The digitization of organic synthesis.

Organic chemistry has largely been conducted in an ad hoc manner by academic laboratories that are funded by grants directed towards the investigation of specific goals or hypotheses. Although modern synthetic methods can provide access to molecules of considerable complexity, predicting the outcome of a single chemical reaction remains a major challenge. Improvements in the prediction of 'above-the-arrow' reaction conditions are needed to enable intelligent decision making to select an optimal synthetic sequence that is guided by metrics including efficiency, quality and yield. Methods for the communication and the sharing of data will need to evolve from traditional tools to machine-readable formats and open collaborative frameworks. This will accelerate innovation and require the creation of a chemistry commons with standardized data handling, curation and metrics.

SeedQuant: a deep learning-based tool for assessing stimulant and inhibitor activity on root parasitic seeds.

Witchweeds (Striga spp.) and broomrapes (Orobanchaceae and Phelipanche spp.) are root parasitic plants that infest many crops in warm and temperate zones, causing enormous yield losses and endangering global food security. Seeds of these obligate parasites require rhizospheric, host-released stimulants to germinate, which opens up possibilities for controlling them by applying specific germination inhibitors or synthetic stimulants that induce lethal germination in the host's absence. To determine their effect on germination, root exudates or synthetic stimulants/inhibitors are usually applied to parasitic seeds in in vitro bioassays, followed by assessment of germination ratios. Although these protocols are very sensitive, the germination recording process is laborious, representing a challenge for researchers and impeding high-throughput screens. Here, we developed an automatic seed census tool to count and discriminate germinated seeds (GS) from non-GS. We combined deep learning, a powerful data-driven framework that can accelerate the procedure and increase its accuracy, for object detection with computer vision latest development based on the Faster Region-based Convolutional Neural Network algorithm. Our method showed an accuracy of 94% in counting seeds of Striga hermonthica and reduced the required time from approximately 5 min to 5 s per image. Our proposed software, SeedQuant, will be of great help for seed germination bioassays and enable high-throughput screening for germination stimulants/inhibitors. SeedQuant is an open-source software that can be further trained to count different types of seeds for research purposes.

Combination treatment optimization using a pan-cancer pathway model.

The design of efficient combination therapies is a difficult key challenge in the treatment of complex diseases such as cancers. The large heterogeneity of cancers and the large number of available drugs renders exhaustive in vivo or even in vitro investigation of possible treatments impractical. In recent years, sophisticated mechanistic, ordinary differential equation-based pathways models that can predict treatment responses at a molecular level have been developed. However, surprisingly little effort has been put into leveraging these models to find novel therapies. In this paper we use for the first time, to our knowledge, a large-scale state-of-the-art pan-cancer signaling pathway model to identify candidates for novel combination therapies to treat individual cancer cell lines from various tissues (e.g., minimizing proliferation while keeping dosage low to avoid adverse side effects) and populations of heterogeneous cancer cell lines (e.g., minimizing the maximum or average proliferation across the cell lines while keeping dosage low). We also show how our method can be used to optimize the drug combinations used in sequential treatment plans-that is, optimized sequences of potentially different drug combinations-providing additional benefits. In order to solve the treatment optimization problems, we combine the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) algorithm with a significantly more scalable sampling scheme for truncated Gaussian distributions, based on a Hamiltonian Monte-Carlo method. These optimization techniques are independent of the signaling pathway model, and can thus be adapted to find treatment candidates for other complex diseases than cancers as well, as long as a suitable predictive model is available.

Development and Pilot Testing of a Patient-Centered Web-Based Reproductive Decision Support Tool for Primary Care.

