Recognition of intraglomerular histological features with deep learning in protocol transplant biopsies and their association with kidney function and prognosis

Imane Farhat; Elise Maréchal; Doris Calmo; Manon Ansart; Michel Paindavoine; Patrick Bard; Georges Tarris; Didier Ducloux; Sophie Adrian Felix; Laurent Martin; Claire Tinel; Jean-Baptiste Gibier; Mathilde Funes de la Vega; Jean-Michel Rebibou; Jamal Bamoulid; Mathieu Legendre

doi:10.1093/ckj/sfae019

Recognition of intraglomerular histological features with deep learning in protocol transplant biopsies and their association with kidney function and prognosis

Clin Kidney J. 2024 Feb 2;17(2):sfae019. doi: 10.1093/ckj/sfae019. eCollection 2024 Feb.

Authors

Imane Farhat¹, Elise Maréchal¹, Doris Calmo², Manon Ansart³, Michel Paindavoine³, Patrick Bard³, Georges Tarris⁴, Didier Ducloux^{2

5}, Sophie Adrian Felix⁶, Laurent Martin⁴, Claire Tinel^{1

5}, Jean-Baptiste Gibier⁷, Mathilde Funes de la Vega⁴, Jean-Michel Rebibou^{1

5}, Jamal Bamoulid^{2

5}, Mathieu Legendre^{1

3

5}

Affiliations

¹ Department of Nephrology, CHU Dijon, Dijon, France.
² Department of Nephrology, CHU Besançon, Besançon, France.
³ LEAD-CNRS, UMR 5022, Université de Bourgogne, Dijon, France.
⁴ Department of Pathology, CHU Dijon, Dijon, France.
⁵ Etablissement Français du sang, Besançon, France.
⁶ Department of Pathology, CHU Besançon, Besançon 25000, France.
⁷ Department of Pathology, CHU Lille, Lille, France.

Abstract

Background: The Banff Classification may not adequately address protocol transplant biopsies categorized as normal in patients experiencing unexplained graft function deterioration. This study seeks to employ convolutional neural networks to automate the segmentation of glomerular cells and capillaries and assess their correlation with transplant function.

Methods: A total of 215 patients were categorized into three groups. In the Training cohort, glomerular cells and capillaries from 37 patients were manually annotated to train the networks. The Test cohort (24 patients) compared manual annotations vs automated predictions, while the Application cohort (154 protocol transplant biopsies) examined predicted factors in relation to kidney function and prognosis.

Results: In the Test cohort, the networks recognized histological structures with Precision, Recall, F-score and Intersection Over Union exceeding 0.92, 0.85, 0.89 and 0.74, respectively. Univariate analysis revealed associations between the estimated glomerular filtration rate (eGFR) at biopsy and relative endothelial area (r = 0.19, P = .027), endothelial cell density (r = 0.20, P = .017), mean parietal epithelial cell area (r = -0.38, P < .001), parietal epithelial cell density (r = 0.29, P < .001) and mesangial cell density (r = 0.22, P = .010). Multivariate analysis retained only endothelial cell density as associated with eGFR (Beta = 0.13, P = .040). Endothelial cell density (r = -0.22, P = .010) and mean podocyte area (r = 0.21, P = .016) were linked to proteinuria at biopsy. Over 44 ± 29 months, 25 patients (16%) reached the primary composite endpoint (dialysis initiation, or 30% eGFR sustained decline), with relative endothelial area, mean endothelial cell area and parietal epithelial cell density below medians linked to this endpoint [hazard ratios, respectively, of 2.63 (P = .048), 2.60 (P = .039) and 3.23 (P = .019)].

Conclusion: This study automated the measurement of intraglomerular cells and capillaries. Our results suggest that the precise segmentation of endothelial and epithelial cells may serve as a potential future marker for the risk of graft loss.

Keywords: artificial intelligence; deep learning; glomerulus; kidney pathology; transplant biopsies.