This paper contributes with a pragmatic evaluation framework for explainable Machine Learning (ML) models for clinical decision support. The study revealed a more nuanced role for ML explanation models, when these are pragmatically embedded in the clinical context. Despite the general positive attitude of healthcare professionals (HCPs) towards explanations as a safety and trust mechanism, for a significant set of participants there were negative effects associated with confirmation bias, accentuating model over-reliance and increased effort to interact with the model. Also, contradicting one of its main intended functions, standard explanatory models showed limited ability to support a critical understanding of the limitations of the model. However, we found new significant positive effects which repositions the role of explanations within a clinical context: these include reduction of automation bias, addressing ambiguous clinical cases (cases where HCPs were not certain about their decision) and support of less experienced HCPs in the acquisition of new domain knowledge.

Specialized transformers-based models (such as BioBERT and BioMegatron) are adapted for the biomedical domain based on publicly available biomedical corpora. As such, they have the potential to encode large-scale biological knowledge. We investigate the encoding and representation of biological knowledge in these models, and its potential utility to support inference in cancer precision medicine—namely, the interpretation of the clinical significance of genomic alterations. We compare the performance of different transformer baselines; we use probing to determine the consistency of encodings for distinct entities; and we use clustering methods to compare and contrast the internal properties of the embeddings for genes, variants, drugs, and diseases. We show that these models do indeed encode biological knowledge, although some of this is lost in fine-tuning for specific tasks. Finally, we analyze how the models behave with regard to biases and imbalances in the dataset.

PURPOSE: Patients with cancer are at increased risk of severe COVID-19 disease, but have heterogeneous presentations and outcomes. Decision-making tools for hospital admission, severity prediction, and increased monitoring for early intervention are critical. We sought to identify features of COVID-19 disease in patients with cancer predicting severe disease and build a decision support online tool, COVID-19 Risk in Oncology Evaluation Tool (CORONET). METHODS: Patients with active cancer (stage I-IV) and laboratory-confirmed COVID-19 disease presenting to hospitals worldwide were included. Discharge (within 24 hours), admission (≥ 24 hours inpatient), oxygen (O₂) requirement, and death were combined in a 0-3 point severity scale. Association of features with outcomes were investigated using Lasso regression and Random Forest combined with Shapley Additive Explanations. The CORONET model was then examined in the entire cohort to build an online CORONET decision support tool. Admission and severe disease thresholds were established through pragmatically defined cost functions. Finally, the CORONET model was validated on an external cohort. RESULTS: The model development data set comprised 920 patients, with median age 70 (range 5-99) years, 56% males, 44% females, and 81% solid versus 19% hematologic cancers. In derivation, Random Forest demonstrated superior performance over Lasso with lower mean squared error (0.801 v 0.807) and was selected for development. During validation (n = 282 patients), the performance of CORONET varied depending on the country cohort. CORONET cutoffs for admission and mortality of 1.0 and 2.3 were established. The CORONET decision support tool recommended admission for 95% of patients eventually requiring oxygen and 97% of those who died (94% and 98% in validation, respectively). The specificity for mortality prediction was 92% and 83% in derivation and validation, respectively. Shapley Additive Explanations revealed that National Early Warning Score 2, C-reactive protein, and albumin were the most important features contributing to COVID-19 severity prediction in patients with cancer at time of hospital presentation. CONCLUSION: CORONET, a decision support tool validated in health care systems worldwide, can aid admission decisions and predict COVID-19 severity in patients with cancer.

Advances in artificial intelligence offer the promise of automated analysis of optical coherence tomography (OCT) scans to detect ocular complications from anticancer drug therapy. To explore how such AI outputs are interpreted in clinical settings, we conducted a survey-based interview study with 27 clinicians -comprising 10 ophthalmic specialists, 10 ophthalmic practitioners, and 7 oncologists. Participants were first introduced to core AI concepts and realistic clinical scenarios, then asked to assess AI-generated OCT analyses using standardized Likert-scale questions, allowing us to gauge their understanding, trust, and readiness to integrate AI into practice. We developed a questionnaire through literature review and consultations with ophthalmologists, computer scientists, and AI researchers. A single investigator interviewed 27 clinicians across three specialties and transcribed their responses. Data were summarized as medians (ranges) and compared with Mann-Whitney U tests (α = 0.05). We noted important differences in the impact of various explainability methods on trust, depending on the clinical or AI scenario nature and the staff expertise. Explanations of AI outputs increased trust in the AI algorithm when outputs simply reflected ground truth expert opinion. When clinical scenarios were complex with incorrect AI outcomes, a mixed response to explainability led to correctly reduced trust in experienced clinicians but mixed feedback amongst less experienced clinicians. All clinicians had a general consensus on lack of current knowledge in interacting with AI and desire more training. Clinicians' trust in AI algorithms are affected by explainability methods and factors, including AI's performance, personal judgments and clinical experience. The development of clinical AI systems should consider the above and these responses ideally be factored into real-world assessments. Use of this study's findings could help improve the real world validity of medical AI systems by enhancing the human-computer interactions, with preferred explainability techniques tailored to specific situations.

In this paper a method of the efficiency evaluation of the duty cycle of Refuse Collection Vehicle is presented. Using real world data, two representative duty cycles were analysed. Total cycle efficiency was calculated, as well as the efficiency of particular cycle phases. Then, energy needed to collect and compact the waste and energy from fuel were compared. Measured and calculated values were shown on the diagrams illustrating hydraulic system parameters and the real load applied to the chassis combustion engine. Then, basic statistical analysis of 524 duty cycles was performed, which shown dependencies between measured parameters and efficiency. Results of the analysis show that presented method can be used to quantitatively assess performance of RCV's regarding efficiency of hydraulic power delivery