Detection, localisation, and quantification of neutrophils to assess disease activity and early response to therapy in Ulcerative Colitis: a novel AI-driven model

Marietta Iacucci; Valentina Vadori; Pablo Meseguer; Irene Zanmarchi; Giovanni Santacroce; Rocio Del Amor; Davide Zardo; Rory Crotty; Louise Burke; Bisi Bode Kolawole; Ujwala Kiran Chaudhari; Silvio Danese; Subrata Ghosh; Valery Naranjo; Enrico Grisan

doi:10.1016/j.eclinm.2025.103658

Detection, localisation, and quantification of neutrophils to assess disease activity and early response to therapy in Ulcerative Colitis: a novel AI-driven model

Marietta Iacucci
, Valentina Vadori
, Pablo Meseguer
, Irene Zanmarchi
, Giovanni Santacroce
, Rocio Del Amor
, Davide Zardo
, Rory Crotty
, Louise Burke
, Bisi Bode Kolawole
, Ujwala Kiran Chaudhari
, Silvio Danese
, Subrata Ghosh
, Valery Naranjo
, Enrico Grisan

University College of Cork,APC Microbiome Ireland, College of Medicine and Health,Cork,Ireland
Instituto de Investigación e Innovación en Bioingeniería, Universitat Politècnica de València, Valencia, Spain
San Bortolo Hospital, Department of Pathology , Vicenza,, Italy
Cork University Hospital, Department of Histopathology , Cork,, Ireland
Vita-Salute San Raffaele University

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

5 Downloads (Pure)

Abstract

Background
Histological remission (HR) is a key treatment goal in ulcerative colitis (UC), suggesting better disease management and treatment response. Absence of mucosal neutrophils is crucial to defining HR. However, the role of neutrophil quantity and localisation remains unclear. We aimed to develop a novel AI-driven pipeline to automate neutrophil detection, localisation, and quantification, supporting assessment of HR and treatment response.
Methods
We developed an AI-driven pipeline by integrating and combining the outputs of two deep learning models to segment whole slide images (WSIs) into epithelium, crypts and lamina propria and detect and quantify neutrophils. Optimal neutrophil density cut-offs to assess disease activity in UC patients from the AMAC phase two Mirikizumab trial (NCT02589665; 2015-2019) were identified and validated in the multicentre prospective PICaSSO cohort (2016-2019). Cut-offs to determine treatment response at weeks 12 and 52 were also evaluated.
Findings
303 WSIs of UC patients from the multicentre, randomised, double-blind, parallel-arm, placebo-controlled phase 2 clinical trial of Mirikizumab were analysed. The models yielded a 65. 44 0% DICE SØrensen for region segmentation on average and 82.3% precision for neutrophil detection. A density cut-off ≥21.7 cells/mm² assessed disease activity with 86% (95% CI: 79%-92%) accuracy, 94% (88%-99%) sensitivity, 74% (54%-89%) specificity, as validated in the PICaSSO cohort. Moreover, a cut-off <21.7 cells/mm² demonstrated 79% (95% CI: 67%-92%) and 85% (95% CI: 70%-96%) accuracy in assessing treatment response at week 12 and 52, respectively.
Interpretation
This novel AI-based pipeline demonstrates strong potential to detect, localise, and quantify neutrophils to assess histological activity in clinical trials and real-world settings. Furthermore, it effectively stratifies treatment response, offering a reliable and objective framework for personalised UC management.

Original language	English
Article number	103658
Journal	eClinicalMedicine
Volume	90
Early online date	25 Nov 2025
DOIs	https://doi.org/10.1016/j.eclinm.2025.103658
Publication status	Published - Dec 2025

Access to Document

10.1016/j.eclinm.2025.103658

1-s2.0-S2589537025005929-mainFinal published version, 929 KBLicence: CC BY-NC-ND 4.0

Cite this

Iacucci, M., Vadori, V., Meseguer, P., Zanmarchi, I., Santacroce, G., Del Amor, R., Zardo, D., Crotty, R., Burke, L., Kolawole, B. B., Chaudhari, U. K., Danese, S., Ghosh, S., Naranjo, V., & Grisan, E. (2025). Detection, localisation, and quantification of neutrophils to assess disease activity and early response to therapy in Ulcerative Colitis: a novel AI-driven model. eClinicalMedicine, 90, Article 103658. https://doi.org/10.1016/j.eclinm.2025.103658

