COMPARISON OF STATISTICAL, MACHINE LEARNING, AND MATHEMATICAL MODELLING METHODS TO INVESTIGATE THE EFFECT OF AGEING ON DOG’S CARDIOVASCULAR SYSTEM

Aug 5, 2026

Upload by Finn

A preliminary comparison evaluates statistical, machine learning, and mathematical modelling approaches for automatically detecting ageing-related effects from in vivo cardiovascular data in laboratory dogs. The study is motivated by safety pharmacology, where pre-clinical assessment determines whether a drug may pose health risks before clinical trials. Using telemetry recordings over multiple weeks, the work also explores whether individual “fingerprints” can be identified from haemodynamic signals by computer algorithms to inform interindividual variability under ageing.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages26

Reading Time1 h 6 min

AccessFree to download

Introduction

Motivation

Methods

ESAIM: PROCEEDINGS AND SURVEYS, September 2023, Vol. 73, p. 2-27 Virginie Ehrlacher, Damiano Lombardi, Olga Mula, Fabio Nobile, Tommaso Taddei COMPARISON OF STATISTICAL, MACHINE LEARNING, AND MATHEMATICAL MODELLING METHODS TO INVESTIGATE THE EFFECT OF AGEING ON DOG’S CARDIOVASCULAR SYSTEM Elham Ataei Alizadeh 1 , Sara Costa Faya 2 , Haibo Liu 32 , Damiano Lombardi 2 , Sylvain Bernasconi 3 , Pieter-Jan Guns 4 and Michael Markert 1 Abstract. The aim of this work is to provide a preliminary comparison of different classes of methods to automatically detect the effect of ageing from in vivo data. The application which motivated this work is related to safety pharmacology, whose major goal is to determine, in a pre-clinical phase, whether a drug is potentially dangerous for the health [1] . In particular, we are going to compare statistical, machine learning and mathematical modelling methods. Résumé . L’objectif de ce travail est de fournir une comparaison préliminaire entre différents classes de méthodes pour la détection automatique de l’effet du viellissement sur le système cardiovasculaire, en exploitant des données in vivo. L’application qui a motivé ce travail est liée à la pharmacologie desécurité, qui vise à établir, dans une phase pre-clinique, si un médicament est potentiellement dangereux pour la santé [1] . En particulier, nous allons comparer des approches statistiques, d’apprentissage statistique et de modélisation mathématique. 1. Introduction 1.1. Motivation This work has been motivated by some questions arising in safety pharmacology. Safety pharmacology studies are designed to identify and assess the potential clinical risk of undesirable drug properties before they enter clinical trials, as described in [2] . Many drug development processes must proceed through several stages to be sure for a product to be safe, efficacious, and has passed all regulatory requirements. The preclinical stage encompasses the use of in vitro and in vivo studies to develop a drug that can safely and effectively be administered for clinical trials. In vivo 1 General Pharmacology Group, Department of Drug Discovery Support, Boehringer Ingelheim Pharma GmbH & Co KG, Biberach an der Riss, Germany 2 Sorbonne Université and COMMEDIA team, Inria, Paris, France 3 NOTOCORD, an Instem company, Le Pecq, France 4 Laboratory of Physiopharmacology, University of Antwerp, Antwerp, Belgium © EDP Sciences, SMAI 2023 This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( [https://creativecommons.org/l](https://creativecommons.org/l)icenses/by/4 .0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Article published online by EDP Sciences and available at [https://www.esaim-proc.org](https://www.esaim-proc.org) or [https://doi.org/10.1051/proc/202373002](https://doi.org/10.1051/proc/202373002) ESAIM: PROCEEDINGS AND SURVEYS 3 studies performed in animals are essential to drug development because they have the ability to evaluate the effects a drug has on a living organism. A particular care is taken in assessing adverse effects and drug-drug interactions that cannot be observed in vitro [3] . When an animal participates in an experiment in safety pharmacology studies, one can anticipate that the compound tested might have an effect on the organism. It is therefore essential to know when the animal can participate again in an experiment after a sufficient wash-out period. This is of particular importance in cross-over design studies (for details, the reader can refer to [4]) . Before an animal will be used in a study, it has to undergo clinical evaluation as well as physiological tests to monitor the condition its cardiovascular system. When these initial tests are successfully performed the particular animal can be labelled as “healthy" and participate in the experiment. Age is one of the f

FAQs

What is the primary goal of this study?

To provide a preliminary comparison of statistical, machine learning, and mathematical modelling methods for automatically detecting the effect of ageing from in vivo cardiovascular data.

Why is ageing detection important in this context?

The work supports safety pharmacology, helping assess potential clinical risk of undesirable drug properties during the pre-clinical phase by understanding cardiovascular effects related to ageing.

What data type and acquisition setup are used to test the methods?

Cardiovascular activity data from laboratory dogs collected via telemetry over several weeks, where recordings are large and require automatic analysis rather than manual inspection.

Aug 5, 2026

Upload by Finn