…Pay special attention to units, time windows, and values read from figures. ``` --- ## Skill 2 — aa-screening **Purpose:** First-pass INCLUDE / EXCLUDE / UNCERTAIN screening of paper abstracts. **File:** `aa-screening/SKILL.md` ```markdown…
aa-screening
Filters
ore for Full CDA and for Refined CDA. Discriminant coefficients generated from the algorithms were also used to rank behavioral items, as proposed p reviously52. Principal component analysis (PCA) Principal component analysis was used as an unsupervised comparison to supervised technique (logistic regres- sion and canonical discriminant analysis). Principal Component Analysis is a dimensionality reduction technique that retains data variation that also might be used for testing the multiple association between variables53. It is performed by reducing the number of variables into principal components (PCs), where the data variation is maximal40. Similarly to CDA, a Full PCA and a Refined PCA without Activity, for the same reason, were per- formed and followed the same procedures described in this section. The number of PCs retained was defined by Horn’s parallel analysis using psych::fa.parallel on the train set. This method compares the factors scree of the observed data to a randomly generated one, of a data matrix of the same size as ours. The correlation matrix used Pearson correlation. The method was computed after 1,001 simulated analyses performed. PCA was then performed (stats::princomp) on the train set. Eigean values were calculated using the standard deviation of the principal components. Loading values were obtained using stats::loadings. The loading values were used to mutate the original scores in the test set, resulting in a new total score based on PCA weightings. The loading values were used as PCA weightings to obtain a new total score. For this purpose, each UPAPS item was multiplied by its respective PCA weighting (loading value), resulting in a new score for each item. The new scores were added, resulting in a new total score for Full PCA
ore for Full CDA and for Refined CDA. Discriminant coefficients generated from the algorithms were also used to rank behavioral items, as proposed p reviously52. Principal component analysis (PCA) Principal component analysis was used as an unsupervised comparison to supervised technique (logistic regres- sion and canonical discriminant analysis). Principal Component Analysis is a dimensionality reduction technique that retains data variation that also might be used for testing the multiple association between variables53. It is performed by reducing the number of variables into principal components (PCs), where the data variation is maximal40. Similarly to CDA, a Full PCA and a Refined PCA without Activity, for the same reason, were per- formed and followed the same procedures described in this section. The number of PCs retained was defined by Horn’s parallel analysis using psych::fa.parallel on the train set. This method compares the factors scree of the observed data to a randomly generated one, of a data matrix of the same size as ours. The correlation matrix used Pearson correlation. The method was computed after 1,001 simulated analyses performed. PCA was then performed (stats::princomp) on the train set. Eigean values were calculated using the standard deviation of the principal components. Loading values were obtained using stats::loadings. The loading values were used to mutate the original scores in the test set, resulting in a new total score based on PCA weightings. The loading values were used as PCA weightings to obtain a new total score. For this purpose, each UPAPS item was multiplied by its respective PCA weighting (loading value), resulting in a new score for each item. The new scores were added, resulting in a new total score for Full PCA
17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles
ntibody with corticoster- operation. Hatched areas represent portions of skin removed. one, cortisone, dexamethasone, prednisolone and prednisone were 0.2, < 0.1, < 0.1, 45.3 and 0.3%, respectively. The intra- assay coefficient of variation (CV) for samples containing 176.5, (oxytetracycline at 10 mg/kg) prior to mulesing. Blood samples 81.9 and 34.8 nmol/L cortisol respectively, were 9.3, 8.2, and were taken from the jugular vein via venepuncture into 5 ml EDTA vacutainers at 5 time points over the course of the exper- 5.9%. The inter-assay CVs for the same samples were 7.2, 9.3 and 8.9%, respectively, iment, commencing prior to analgesic treatment and again at 0.5, 6, 12 and 24h post-mulesing. On each occasion, lambs had Behaviour blood samples taken in the sequence in which they had been mulesed. The behaviour of each lamb was captured using four digital video cameras for a total of 12 hours after mulesing, consisting of The lambs received their experimental treatments at the times three periods of 4 hours’ duration (0 to 4, 4 to 8 and 8 to 12h relative to the mules operation described in Table 1. At the time post-mulesing). One operator, blinded to the treatment codes, of mulesing, lambs were caught and placed in a mulesing cradle classified the behaviour of all lambs in the experiment, into the and their tails were docked with a gas fired hot knife. The mules- categories shown in Table 2. The lambs were observed and their ing operator had many years of experience mulesing lambs and behaviours recorded at 5 minute intervals. Lamb postures were had completed accreditation training provided by the Livestock used as the main measure as these were identified in preliminary Contractors Association, Mulesing involved the excision of two research as most indicative of