@article{13f4c575616446778aeddcea77ca3f5d,

title = "Detection, localisation, and quantification of neutrophils to assess disease activity and early response to therapy in Ulcerative Colitis: a novel AI-driven model",

abstract = "Background Histological remission (HR) is a key treatment goal in ulcerative colitis (UC), suggesting better disease management and treatment response. Absence of mucosal neutrophils is crucial to defining HR. However, the role of neutrophil quantity and localisation remains unclear. We aimed to develop a novel AI-driven pipeline to automate neutrophil detection, localisation, and quantification, supporting assessment of HR and treatment response. Methods We developed an AI-driven pipeline by integrating and combining the outputs of two deep learning models to segment whole slide images (WSIs) into epithelium, crypts and lamina propria and detect and quantify neutrophils. Optimal neutrophil density cut-offs to assess disease activity in UC patients from the AMAC phase two Mirikizumab trial (NCT02589665; 2015-2019) were identified and validated in the multicentre prospective PICaSSO cohort (2016-2019). Cut-offs to determine treatment response at weeks 12 and 52 were also evaluated. Findings 303 WSIs of UC patients from the multicentre, randomised, double-blind, parallel-arm, placebo-controlled phase 2 clinical trial of Mirikizumab were analysed. The models yielded a 65. 44 0\% DICE S{\O}rensen for region segmentation on average and 82.3\% precision for neutrophil detection. A density cut-off ≥21.7 cells/mm² assessed disease activity with 86\% (95\% CI: 79\%-92\%) accuracy, 94\% (88\%-99\%) sensitivity, 74\% (54\%-89\%) specificity, as validated in the PICaSSO cohort. Moreover, a cut-off <21.7 cells/mm² demonstrated 79\% (95\% CI: 67\%-92\%) and 85\% (95\% CI: 70\%-96\%) accuracy in assessing treatment response at week 12 and 52, respectively. Interpretation This novel AI-based pipeline demonstrates strong potential to detect, localise, and quantify neutrophils to assess histological activity in clinical trials and real-world settings. Furthermore, it effectively stratifies treatment response, offering a reliable and objective framework for personalised UC management. ",

author = "Marietta Iacucci and Valentina Vadori and Pablo Meseguer and Irene Zanmarchi and Giovanni Santacroce and \{Del Amor\}, Rocio and Davide Zardo and Rory Crotty and Louise Burke and Kolawole, \{Bisi Bode\} and Chaudhari, \{Ujwala Kiran\} and Silvio Danese and Subrata Ghosh and Valery Naranjo and Enrico Grisan",

year = "2025",

month = dec,

doi = "10.1016/j.eclinm.2025.103658",

language = "English",

volume = "90",

journal = "eClinicalMedicine",

issn = "2589-5370",

publisher = "Elsevier",

}

Iacucci, M, Vadori, V, Meseguer, P, Zanmarchi, I, Santacroce, G, Del Amor, R, Zardo, D, Crotty, R, Burke, L, Kolawole, BB , Chaudhari, UK, Danese, S, Ghosh, S, Naranjo, V & Grisan, E 2025, 'Detection, localisation, and quantification of neutrophils to assess disease activity and early response to therapy in Ulcerative Colitis: a novel AI-driven model', eClinicalMedicine, vol. 90, 103658. https://doi.org/10.1016/j.eclinm.2025.103658

TY - JOUR

T1 - Detection, localisation, and quantification of neutrophils to assess disease activity and early response to therapy in Ulcerative Colitis: a novel AI-driven model