BACKGROUND: Patient-centered counseling to help women achieve their reproductive goals is an essential yet often absent component of primary care. OBJECTIVE: We developed and piloted MyPath, a novel web-based decision support tool integrating reproductive goals assessment, information about optimizing health before pregnancy, and contraceptive decision support, for use prior to primary care visits in the Veterans Administration (VA). DESIGN: We created MyPath using best practices for decision tool development, including a conceptual framework informed by theory and user-centered design with input from patients, providers, and scientific experts. We conducted a non-randomized pilot in two VA Women's Health primary care clinics. A control group (n = 28) was recruited prior to and intervention group (n = 30) recruited after introduction of MyPath into clinics. PARTICIPANTS: Women Veterans ages 18-44 with an upcoming visit scheduled with one of eight providers. INTERVENTIONS: After recruitment of controls, providers and staff received a brief introduction to MyPath. Patients scheduled to see providers in the intervention phase used MyPath on an iPad in the waiting room prior to their visit. MAIN MEASURES: Acceptability, feasibility, discussions about pregnancy and/or contraceptive needs, and contraceptive decision quality by a survey of participants and providers. KEY RESULTS: Nearly all participants who used MyPath reported they learned new information (97%) and would recommend it to other Veterans (93%). No providers reported that MyPath significantly increased workload. A greater proportion of intervention participants reported having discussions about reproductive needs in their visit compared to controls (93% vs 68%; p = 0.02). Intervention participants also experienced greater increases in pre-/post-visit knowledge and communication self-efficacy and a trend towards greater reduction in contraceptive decision conflict compared to controls. CONCLUSIONS: MyPath was highly acceptable to women, increased the proportion of primary care visits addressing reproductive needs, and improved decision quality without increasing providers' perceived workload. A larger randomized evaluation of effectiveness is warranted.

Challenges of developing artificial intelligence-assisted tools for clinical medicine.

Machine learning, a subset of artificial intelligence (AI), is a set of computational tools that can be used to enhance provision of clinical care in all areas of medicine. Gastroenterology and hepatology utilize multiple sources of information, including visual findings on endoscopy, radiologic imaging, manometric testing, genomes, proteomes, and metabolomes. However, clinical care is complex and requires a thoughtful approach to best deploy AI tools to improve quality of care and bring value to patients and providers. On the operational level, AI-assisted clinical management should consider logistic challenges in care delivery, data management, and algorithmic stewardship. There is still much work to be done on a broader societal level in developing ethical, regulatory, and reimbursement frameworks. A multidisciplinary approach and awareness of AI tools will create a vibrant ecosystem for using AI-assisted tools to guide and enhance clinical practice. From optically enhanced endoscopy to clinical decision support for risk stratification, AI tools will potentially transform our practice by leveraging massive amounts of data to personalize care to the right patient, in the right amount, at the right time.

Efficacy of an Online Curriculum for Perioperative Goals of Care and Code Status Discussions: A Randomized Controlled Trial.

BACKGROUND: Preoperative goals of care (GOC) and code status (CS) discussions are important in achieving an in-depth understanding of the patient's care goals in the setting of a serious illness, enabling the clinician to ensure patient autonomy and shared decision making. Past studies have shown that anesthesiologists are not formally trained in leading these discussions and may lack the necessary skill set. We created an innovative online video curriculum designed to teach these skills. This curriculum was compared to a traditional method of learning from reading the medical literature. METHODS: In this bi-institutional randomized controlled trial at 2 major academic medical centers, 60 anesthesiology trainees were randomized to receive the educational content in 1 of 2 formats: (1) the novel video curriculum (video group) or (2) journal articles (reading group). Thirty residents were assigned to the experimental video curriculum group, and 30 were assigned to the reading group. The content incorporated into the 2 formats focused on general preoperative evaluation of patients and communication strategies pertaining to GOC and CS discussions. Residents in both groups underwent a pre- and postintervention objective structured clinical examination (OSCE) with standardized patients. Both OSCEs were scored using the same 24-point rubric. Score changes between the 2 OSCEs were examined using linear regression, and interrater reliability was assessed using weighted Cohen's kappa. RESULTS: Residents receiving the video curriculum performed significantly better overall on the OSCE encounter, with a mean score of 4.19 compared to 3.79 in the reading group. The video curriculum group also demonstrated statistically significant increased scores on 8 of 24 rubric categories when compared to the reading group. CONCLUSIONS: Our novel video curriculum led to significant increases in resident performance during simulated GOC discussions and modest increases during CS discussions. Further development and refinement of this curriculum are warranted.

