Traditional approaches for finding well-performing parameterizations of complex imaging systems, such as super-resolution microscopes rely on an extensive exploration phase over the illumination and acquisition settings, prior to the imaging task. This strategy suffers from several issues: it requires a large amount of parameter configurations to be evaluated, it leads to discrepancies between well-performing parameters in the exploration phase and imaging task, and it results in a waste of time and resources given that optimization and final imaging tasks are conducted separately. Here we show that a fully automated, machine learning-based system can conduct imaging parameter optimization toward a trade-off between several objectives, simultaneously to the imaging task. Its potential is highlighted on various imaging tasks, such as live-cell and multicolor imaging and multimodal optimization. This online optimization routine can be integrated to various imaging systems to increase accessibility, optimize performance and improve overall imaging quality.

Background Since screening programs identify only a small proportion of the population as eligible for an intervention, genomic prediction of heritable risk factors could decrease the number needing to be screened by removing individuals at low genetic risk. We therefore tested whether a

In all, 80% of antenatal karyotypes are generated by Down's syndrome screening programmes (DSSP). After a positive screening, women are offered prenatal foetus karyotyping, the gold standard. Reliable molecular methods for rapid aneuploidy diagnosis (RAD: fluorescence in situ hybridization (FISH) and quantitative fluorescence PCR (QF-PCR)) can detect common aneuploidies, and are faster and less expensive than karyotyping.In the UK, RAD is recommended as a standalone approach in DSSP, whereas the US guidelines recommend that RAD be followed up by karyotyping. A cost-effectiveness (CE) analysis of RAD in various DSSP is lacking. There is a debate over the significance of chromosome abnormalities (CA) detected with karyotyping but not using RAD. Our objectives were to compare the CE of RAD versus karyotyping, to evaluate the clinically significant missed CA and to determine the impact of detecting the missed CA. We performed computer simulations to compare six screening options followed by FISH, PCR or karyotyping using a population of 110 948 pregnancies. Among the safer screening strategies, the most cost-effective strategy was contingent screening with QF-PCR (CE ratio of $24 084 per Down's syndrome (DS) detected). Using karyotyping, the CE ratio increased to $27 898. QF-PCR missed only six clinically significant CA of which only one was expected to confer a high risk of an abnormal outcome. The incremental CE ratio (ICER) to find the CA missed by RAD was $66 608 per CA. These costs are much higher than those involved for detecting DS cases. As the DSSP are mainly designed for DS detection, it may be relevant to question the additional costs of karyotyping.

Generative Adversarial Networks (GANs) can successfully approximate a probability distribution and produce realistic samples. However, open questions such as sufficient convergence conditions and mode collapse still persist. In this paper, we build on existing work in the area by proposing a novel framework for training the generator against an ensemble of discriminator networks, which can be seen as a one-student/multiple-teachers setting. We formalize this problem within the full-information adversarial bandit framework, where we evaluate the capability of an algorithm to select mixtures of discriminators for providing the generator with feedback during learning. To this end, we propose a reward function which reflects the progress made by the generator and dynamically update the mixture weights allocated to each discriminator. We also draw connections between our algorithm and stochastic optimization methods and then show that existing approaches using multiple discriminators in literature can be recovered from our framework. We argue that less expressive discriminators are smoother and have a general coarse grained view of the modes map, which enforces the generator to cover a wide portion of the data distribution support. On the other hand, highly expressive discriminators ensure samples quality. Finally, experimental results show that our approach improves samples quality and diversity over existing baselines by effectively learning a curriculum. These results also support the claim that weaker discriminators have higher entropy improving modes coverage.

A patient-level Markov decision model was used to simulate a virtual cohort of 500,000 women 40 years old and over, in relation to osteoporosis-related hip, clinical vertebral, and wrist bone fractures events. Sixteen different screening options of three main scenario groups were compared: (1) the status quo (no specific national prevention program); (2) a universal primary prevention program; and (3) a universal screening and treatment program based on the 10-year absolute risk of fracture. The outcomes measured were total directs costs from the perspective of the public health care system, number of fractures, and quality-adjusted life-years (QALYs). Results show that an option consisting of a program promoting physical activity and treatment if a fracture occurs is the most cost-effective (CE) (cost/fracture averted) alternative and also the only cost saving one, especially for women 40 to 64 years old. In women who are 65 years and over, bone mineral density (BMD)-based screening and treatment based on the 10-year absolute fracture risk calculated using a Canadian Association of Radiologists and Osteoporosis Canada (CAROC) tool is the best next alternative. In terms of cost-utility (CU), results were similar. For women less than 65 years old, a program promoting physical activity emerged as cost-saving but BMD-based screening with pharmacological treatment also emerged as an interesting alternative. In conclusion, a program promoting physical activity is the most CE and CU option for women 40 to 64 years old. BMD screening and pharmacological treatment might be considered a reasonable alternative for women 65 years old and over because at a healthcare capacity of $50,000 Canadian dollars ($CAD) for each additional fracture averted or for one QALY gained its probabilities of cost-effectiveness compared to the program promoting physical activity are 63% and 75%, respectively, which could be considered socially acceptable. Consideration of the indirect costs could change these findings.

No abstract

Objectives The study sought to assess the clinical performance of a machine learning model aiming to identify unusual medication orders. Materials and Methods This prospective study was conducted at CHU Sainte-Justine, Canada, from April to August 2020. An unsupervised machine learning model based on GANomaly and 2 baselines were trained to learn medication order patterns from 10 years of data. Clinical pharmacists dichotomously (typical or atypical) labeled orders and pharmacological profiles (patients’ medication lists). Confusion matrices, areas under the precision-recall curve (AUPRs), and F1 scores were calculated. Results A total of 12 471 medication orders and 1356 profiles were labeled by 25 pharmacists. Medication order predictions showed a precision of 35%, recall (sensitivity) of 26%, and specificity of 97% as compared with pharmacist labels, with an AUPR of 0.25 and an F1 score of 0.30. Profile predictions showed a precision of 49%, recall of 75%, and specificity of 82%, with an AUPR of 0.60, and an F1 score of 0.59. The model performed better than the baselines. According to the pharmacists, the model was a useful screening tool, and 9 of 15 participants preferred predictions by medication, rather than by profile. Discussion Predictions for profiles had higher F1 scores and recall compared with medication order predictions. Although the performance was much better for profile predictions, pharmacists generally preferred medication order predictions. Conclusions Based on the AUPR, this model showed better performance for the identification of atypical pharmacological profiles than for medication orders. Pharmacists considered the model a useful screening tool. Improving these predictions should be prioritized in future research to maximize clinical impact.

No abstract