Towards the individualization of vaginal dilatation exercises: A Quantitative analysis of the variability of vaginal pressure responses
Varagnolo, D1; Knorn, S2; Chiva, EO1; Melles, RJ3; Dewitte, M4
1: Luleå University of Technology, Sweden; 2: Uppsala University, Sweden; 3: Maastricht University Medical Center, The Netherlands; 4: Maastricht University, The Netherlands
Objective(s): We consider the broad question of whether it is meaningful to individualize vaginal dilation patterns when treating patients affected by dyspareunia using a vaginal dilator. More precisely, we aim at quantitatively evaluating: 1) how different individuals may have different physiological responses to the same vaginal dilation patterns with the aim of obtaining models that are useful to decide how to personalize the vaginal dilation patterns; 2) what are the drawbacks, from modeling perspectives, of grouping individuals into one single class.
Material and Method(s): We consider time-series data collected from 46 healthy patients during ad-hoc medical trials consisting of gradually dilating the vaginal duct while registering the exerted pelvic floor pressure. Starting from this data we perform a series of individualized data-driven modeling of the physiological responses of the various patients. We then evaluate the differences among the estimated models, and quantitatively assess how well an “average-patient” model can reproduce the measured responses of the various individuals.
Result(s): We first establish that the measured responses contain dynamics, and this implies that the personalization of vaginal dilation patterns cannot be performed using static models. We then verify that different patients may have physiological responses that are variegate. Hence designing vaginal dilation patterns using an “average-patient” model is likely to ignore variations in the responses and needs of the individual patients, and be for this reason suboptimal.
Conclusion(s): The first step of the personalization of clinical interventions using vaginal dilatation exercises should be to estimate how the patient dynamically responds to stimuli. Each patient has indeed an individual physiological response that is unlikely to be accurately reproduceable through pre-compiled “average-patient” models.
Work supported by industry: no.Go Back