Inline classification of polymer films using Machine learning methods.

Improving the sortability of plastic packaging film waste (PPFW) is crucial for increasing the recycling rate in Austria as they account for 150,000 t of the annually produced 300,000 t of plastic packaging waste. Currently PPFW is thermally recovered, as it is impossible to separate the mechanically recyclable monomaterial films from the non mechanically-recyclable multimaterial films. In this study, machine learning models capable of classifying inline into monolayer and multilayer films of PPFW according to their spectral fingerprint taken in transflection were created. Feature selection methods, like PCA and MRMR F-Tests, identified the most relevant spectral ranges for classification, that show the least redundancy and highest relevance. This effective subset of features decreases the required complexity of the model while reducing prediction time without compromising accuracy. The resulting models achieved a prediction accuracy of 85 % on unseen specimens with minimal prediction latency, effectively showing the inline applicability of these models in sorting aggregates.

Alloreactivity: the Janus-face of hematopoietic stem cell transplantation.

Differences in major and minor histocompatibility antigens between donor and recipient trigger powerful graft-versus-host reactions after allogeneic hematopoietic stem cell transplantation (HSCT). The clinical effects of alloreactivity present a Janus-face: detrimental graft-versus-host disease increases non-relapse mortality, beneficial graft-versus-malignancy may cure the recipient. The ultimate consequences on long-term outcome remain a matter of debate. We hypothesized that increasing donor-recipient antigen matching would decrease the negative effects, while preserving antitumor alloreactivity. We analyzed retrospectively a predefined cohort of 32 838 such patients and compared it to 59 692 patients with autologous HSCT as reference group. We found a significant and systematic decrease in non-relapse mortality with decreasing phenotypic and genotypic antigen disparity, paralleled by a stepwise increase in overall and relapse-free survival (Spearman correlation coefficients of cumulative excess event rates at 5 years 0.964; P<0.00; respectively 0.976; P<0.00). We observed this systematic stepwise effect in all main disease and disease-stage categories. The results suggest that detrimental effects of alloreactivity are additive with each step of mismatching; the beneficial effects remain preserved. Hence, if there is a choice, the best match should be donor of choice. The data support an intensified search for predictive genomic and environmental factors of 'no-graft-versus-host disease'.