Purpose: To evaluate factors impacting the Segment Anything Model (SAM) and variant MedSAM performance for segmenting liver observations on contrast-enhanced (CE) magnetic resonance imaging (MRI) in high-risk patients with probable hepatocellular carcinoma (HCC) (LR-4) and definite HCC (LR-5). Methods: A retrospective cohort of liver observations (LR-4/LR-5) on CE-MRI from 97 patients at high-risk for HCC was derived (2013-2018). Using bounding-boxes as prompts under 5-fold cross-validation, segmentation performance was evaluated at the model and liver observation-levels for: (1) model types: SAM versus MedSAM, (2) image sizes: 256 × 256 versus 512 × 512, (3) image channel composition: CE sequences at 3 phases of enhancement independently and combined, (4) liver observation size: >10 mm versus >20 mm, (5) certainty of diagnosis: LR-4 versus LR-5, and (6) contrast-agent type: hepatobiliary versus extracellular. Segmentation performance, quantified using Dice coefficient, were compared using univariate (Wilcoxon signed-rank and t-test) and multivariable analyses (multiple correspondence analysis and subsequent linear modelling). Results: MedSAM trained on 512 × 512 combined CE sequences performed best with mean Dice coefficient 0.68 (95% confidence interval 0.66, 0.69). Overall, all factors except contrast-agent type affected performance, with larger image size resulting in the highest performance improvement (512 × 512: 0.57, 256 × 256: 0.26, P < .001) at the model-level. Contrast-agents affected performance for patients with LR-4 observations using MedSAM-based models (P < .03). Larger observation size, image size, and higher certainty of diagnosis were associated with better segmentation on multivariable analysis. Conclusion: A variety of factors were found to impact SAM/MedSAM performance for segmenting liver observations in patients with probable and definite HCC on CE-MRI. Future models may be optimized by accounting for these factors.

Lack of eye contact and imitation deficits are frequently targeted in behavioral interventions for children with autism spectrum disorder (ASD). In this study, we examined the effects of prompting and modeling on the imitation skills and eye contact of three Arabic-speaking young children with ASD in Syria. A multiple baseline design with a withdrawal component was used to evaluate the effects of the intervention in a clinical setting, at a center for children with special needs, and in follow-up sessions conducted in the participants\' homes. All participants\' imitative responses and eye contact increased when prompting and modeling were used. Our findings support the effectiveness of prompting and modeling on imitation skills.

BACKGROUND: Large language models (LLMs) have shown remarkable capabilities in natural language processing (NLP), especially in domains where labeled data are scarce or expensive, such as the clinical domain. However, to unlock the clinical knowledge hidden in these LLMs, we need to design effective prompts that can guide them to perform specific clinical NLP tasks without any task-specific training data. This is known as in-context learning, which is an art and science that requires understanding the strengths and weaknesses of different LLMs and prompt engineering approaches.
OBJECTIVE: The objective of this study is to assess the effectiveness of various prompt engineering techniques, including 2 newly introduced types-heuristic and ensemble prompts, for zero-shot and few-shot clinical information extraction using pretrained language models.
METHODS: This comprehensive experimental study evaluated different prompt types (simple prefix, simple cloze, chain of thought, anticipatory, heuristic, and ensemble) across 5 clinical NLP tasks: clinical sense disambiguation, biomedical evidence extraction, coreference resolution, medication status extraction, and medication attribute extraction. The performance of these prompts was assessed using 3 state-of-the-art language models: GPT-3.5 (OpenAI), Gemini (Google), and LLaMA-2 (Meta). The study contrasted zero-shot with few-shot prompting and explored the effectiveness of ensemble approaches.
RESULTS: The study revealed that task-specific prompt tailoring is vital for the high performance of LLMs for zero-shot clinical NLP. In clinical sense disambiguation, GPT-3.5 achieved an accuracy of 0.96 with heuristic prompts and 0.94 in biomedical evidence extraction. Heuristic prompts, alongside chain of thought prompts, were highly effective across tasks. Few-shot prompting improved performance in complex scenarios, and ensemble approaches capitalized on multiple prompt strengths. GPT-3.5 consistently outperformed Gemini and LLaMA-2 across tasks and prompt types.
CONCLUSIONS: This study provides a rigorous evaluation of prompt engineering methodologies and introduces innovative techniques for clinical information extraction, demonstrating the potential of in-context learning in the clinical domain. These findings offer clear guidelines for future prompt-based clinical NLP research, facilitating engagement by non-NLP experts in clinical NLP advancements. To the best of our knowledge, this is one of the first works on the empirical evaluation of different prompt engineering approaches for clinical NLP in this era of generative artificial intelligence, and we hope that it will inspire and inform future research in this area.

