0,02) and CRR2 (OR=3,286 95% CI 1,359-7,944 p=0,008) indicated better graft function Conclusion Higher absolute lymphocyte count (lnLymphocytes) and lnLMR as well as lower lnNLR and lnPLR were associated with lower eGFR on the 21st day after kidney transplantation. On multivariate analysis CRR2 in combination with either lnLymphocytes, lnNLR or lnPLR improved the accuracy of detecting patients with poor graft function.

A Panel of CA19-9, Ca125, and Ca15-3 as the Enhanced Test for the Differential Diagnosis of the Pancreatic Lesion.

Background. Proper diagnosis of pancreatic lesion etiology is a challenging clinical dilemma. Studies suggest that surgery for suspected pancreatic ductal adenocarcinoma (PDAC) reveals a benign lesion in 5% to 13% of cases. The aim of our study was to assess whether routinely used biomarkers such as CA19-9, Ca125, Ca15-3, and CEA, when combined, can potentially yield an accurate test predicting pancreatic lesion etiology. Methods. We retrospectively analyzed data of 326 patients who underwent a diagnostic process due to pancreatic lesions of unknown etiology. Results. We found statistically significant differences in mean levels of all biomarkers. In logistic regression model, we applied levels CA19-9, Ca125, and Ca15-3 as variables. Two validation methods were used, namely, random data split into training and validation groups and bootstrapping. Afterward, we built ROC curve using the model that we had created, reaching AUC = 0,801. With an optimal cut-off point, it achieved specificity of 81,2% and sensitivity of 63,10%. Our proposed model has superior diagnostic accuracy to both CA19-9 (p = 0,0194) and CA125 (p = 0,0026). Conclusion. We propose a test that is superior to CA19-9 in a differential diagnosis of pancreatic lesion etiology. Although our test fails to reach exceptionally high accuracy, its feasibility and cost-effectiveness make it clinically useful.