Clinical implementation of algorithm-based embryo selection is associated with improved pregnancy outcomes in single vitrified warmed euploid embryo transfers.

PURPOSE: To assess whether utilization of a mathematical ranking algorithm for assistance with embryo selection improves clinical outcomes compared with traditional embryo selection via morphologic grading in single vitrified warmed euploid embryo transfers (euploid SETs). METHODS: A retrospective cohort study in a single, academic center from September 2016 to February 2020 was performed. A total of 4320 euploid SETs met inclusion criteria and were included in the study. Controls included all euploid SETs in which embryo selection was performed by a senior embryologist based on modified Gardner grading (traditional approach). Cases included euploid SETs in which embryo selection was performed using an automated algorithm-based approach (algorithm-based approach). Our primary outcome was implantation rate. Secondary outcomes included ongoing pregnancy/live birth rate and clinical loss rate. RESULTS: The implantation rate and ongoing pregnancy/live birth rate were significantly higher when using the algorithm-based approach compared with the traditional approach (65.3% vs 57.8%, p<0.0001 and 54.7% vs 48.1%, p=0.0001, respectively). After adjusting for potential confounding variables, utilization of the algorithm remained significantly associated with improved odds of implantation (aOR 1.51, 95% CI 1.04, 2.18, p=0.03) ongoing pregnancy/live birth (aOR 1.99, 95% CI 1.38, 2.86, p=0.0002), and decreased odds of clinical loss (aOR 0.42, 95% CI 0.21, 0.84, p=0.01). CONCLUSIONS: Clinical implementation of an automated mathematical algorithm for embryo ranking and selection is significantly associated with improved implantation and ongoing pregnancy/live birth as compared with traditional embryo selection in euploid SETs.

Can methods of artificial intelligence aid in optimizing patient selection in patients undergoing intrauterine inseminations?

PURPOSE: AI and its machine learning algorithms have proven useful in several fields of medicine, including medically assisted reproduction. The purpose of the study was to construct several predictive models based on clinical data and select the best models to predict IUI procedure outcomes. METHODS: Clinical data (patient baseline characteristics, sperm quality, hormonal status, and cycle data) from 1029 IUI procedures performed in 413 couples stimulated by clomiphene citrate, letrozole, or gonadotropins were used to build several models to predict clinical pregnancy. The models included ANN, random forest, PLS, SVM, and linear models using the caret package in R. The models were evaluated using ROC analysis by means of random CV on test data. RESULTS: Out of the best performing models, the random forest model achieved an AUC of 0.66, a sensitivity of 0.432, and a specificity of 0.756. This performance was followed by the PLS model, which achieved a sensitivity of 0.459 and specificity of 0.734. The other models achieved significantly lower AUCs. When adjusting the predictive cutoff value, confusion matrices show that clinical pregnancy is twice as likely in the case of positive prediction. CONCLUSION: Among the compared methods, the random forest and PLS models demonstrated superior performance in predicting the clinical outcome of IUI. With additional research and clinical validation, AI methods may be successfully used in improving patient selection and consequently lead to better clinical results.

Embryo selection with artificial intelligence: how to evaluate and compare methods?