17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 17510813, 2020, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/avj.12930. By National Health And Medical Research Council, Wiley Online Library on [30/03/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles
e FOS FOS staining staining intensity and frequency. (B) Area evaluated in the spinal cord. (C) Examples of different scores. intensity and frequency. (B) Area evaluated in the spinal cord. (C) Examples of different scores. The analysis 2.3. Statistical method was first established in a preliminary study (data not shown) to Analysis develop and define a scoring approach for formalin-fixed, paraffin-embedded and stained 2.3.1. EEG tissue sections. The staining intensity (none, slight, moderate, or strong) and the percentage The power of positive spectral neuronal density nuclei in bothof dorsal the selected horns EEG of theepisodes was calculated spinal cords with the were calculated. A MATLAB semiquantitative score including these two parameters (Figure 1A) was applied. from pwelch function and a frequency resolution of 0.25 Hz for the pooled data Four the interdigital different segmentspinch stimulus (L1, and 1S1-3) L2, L3, and Hz for the data in each recorded animal were evaluated. to other in responseThe resultsstimuli. of this Because evaluationof the larger clearly sample showed thesize in thescores highest pooledingroup for thecastrated the animals withoutwe pinch stimulus, decided pain relief. to useitawas Thus, higher frequency decided resolution to proceed to methodological with this better describe approach. the response. Density spectral arrays were also constructed, and the results are presented as heatmaps to depict the temporal evolution of the spectral EEG features. The DSA was derived from 5-s EEG episodes overlapping by 4 s. The results are presented as changes in the EEG power relative to a prestimulus baseline defined as the episode 50–40 s before the stimulus. The poststimulus value was divided by the prestimulus value. Thus, a value <1 indicates
e FOS FOS staining staining intensity and frequency. (B) Area evaluated in the spinal cord. (C) Examples of different scores. intensity and frequency. (B) Area evaluated in the spinal cord. (C) Examples of different scores. The analysis 2.3. Statistical method was first established in a preliminary study (data not shown) to Analysis develop and define a scoring approach for formalin-fixed, paraffin-embedded and stained 2.3.1. EEG tissue sections. The staining intensity (none, slight, moderate, or strong) and the percentage The power of positive spectral neuronal density nuclei in bothof dorsal the selected horns EEG of theepisodes was calculated spinal cords with the were calculated. A MATLAB semiquantitative score including these two parameters (Figure 1A) was applied. from pwelch function and a frequency resolution of 0.25 Hz for the pooled data Four the interdigital different segmentspinch stimulus (L1, and 1S1-3) L2, L3, and Hz for the data in each recorded animal were evaluated. to other in responseThe resultsstimuli. of this Because evaluationof the larger clearly sample showed thesize in thescores highest pooledingroup for thecastrated the animals withoutwe pinch stimulus, decided pain relief. to useitawas Thus, higher frequency decided resolution to proceed to methodological with this better describe approach. the response. Density spectral arrays were also constructed, and the results are presented as heatmaps to depict the temporal evolution of the spectral EEG features. The DSA was derived from 5-s EEG episodes overlapping by 4 s. The results are presented as changes in the EEG power relative to a prestimulus baseline defined as the episode 50–40 s before the stimulus. The poststimulus value was divided by the prestimulus value. Thus, a value <1 indicates
and a predictive variable that is the parameter to be tested78. AUC above 90 is considered ‘good’ and above 95 is ‘excellent’. The ROC curve for each algorithm was constructed using the condition (painful and pain-free) as a predictive variable and the UPAPS total sum or the probability of each pig needing analgesia based each algorithm as a predictive variable (pROC::ROC). To obtain the AUC, it is necessary to establish an optimal cut-off point for each predictive variable, which was based on the Youden index (YI). The YI is estimated based on the sum of specificity and sensitivity subtracted from 1, calculated for each value of the predictive variable. The YI represents the maximum specificity and sensitivity concomitantly, attributing similar importance of specificity and sensitivity to the optimal cut-off point. The algorithms’ AUC were compared with UPAPS’s AUC by the DeLong test (pROC::roc.test). Refining random forest We use the ranking established by random forest training to reduce the number of pain-altered behaviors in the algorithm based on a step-up backward procedure. In this way, an algorithm was built using the training base, with exclusively the 16 best-ranked pain-altered behaviors, and this procedure was repeated sequentially until an algorithm was built with only the two best-ranked pain-altered behaviors. Using the testing base, the AUC of all algorithms with different numbers of pain-altered behaviors was compared with the AUC of the algorithm containing all behaviors using the DeLong test. The algorithm with the smallest number of pain- altered behaviors (lowest complexity) having an AUC statistically equivalent to or greater than the algorithm with all pain-altered behaviors was understood as the best refinement. Scientific Reports |