Adverse Outcome Pathways (AOPs) are conceptual frameworks that tie an initial perturbation (molecular initiat- ing event) to a phenotypic toxicological manifestation (adverse outcome), through a series of steps (key events). They provide therefore a standardized way to map and organize toxicological mechanistic information. As such, AOPs inform on key events underlying toxicity, thus supporting the development of New Approach Methodologies (NAMs), which aim to reduce the use of animal testing for toxicology purposes. However, the establishment of a novel AOP relies on the gathering of multiple streams of evidence and infor- mation, from available literature to knowledge databases. Often, this information is in the form of free text, also called unstructured text, which is not immediately digestible by a computer. This information is thus both tedious and increasingly time-consuming to process manually with the growing volume of data available. The advance- ment of machine learning provides alternative solutions to this challenge. To extract and organize information from relevant sources, it seems valuable to employ deep learning Natural Language Processing techniques. We review here some of the recent progress in the NLP field, and show how these techniques have already demonstrated value in the biomedical and toxicology areas. We also propose an approach to efficiently and reliably extract and combine relevant toxicological information from text. This data can be used to map underlying mechanisms that lead to toxicological effects and start building quantitative models, in particular AOPs, ultimately allowing animal-free human-based hazard and risk assessment.
Purpose: Instability of the knee joint is reported by a majority (>65%) of patients with knee osteoarthritis (KOA) and is hypothesized to play a crucial role in the initiation and progression of KOA. A generally accepted objective metric of knee joint stability is lacking, making development of diagnostics and treatment options for knee joint instability more difficult. Such a metric should be based on how gait biomechanics and muscle activation in the unstable knee joint differ from those in a stable knee joint. To challenge knee joint instability, external perturbations during gait are needed to replicate the situations in daily life that require stability of the knee joint. Therefore, the aim of this study was to compare the responses in knee biomechanics and muscle activation patterns to different types of external perturbations during gait of patients with self-reported knee joint instability (KOA-I) versus patients reporting stable knees (KOA-S) and healthy control subjects.Methods: Forty patients (60% female) were included in this study with a mean age of 66 years (range: 52-82), body mass index of 26 (range: 19-32) and Kellgren and Lawrence grade of 2.5 (range 0-4). Patients were dichotomized in a KOA-I group (n=20) and KOA-S group (n=20) based on if they had perceived an episode of knee joint instability in the past four weeks. Furthermore, twenty age-, gender- and BMI-matched healthy control subjects were measured. The participants walked on a dual-belt instrumented treadmill while different external perturbations were applied, triggered by heel strike of the most affected leg (figure 1). The external perturbations consisted of sway left (SL) or sway right (SR) translations (4 cm) or accelerations (AC) or decelerations (DC) of one belt (1.6 m/s walking speed change in 0.23 seconds). Knee kinematics and muscle activation patterns of the perturbed gait cycles were collected using a motion capture system and surface electromyography. The three groups were compared using statistical parametric mapping (SPM) and discrete values by analysis of variance. The discrete values of the knee angles (initial contact, peak and range of motion (ROM) values) and muscle activation patterns (peak, mean and co-contraction index (CCI) values) were corrected for walking speed.Results: The SPM analysis results (example provided in figure 2) showed that in response to the SL perturbations the KOA-I group walked with greater knee flexion angles (KFA) during pre-swing compared to the control group (SPM, p<0.01) and during mid-swing compared to the KOA-S group and control group (SPM, p<0.01). Moreover, during the SR perturbed gait cycles the KOA-I group had greater KFA during mid-swing compared to the KOA-S group (SPM, p=0.01). In response to the AC perturbations the KOA-I group walked with a greater KFA during late terminal stance compared to the control group (SPM, p<0.01). Furthermore, the KOA-I group had greater KFA during the pre-swing phase of the DC perturbed gait cycles compared to the control group (SPM, p<0.01). The significant results from the comparison of the discrete values are presented in table 1. The KOA-I group had greater peak KFA during the swing phase of all perturbed gait cycles (independent of perturbation type) compared to the KOA-S group and control group (p<0.01). Moreover, during both sway perturbations (SL, SR) higher KFA ROM were observed in the KOA-I group compared to the KOA-S group (p<0.05). Besides this, the KOA-I group presented higher CCI of the medial muscles (vastus medialis and medial hamstring) compared to the KOA-S group during the DC perturbation (p=0.03). Furthermore, changes in vastus medialis and gluteus medius muscle activation in response to different external perturbations were observed in the KOA-S group compared to the control group and the KOA-I group (p<0.05).Conclusions: Patients with KOA-I walked with greater knee flexion angles during peak stance, late-terminal stance, pre-swing and mid-swing in response to different external perturbations, which could be a distinctive strategy of these patients to maintain stability of the knee joint during these phases of gait. Besides this, only few alterations were observed in the knee muscle activation patterns between the groups. This could be explained by the large variation between subjects in the muscle activations patterns which might indicate different neuromuscular strategies to respond to the external perturbations. Future studies with larger sample sizes are required to test the reliability and validity of the knee flexion angle as a candidate for the objective measurement of knee joint stability and to further investigate neuromuscular control of the unstable osteoarthritic knee.
Background: Steady-state gait characteristics appear promising as predictors of falls in stroke survivors. However, assessing how stroke survivors respond to actual gait perturbations may result in better fall predictions. We hypothesize that stroke survivors who fall have a diminished ability to adequately adjust gait characteristics after gait is perturbed. This study explored whether gait characteristics of perturbed gait differ between fallers and non fallers. Method: Chronic stroke survivors were recruited by clinical therapy practices. Prospective falls were monitored over a six months follow up period. We used the Gait Real-time Analysis Interactive Lab (GRAIL, Motekforce Link B.V., Amsterdam) to assess gait. First we assessed gait characteristics during steady-state gait and second we examined gait responses after six types of gait perturbations. We assessed base of support gait characteristics and margins of stability in the forward and medio-lateral direction. Findings: Thirty eight stroke survivors complete our gait protocol. Fifteen stroke survivors experienced falls. All six gait perturbations resulted in a significant gait deviation. Forward stability was reduced in the fall group during the second step after a ipsilateral perturbation. Interpretation: Although stability was different between groups during a ipsilateral perturbation, it was caused by a secondary strategy to keep up with the belt speed, therefore, contrary to our hypothesis fallers group of stroke survivors have a preserved ability to cope with external gait perturbations as compared to non fallers. Yet, our sample size was limited and thereby, perhaps minor group differences were not revealed in the present study.
LINK