Embryo selection within in vitro fertilization (IVF) is the process of evaluating qualities of fertilized oocytes (embryos) and selecting the best embryo(s) available within a patient cohort for subsequent transfer or cryopreservation. In recent years, artificial intelligence (AI) has been used extensively to improve and automate the embryo ranking and selection procedure by extracting relevant information from embryo microscopy images. The AI models are evaluated based on their ability to identify the embryo(s) with the highest chance(s) of achieving a successful pregnancy. Whether such evaluations should be based on ranking performance or pregnancy prediction, however, seems to divide studies. As such, a variety of performance metrics are reported, and comparisons between studies are often made on different outcomes and data foundations. Moreover, superiority of AI methods over manual human evaluation is often claimed based on retrospective data, without any mentions of potential bias. In this paper, we provide a technical view on some of the major topics that divide how current AI models are trained, evaluated and compared. We explain and discuss the most common evaluation metrics and relate them to the two separate evaluation objectives, ranking and prediction. We also discuss when and how to compare AI models across studies and explain in detail how a selection bias is inevitable when comparing AI models against current embryo selection practice in retrospective cohort studies.

Nephrology in the Academic Intensive Care Unit: A Qualitative Study of Interdisciplinary Collaboration.

RATIONALE & OBJECTIVE: Collaboration between nephrology consultants and intensive care unit (ICU) teams is important in light of the high incidence of acute kidney injury in today's ICUs. Although there is considerable debate about how nephrology consultants and ICU teams should collaborate, communicative dynamics between the 2 parties remain poorly understood. This article describes interactions between nephrology consultants and ICU teams in the academic medical setting. STUDY DESIGN: Focused ethnography using semi-structured interviews and participant observation. SETTING & PARTICIPANTS: Purposive sampling was used to enroll nephrologists, nephrology fellows, and ICU practitioners across several roles collaborating in 3 ICUs (a medical ICU, a surgical ICU, and a cardiothoracic surgical ICU) of a large urban US academic medical center. Participant observation (150 hours) and semi-structured interviews (35) continued until theoretical saturation. ANALYTICAL APPROACH: Interview and fieldnote transcripts were coded in an iterative team-based process. Explanation was developed using an abductive approach. RESULTS: Nephrology consultants and surgical ICU teams exhibited discordant preferences about the aggressiveness of renal replacement therapy based on different understandings of physiology, goals of care, and acuity. Collaborative difficulties resulting from this discordance led to nephrology consultants often serving as dialysis proceduralists rather than diagnosticians in surgical ICUs and to consultants sometimes choosing not to express disagreements about clinical care because of the belief that doing so would not lead to changes in the course of care. LIMITATIONS: Aspects of this single-site study of an academic medical center may not be generalizable to other clinical settings and samples. Surgical team perspectives would provide further detail about nephrology consultation in surgical ICUs. The effects of findings on patient care were not examined. CONCLUSIONS: Differences in approach between internal medicine-trained nephrologists and anesthesia- and surgery-trained intensivists and surgeons led to collaborative difficulties in surgical ICUs. These findings stress the need for medical teamwork research and intervention to address issues stemming from disciplinary siloing rooted in long-term socialization to different disciplinary practices.

Machine learning-based prediction of transfusion.

BACKGROUND: The ability to predict transfusions arising during hospital admission might enable economized blood supply management and might furthermore increase patient safety by ensuring a sufficient stock of red blood cells (RBCs) for a specific patient. We therefore investigated the precision of four different machine learning-based prediction algorithms to predict transfusion, massive transfusion, and the number of transfusions in patients admitted to a hospital. STUDY DESIGN AND METHODS: This was a retrospective, observational study in three adult tertiary care hospitals in Western Australia between January 2008 and June 2017. Primary outcome measures for the classification tasks were the area under the curve for the receiver operating characteristics curve, the F1 score, and the average precision of the four machine learning algorithms used: neural networks (NNs), logistic regression (LR), random forests (RFs), and gradient boosting (GB) trees. RESULTS: Using our four predictive models, transfusion of at least 1 unit of RBCs could be predicted rather accurately (sensitivity for NN, LR, RF, and GB: 0.898, 0.894, 0.584, and 0.872, respectively; specificity: 0.958, 0.966, 0.964, 0.965). Using the four methods for prediction of massive transfusion was less successful (sensitivity for NN, LR, RF, and GB: 0.780, 0.721, 0.002, and 0.797, respectively; specificity: 0.994, 0.995, 0.993, 0.995). As a consequence, prediction of the total number of packed RBCs transfused was also rather inaccurate. CONCLUSION: This study demonstrates that the necessity for intrahospital transfusion can be forecasted reliably, however the amount of RBC units transfused during a hospital stay is more difficult to predict.

