Evaluating the predication model of metaphor comprehension: Using word2vec to model best/worst quality judgments of 622 novel metaphors.

In this paper our goal is to undertake a systematic assessment of the first, most widely known, and simplest computational model of metaphor comprehension, the predication model developed by Kintsch (Cognitive Science, 25(2), 173-202, 2000). 622 metaphors of the form "x is a y" were selected from a much larger set generated randomly. The metaphors were judged for quality using best/worst judgments, which asks judges to pick the best and worst metaphor from among four presented metaphors. The metaphors and their judgments have been publicly released. We modeled the judgments by extending Kintsch's predication model (2000) by systematically walking through the parameter space of that model. Our model successfully differentiated metaphors rated as good (> 1.5z) from metaphors rated as bad (< -1.5z; Cohen's d = 0.72) and was able to successfully classify good metaphors with an accuracy of 82.9%. However, it achieved a true negative rate below chance at 36.3% and had a resultantly low kappa of 0.037. The model could not distinguish unselected random metaphors from those selected by humans as having metaphorical potential. In a follow-up study we showed that the model's quality estimates reliably predict metaphor decision times, with better metaphors being judged more quickly than worse metaphors.

In vitro assessment of an idealized nose for nasal spray testing: Comparison with regional deposition in realistic nasal replicas.

An idealized nasal replica that mimics average regional deposition of nasal spray pump droplets in human nasal airways would potentially be useful in expediting the development of nasal spray products. The aim of this study was to validate an idealized nose, previously developed using in silico simulations, by comparing with regional deposition in realistic, sectioned nasal replicas obtained from in vitro deposition experiments. The realistic nasal airway replicas of five subjects obtained from computerized tomography were manufactured in plastic using rapid prototyping. The idealized nose was made using the same build procedure. A commercial nasal spray pump (NasalCrom, 5.2 mg cromolyn sodium per spray) was then actuated repeatably into each replica under a steady inspiratory flow of 7.5 L/min at two different orientations (45° and 60° from the horizontal). Sectioned replicas were disassembled, and the mass fraction of drug deposited on the surface of each anatomical region was determined. It was found that regional deposition of spray droplets in the idealized replica agreed well with average regional deposition in the realistic replicas. Regional deposition also agreed with previously published in vivo regional deposition using the same spray pump.

An idealized geometry that mimics average nasal spray deposition in adults: A computational study.

This work describes the development of an idealized geometry that mimics average regional deposition of nasal sprays within realistic adult nasal geometries. Previous simulation results in seven realistic nasal airways (Kiaee et al. Int. J. Num. Methods Biomed. Eng. 34: e2968, 2018) were used to establish target values of regional deposition. Characteristic geometric features observed to be common to all the realistic nasal airway geometries studied were extracted and included in the idealized geometry. Additional geometric features and size scaling were explored, in order to enhance deposition in specific regions based on the results of simulations done in preliminary versions of the idealized geometry. In total, more than one hundred thousand simulation cases were conducted across a range of particle parameters and geometric shapes in order to reach the final idealized geometry presented herein. For droplet velocities of 0-20 m/s, droplet sizes of 5-40 µm and at an inhalation flow rate of 15 l/min, regional deposition in the final idealized geometry compares favourably with average deposition in each of the vestibule, valve, olfactory, turbinate, nasopharynx, and outlet regions in the realistic geometries. The proposed idealized nasal geometry has potential for use in the development and testing of nasal drug delivery systems, allowing researchers to estimate in vivo regional nasal deposition patterns using a simple benchtop test apparatus.

Regional deposition of nasal sprays in adults: A wide ranging computational study.

The present work examines regional deposition within the nose for nasal sprays over a large and wide ranging parameter space by using numerical simulation. A set of 7 realistic adult nasal airway geometries was defined based on computed tomography images. Deposition in 6 regions of each nasal airway geometry (the vestibule, valve, anterior turbinate, posterior turbinate, olfactory, and nasopharynx) was determined for varying particle diameter, spray cone angle, spray release direction, particle injection speed, and particle injection location. Penetration of nasal spray particles through the airway geometries represented unintended lung exposure. Penetration was found to be relatively insensitive to injection velocity, but highly sensitive to particle size. Penetration remained at or above 30% for particles exceeding 10 µm in diameter for several airway geometries studied. Deposition in the turbinates, viewed as desirable for both local and systemic nasal drug delivery, was on average maximized for particles ranging from ~20 to 30 µm in diameter, and for low to zero injection velocity. Similar values of particle diameter and injection velocity were found to maximize deposition in the olfactory region, a potential target for nose-to-brain drug delivery. However, olfactory deposition was highly variable between airway geometries, with maximum olfactory deposition ranging over 2 orders of magnitude between geometries. This variability is an obstacle to overcome if consistent dosing between subjects is to be achieved for nose-to-brain drug delivery.