BACKGROUND: School canteens are a recommended setting to influence adolescent nutrition due to their scope to improve student food choices. Online lunch ordering systems (\"online canteens\") are increasingly used and represent attractive infrastructure to implement choice architecture interventions that nudge users toward healthier food choices. A recent cluster randomized controlled trial demonstrated the short-term effectiveness (2-month follow-up) of a choice architecture intervention to increase the healthiness of foods purchased by high school students from online canteens. However, there is little evidence regarding the long-term effectiveness of choice architecture interventions targeting adolescent food purchases, particularly those delivered online.
OBJECTIVE: This study aimed to determine the long-term effectiveness of a multi-strategy choice architecture intervention embedded within online canteen infrastructure in high schools at a 15-month follow-up.
METHODS: A cluster randomized controlled trial was undertaken with 1331 students (from 9 high schools) in New South Wales, Australia. Schools were randomized to receive the automated choice architecture intervention (including menu labeling, positioning, feedback, and prompting strategies) or the control (standard online ordering). The foods purchased were classified according to the New South Wales Healthy Canteen strategy as either \"everyday,\" \"occasional,\" or \"should not be sold.\" Primary outcomes were the average proportion of \"everyday,\" \"occasional,\" and \"should not be sold\" items purchased per student. Secondary outcomes were the mean energy, saturated fat, sugar, and sodium content of purchases. Outcomes were assessed using routine data collected by the online canteen.
RESULTS: From baseline to 15-month follow-up, on average, students in the intervention group ordered significantly more \"everyday\" items (+11.5%, 95% CI 7.3% to 15.6%; P<.001), and significantly fewer \"occasional\" (-5.4%, 95% CI -9.4% to -1.5%; P=.007) and \"should not be sold\" items (-6%, 95% CI -9.1% to -2.9%; P<.001), relative to controls. There were no between-group differences over time in the mean energy, saturated fat, sugar, or sodium content of lunch orders.
CONCLUSIONS: Given their longer-term effectiveness, choice architecture interventions delivered via online canteens may represent a promising option for policy makers to support healthy eating among high school students.
BACKGROUND: Australian Clinical Trials ACTRN12620001338954, https://anzctr.org.au/Trial/Registration/TrialReview.aspx?id=380546 ; Open Science Framework osf.io/h8zfr, https://osf.io/h8zfr/.

In simultaneous prompting procedures, an immediate (i.e., 0-s) prompt is presented during all training trials, and transfer to the target discriminative condition is assessed during daily probes. Previous research suggests that simultaneous prompting procedures are efficacious and may produce acquisition in fewer errors to mastery when compared to prompt delay procedures. To date, only a single study on simultaneous prompting has included intraverbal targets. The current study evaluated the efficacy of a simultaneous prompting procedure on the acquisition of intraverbal synonyms for six children at risk for reading failure. Simultaneous prompting alone produced responding at mastery levels in seven of the 12 evaluations. Antecedent-based procedural modifications were effective in four of the five remaining evaluations. Errors were generally low for all but one participant. The current findings support the use of simultaneous prompting procedures when targeting intraverbals for young children exhibiting reading deficits.

To classify sentences in cardiovascular German doctor\'s letters into eleven section categories, we used pattern-exploiting training, a prompt-based method for text classification in few-shot learning scenarios (20, 50 and 100 instances per class) using language models with various pre-training approaches evaluated on CARDIO:DE, a freely available German clinical routine corpus. Prompting improves results by 5-28% accuracy compared to traditional methods, reducing manual annotation efforts and computational costs in a clinical setting.