Computerized Medical Evidence-Based Decision Assistance System "MEBDAS®" improves in-hospital outcome after pancreatoduodenectomy for pancreatic cancer.

BACKGROUND: Indication for pancreatoduodenectomy for pancreatic cancer can be challenging. Wrong decisions in indication processes lead to significant health impairments. Computerized decision support systems can take over parts of decision-making processes, making them more accurate. MEBDAS® is a decision-supporting software that predicts outcomes of proposed treatments. AIM: to determine the decision concordance between MEBDAS® and multidisciplinary tumour board (MTB) and the impact of MEBDAS® on in-hospital outcome at different indication thresholds. METHODS: 126 patients with pancreatoduodenectomy from a high-volume university hospital were included. Outcome indicators were in-hospital mortality, Comprehensive Complication Index (CCI®), therapy-related loss of "Quality-Adjusted-Life-Day" (QALD-loss) and prognostic gain of treatment-related "Quality-Adjusted-Life-Year" (QALY-gain). RESULTS: The concordance of decisions was 94.4% at the indication threshold of 0. By raising the indication threshold to 1 year, the concordance decreased to 0%, the in-hospital-mortality dropped from 2.52% to 0%, the CCI® decreased from 26.47 to 13.90, the therapy-related QALD-loss declined from 21.53 to 16.22 days and the prognostic QALY-gain increased from 0.374 to 0.906 years. At IT = 0.250 years, the concordance was 61.11% and differences between MTB and MEBDAS®-group were highly significant (p < 0.001) for all outcome parameters: mortality (3.97% vs. 1.30%), CCI® (28.96 vs. 18.29), therapy-related QALD-loss (24.41 vs. 15.19 days) and QALY-gain (0.351 vs. 0.501 years). CONCLUSION: MEBDAS® decisions are superior to those of MTB in terms of in-hospital-outcome. The inclusion of MEBDAS® in decision procedure makes the indication more accurate and reduces morbidity and mortality. In addition, MEBDAS® can increase patients' competence by involving them in decision-making process.

An adaptive trial design to optimize dose-schedule regimes with delayed outcomes.

This paper proposes a two-stage phase I-II clinical trial design to optimize dose-schedule regimes of an experimental agent within ordered disease subgroups in terms of the toxicity-efficacy trade-off. The design is motivated by settings where prior biological information indicates it is certain that efficacy will improve with ordinal subgroup level. We formulate a flexible Bayesian hierarchical model to account for associations among subgroups and regimes, and to characterize ordered subgroup effects. Sequentially adaptive decision-making is complicated by the problem, arising from the motivating application, that efficacy is scored on day 90 and toxicity is evaluated within 30 days from the start of therapy, while the patient accrual rate is fast relative to these outcome evaluation intervals. To deal with this in a practical manner, we take a likelihood-based approach that treats unobserved toxicity and efficacy outcomes as missing values, and use elicited utilities that quantify the efficacy-toxicity trade-off as a decision criterion. Adaptive randomization is used to assign patients to regimes while accounting for subgroups, with randomization probabilities depending on the posterior predictive distributions of utilities. A simulation study is presented to evaluate the design's performance under a variety of scenarios, and to assess its sensitivity to the amount of missing data, the prior, and model misspecification.

Diagnosis of peanut allergy using continuous likelihood ratios.