AU - Iacucci, Marietta

AU - Vadori, Valentina

AU - Meseguer, Pablo

AU - Zanmarchi, Irene

AU - Santacroce, Giovanni

AU - Del Amor, Rocio

AU - Zardo, Davide

AU - Crotty, Rory

AU - Burke, Louise

AU - Kolawole, Bisi Bode

AU - Chaudhari, Ujwala Kiran

AU - Danese, Silvio

AU - Ghosh, Subrata

AU - Naranjo, Valery

AU - Grisan, Enrico

PY - 2025/12

Y1 - 2025/12

N2 - Background Histological remission (HR) is a key treatment goal in ulcerative colitis (UC), suggesting better disease management and treatment response. Absence of mucosal neutrophils is crucial to defining HR. However, the role of neutrophil quantity and localisation remains unclear. We aimed to develop a novel AI-driven pipeline to automate neutrophil detection, localisation, and quantification, supporting assessment of HR and treatment response. Methods We developed an AI-driven pipeline by integrating and combining the outputs of two deep learning models to segment whole slide images (WSIs) into epithelium, crypts and lamina propria and detect and quantify neutrophils. Optimal neutrophil density cut-offs to assess disease activity in UC patients from the AMAC phase two Mirikizumab trial (NCT02589665; 2015-2019) were identified and validated in the multicentre prospective PICaSSO cohort (2016-2019). Cut-offs to determine treatment response at weeks 12 and 52 were also evaluated. Findings 303 WSIs of UC patients from the multicentre, randomised, double-blind, parallel-arm, placebo-controlled phase 2 clinical trial of Mirikizumab were analysed. The models yielded a 65. 44 0% DICE SØrensen for region segmentation on average and 82.3% precision for neutrophil detection. A density cut-off ≥21.7 cells/mm² assessed disease activity with 86% (95% CI: 79%-92%) accuracy, 94% (88%-99%) sensitivity, 74% (54%-89%) specificity, as validated in the PICaSSO cohort. Moreover, a cut-off <21.7 cells/mm² demonstrated 79% (95% CI: 67%-92%) and 85% (95% CI: 70%-96%) accuracy in assessing treatment response at week 12 and 52, respectively. Interpretation This novel AI-based pipeline demonstrates strong potential to detect, localise, and quantify neutrophils to assess histological activity in clinical trials and real-world settings. Furthermore, it effectively stratifies treatment response, offering a reliable and objective framework for personalised UC management.

AB - Background Histological remission (HR) is a key treatment goal in ulcerative colitis (UC), suggesting better disease management and treatment response. Absence of mucosal neutrophils is crucial to defining HR. However, the role of neutrophil quantity and localisation remains unclear. We aimed to develop a novel AI-driven pipeline to automate neutrophil detection, localisation, and quantification, supporting assessment of HR and treatment response. Methods We developed an AI-driven pipeline by integrating and combining the outputs of two deep learning models to segment whole slide images (WSIs) into epithelium, crypts and lamina propria and detect and quantify neutrophils. Optimal neutrophil density cut-offs to assess disease activity in UC patients from the AMAC phase two Mirikizumab trial (NCT02589665; 2015-2019) were identified and validated in the multicentre prospective PICaSSO cohort (2016-2019). Cut-offs to determine treatment response at weeks 12 and 52 were also evaluated. Findings 303 WSIs of UC patients from the multicentre, randomised, double-blind, parallel-arm, placebo-controlled phase 2 clinical trial of Mirikizumab were analysed. The models yielded a 65. 44 0% DICE SØrensen for region segmentation on average and 82.3% precision for neutrophil detection. A density cut-off ≥21.7 cells/mm² assessed disease activity with 86% (95% CI: 79%-92%) accuracy, 94% (88%-99%) sensitivity, 74% (54%-89%) specificity, as validated in the PICaSSO cohort. Moreover, a cut-off <21.7 cells/mm² demonstrated 79% (95% CI: 67%-92%) and 85% (95% CI: 70%-96%) accuracy in assessing treatment response at week 12 and 52, respectively. Interpretation This novel AI-based pipeline demonstrates strong potential to detect, localise, and quantify neutrophils to assess histological activity in clinical trials and real-world settings. Furthermore, it effectively stratifies treatment response, offering a reliable and objective framework for personalised UC management.

U2 - 10.1016/j.eclinm.2025.103658

DO - 10.1016/j.eclinm.2025.103658

M3 - Article

SN - 2589-5370

VL - 90

JO - eClinicalMedicine

JF - eClinicalMedicine

M1 - 103658

ER -