Although nudging has been found to promote the choice of healthy foods in lab studies and ad-hoc field studies, relatively little research is available regarding effectiveness in real food venues that operate for profit. The paucity of empirical studies providing \"proof of implementation\" reveals the difficulty of applying previous empirical findings on nudging to mass-eating food services contexts, which serve meals to a lot of individuals daily. Based on the typology of choice architecture in food choice contexts, we closely collaborated with the in-house food service operator to devise and implement five nudge interventions to promote fruits and vegetables (FV) in university cafeterias. Each study was conducted for one 12-week semester or more over a three-year period. In the first two studies, non-verbal point-of-purchase prompting increased the choice of kale/spinach supplemented smoothies and whole fruits from baskets. In Study 3, the combination of sizing and point-of-purchase non-verbal prompting increased the sale of large size vegetable-rich bowls from a stir-fry grill. In Study 4, the proximity type of nudging by altering the position of the healthier option in a sandwich bar in combination with non-verbal prompting increased the sale of sandwiches containing spinach. In Study 5, the combination of sizing and proximity of large vs. small sized plates and serving spoons had no effect on sale of self-serve items in a salad bar. All the interventions except for Study 5 produced a moderate effect in increasing the choice of FV-rich items. We recommend that hospitality and food service operators consider operational parameters and simultaneously adopt more than one nudging components to achieve a sizable effect. Future randomized controlled trials are needed to implement choice architecture techniques in collaboration with food service companies.

Applied research on decreasing pedestrian injuries often focuses on how to increase driver yielding behavior but rarely studies what pedestrians can do to increase their safety. There is a lack of empirical research focusing on how pedestrians can effectively signal their need to cross the street when there is no traffic light directing the pedestrian and oncoming traffic. As a replication and extension of Crowley-Koch et al. (2011), this study examined the effects of two pedestrian gestures, an extended arm and raised hand, on driver yielding behavior at 3 crosswalks in Oklahoma City. Research assistants implemented gestures prior to crossing the street as cars approached the crosswalk. Data were collected on the percentage of drivers yielding to the pedestrian. Both pedestrian gestures increased driver yielding across all 3 sites when compared to no gesture. Results were discussed in terms of future research and practical solutions towards increasing pedestrian safety.

When reported, the methods for prompting functional communication responses (FCRs) following problem behavior during functional communication training (FCT) vary. Some researchers have prompted the FCR immediately following problem behavior but doing so may inadvertently strengthen problem behavior as the first link in an undesirable response chain. This study investigated the effects of prompting FCRs following problem behavior during FCT with 4 children who exhibited severe problem behavior. Problem behavior remained low and FCR rates were near optimal when prompts were delivered immediately following problem behavior for 2 participants. Delaying prompts following problem behavior was instrumental for FCR acquisition for 1 participant but led to escalation of problem behavior for a 2nd participant. The conditions under which immediate prompts following problem behavior may improve or worsen FCT are discussed.

Prompting children to look at print and picture content during shared book reading (SBR) facilitates joint attention and early language and literacy learning opportunities for typically developing (TD) children. Whether preschoolers with autism spectrum disorder (ASD) respond similarly to bids for joint attention during SBR and how autism characteristics impact upon their responsiveness is currently unclear. This is important given these children are at risk of persistent language and literacy challenges. To address this, we examined the effects of prompts that were solely verbal versus verbal with pointing on visual attention to print and picture targets during SBR with digital storybooks for 34 children with ASD and 27 TD peers. Children with ASD looked as frequently at print targets, but less frequently at picture targets, when prompted compared to TD peers. Both prompt types showed similar effects in shifting children\'s visual attention to print and picture targets at group level. When groups were combined, autism characteristics influenced children\'s responsiveness to verbal versus verbal with pointing prompts to print targets, but not to picture targets; children looked more frequently at print targets as autism characteristics increased when verbal prompts were used, with a large effect shown (d = 0.91). Overall, findings suggest that prompting children with ASD to look at print and pictures during SBR with digital storybooks may be helpful in facilitating joint attention to storybook content. Implications for the development of effective early interventions aimed at providing emergent literacy support for children with ASD are discussed. LAY SUMMARY: We looked at how prompting impacted upon where children with autism look during shared book reading with digital storybooks compared to children without autism using eye-tracking. We found the target (pictures or print) was what mattered and number of autism characteristics impacted how responsive children were to different types of prompts. This helps us to understand how prompting may help children with autism to look at print or pictures during shared book reading which could support language or reading interventions.