Tommaso Giacometti

Dr. Tommaso Giacometti

DIFA, University of Bologna, Bologna, Emilia-Romagna, Italy
orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 June 2025 in Entropy
Verified
Journal Article
Open access
PDF file
TGTommaso Giacometti
Tommaso Giacometti
EREttore Rocchi
Ettore Rocchi
Pier Giorgio CojuttiPier Giorgio Cojutti
Pier Giorgio Cojutti
FMFederico Magnani
Federico Magnani
Daniel RemondiniDaniel Remondini
Daniel Remondini
Gastone CastellaniGastone Castellani
Gastone Castellani
This study investigates the use of Neural Ordinary Differential Equations (NODEs) as an alternative to traditional compartmental models and Nonlinear Mixed-Effects (NLME) models for drug concentration prediction in pharmacokinetics. Unlike standard models that rely on strong assumptions and often struggle with high-dimensional covariate relationships, NODEs offer a data-driven approach, learning differential equations directly from data while integrating covariates. To evaluate their performance, NODEs were applied to a real-world Dalbavancin pharmacokinetic dataset comprising 218 patients and compared against a two-compartment model and an NLME within a cross-validation framework, which ensures an evaluation of robustness. Given the challenge of limited data availability, a data augmentation strategy was employed to pre-train NODEs. Their predictive performance was assessed both with and without covariates, while model explainability was analyzed using Shapley additive explanations (SHAP) values. Results show that, in the absence of covariates, NODEs performed comparably to state-of-the-art NLME models. However, when covariates were incorporated, NODEs demonstrated superior predictive accuracy. SHAP analyses further revealed how NODEs leverage covariates in their predictions. These results establish NODEs as a promising alternative for pharmacokinetic modeling, particularly in capturing complex covariate interactions, even when dealing with sparse and small datasets, thus paving the way for improved drug concentration predictions and personalized treatment strategies in precision medicine.

20 February 2026
Verified
Book Chapter
Closed access
TGTommaso Giacometti
Tommaso Giacometti
Nico CurtiNico Curti
Nico Curti
AZAdriano Zaghi
Adriano Zaghi
Daniel RemondiniDaniel Remondini
Daniel Remondini
Gastone CastellaniGastone Castellani
Gastone Castellani

5 June 2026 in Experimental Dermatology
Verified
Journal Article
Open access
CZCorrado Zengarini
Corrado Zengarini
TGTommaso Giacometti
Tommaso Giacometti
YMYuri Merli
Yuri Merli
DGDavide Griffa
Davide Griffa
LRLuca Rapparini
Luca Rapparini
Nico CurtiNico Curti
Nico Curti
Wound edge assessment is a key component of chronic wound evaluation, but it remains highly subjective and affected by inter‐observer variability, particularly when performed on two‐dimensional clinical photographs. We retrospectively analysed 1 860 wound images acquired during routine clinical practice and independently annotated by four expert clinicians. An automated image‐analysis pipeline was used to segment the wound, standardise the peri‐wound border region, and estimate the three‐dimensional profile of the wound edge. We first tested whether geometry‐derived edge profiles alone could reproduce clinical wound edge categories. We then evaluated whether adding global visual descriptors of wound shape, colour appearance, and surface pattern improved agreement with clinicians. Inter‐clinician agreement was low, confirming the intrinsic subjectivity of wound edge classification. Geometry‐based analysis identified coherent edge‐profile patterns but showed poor correspondence with clinical annotations. In contrast, a supervised classifier incorporating both geometric and visual features achieved agreement comparable to, and in some comparisons higher than, the agreement observed among clinicians. Clinical wound edge assessment is not driven by edge geometry alone. Visual cues such as wound shape, colour appearance, and surface pattern appear to influence expert classification and may contribute to variability. Automated image‐based analysis may support more reproducible wound edge assessment, provided that it is externally validated in diverse clinical settings.