BACKGROUND: Peanut allergy (PA) is associated with an economic and psychological burden on patients and families. Its diagnosis includes tests for peanut specific immunoglobulin E (sIgE), the values of which usually are categorized as positive or negative using a predefined cutoff (usually 0.35 kU/L). With the use of Bayes' theorem, this categorization can be replaced with a continuous interpretation of sIgE, which accounts for the prevalence of PA and history of ingestion. OBJECTIVE: To evaluate a method for estimating the likelihood ratio (LR) for each value of sIgE by performing a pilot investigation with the results of oral food challenges. The LR could be used to estimate the probability of PA. METHODS: The outcomes of oral food challenges and serum IgE values from 117 children seen in an allergy clinic between January 2017 and November 2019 were obtained. Polynomial regression of the receiver operation characteristics curve was used to determine an LR for each value of sIgE. Linear regression was used to estimate an LR for each value of sIgE. RESULTS: sIgE ranged from less than 0.1 kU/L to 35 kU/L. Bayes' theorem and a receiver operation characteristics curve were used to estimate LRs for each value of peanut sIgE. The value of IgE associated with an LR of 1 was 0.22 kU/L, which is comparable to other studies that used a value of 0.35 kU/L to separate positive from negative results. CONCLUSION: When combined with estimates of pretest probability, this method should permit the development of computerized decision-making algorithms to estimate the probability that a patient has PA.

Epidemics Fuzzy Decision-Making Applications and Fuzzy Genetic Algorithms Efficiency Enhancement.

Fuzzy logic is an innovative scientific field with several successful applications. Genetic algorithms and fuzzy logic systems fusion provide real-world problems modeling through the development of intelligent and adaptive systems. Moreover, the statistical analysis of the epidemiology of infectious diseases, which combines fuzzy logic aspects, is vital for perceiving their evolution and control potential. Author's objective is initially to provide a review of the efficiency of fuzzy logic applications. The advanced implementation of fuzzy logic theory in epidemiology and the application of fuzzy logic for controlling genetic algorithms within strategies based on the human experience and knowledge known as fuzzy logic controllers (FLCs) are analyzed. Outcomes of this review study show that not only can fuzzy sets be efficiently implemented in epidemiology but also prove the effectiveness of fuzzy genetic algorithms applications, thus suggesting that fuzzy logic applications are a really promising field of research.

Computer-assisted preoperative planning improves the learning curve of PFNA-II in the treatment of intertrochanteric femoral fractures.

BACKGROUND: Intertrochanteric femoral fractures are prevalent among the elderly, and usually demands surgical treatments. Proximal femoral nail antirotation Asian version (PFNA-II) is widely used for intertrochanteric fracture treatment. The computer-assisted preoperative planning (CAPP) system has the potential to reduce the difficulty of PFNA-II in the treatment of intertrochanteric fractures. The aim of the study was to investigate and compare the learning curves of PFNA-II treatment with CAPP and conventional preoperational planning methods for intertrochanteric femoral fractures. METHODS: A total of 125 patients with intertrochanteric fracture who were treated with PFNA-II between March 2012 and June 2015 were retrospectively analyzed. Patients who underwent surgery with CAPP procedure by a junior surgeon were regarded as group A (n = 53); patients who underwent the conventional surgery by another junior surgeon were regarded as group B (n = 72). Each group was divided into three subgroups (case 1-20, case 21-40, case 41-53 or case 41-72). RESULTS: The average operation time of group A was 45.00(42.00, 50.00) minutes, and in group B was 55.00 (50.00, 60.00) minutes (P < 0.01). Average radiation frequency and blood loss were 13.02 ± 2.32, 160.00 (140.00, 170.00) ml and 20.92 ± 3.27, 250.00 (195.00, 279.50) ml, respectively, with significant differences (P < 0.01). The learning curve of the surgical procedure in group A was steeper than that in group B. There were no significant differences in patient reported outcomes, hospital stay and complication rate between the two groups. Significant differences were observed between group A and B in Harris score at last follow-up in the AO/OTA type 31-A2 intertrochanteric fracture (P < 0.05). CONCLUSION: Compared with conventional preoperative planning methods, CAPP system significantly reduced operation time, radiation frequency and blood loss, thus reshaped the learning curve of PFNA-II treatment with lower learning difficulty. TRIAL REGISTRATION: researchregistry4770. Registered 25 March 2019.