Evaluating tubulointerstitial compartments in renal biopsy specimens using a deep learning-based approach for classifying normal and abnormal tubules

A Machine Learning-Driven Virtual Biopsy System For Kidney Transplant Patients

In kidney transplantation, day-zero biopsies are used to assess organ quality and discriminate between donor-inherited lesions and those acquired post-transplantation. However, many centers do not perform such biopsies since they are invasive, costly and may delay the transplant procedure. We aim to generate a non-invasive virtual biopsy system using routinely collected donor parameters. Using 14,032 day-zero kidney biopsies from 17 international centers, we develop a virtual biopsy system. 11 basic donor parameters are used to predict four Banff kidney lesions: arteriosclerosis, arteriolar hyalinosis, interstitial fibrosis and tubular atrophy, and the percentage of renal sclerotic glomeruli. Six machine learning models are aggregated into an ensemble model. The virtual biopsy system shows good performance in the internal and external validation sets. We confirm the generalizability of the system in various scenarios. This system could assist physicians in assessing organ quality, optimizing allograft allocation together with discriminating between donor derived and acquired lesions post-transplantation.

A Spatially Guided Machine-Learning Method to Classify and Quantify Glomerular Patterns of Injury in Histology Images

Introduction The diagnosis of glomerular diseases is primarily based on visual assessment of histologic patterns. Semi-quantitative scoring of active and chronic lesions is often required to assess individual characteristics of the disease. Reproducibility of the visual scoring systems remains debatable, while digital and machine-learning technologies present opportunities to detect, classify and quantify glomerular lesions, also considering their inter- and intraglomerular heterogeneity.

MATERIALS AND METHODS

We performed a cross-validated comparison of three modifications of a convolutional neural network (CNN)-based approach for recognition and intraglomerular quantification of nine main glomerular patterns of injury. Reference values provided by two nephropathologists were used for validation. For each glomerular image, visual attention heatmaps were generated with a probability of class attribution for further intraglomerular quantification. The quality of classifier-produced heatmaps was evaluated by intersection over union metrics (IoU) between predicted and ground truth localization heatmaps.

RESULTS

A proposed spatially guided modification of the CNN classifier achieved the highest glomerular pattern classification accuracies, with area under curve (AUC) values up to 0.981. With regards to heatmap overlap area and intraglomerular pattern quantification, the spatially guided classifier achieved a significantly higher generalized mean IoU value compared to single-multiclass and multiple-binary classifiers.

CONCLUSIONS

We propose a spatially guided CNN classifier that in our experiments reveals the potential to achieve high accuracy for the localization of intraglomerular patterns.

Omni-Seg: A Scale-Aware Dynamic Network for Renal Pathological Image Segmentation

Comprehensive semantic segmentation on renal pathological images is challenging due to the heterogeneous scales of the objects. For example, on a whole slide image (WSI), the cross-sectional areas of glomeruli can be 64 times larger than that of the peritubular capillaries, making it impractical to segment both objects on the same patch, at the same scale. To handle this scaling issue, prior studies have typically trained multiple segmentation networks in order to match the optimal pixel resolution of heterogeneous tissue types. This multi-network solution is resource-intensive and fails to model the spatial relationship between tissue types. In this article, we propose the Omni-Seg network, a scale-aware dynamic neural network that achieves multi-object (six tissue types) and multi-scale (5× to 40× scale) pathological image segmentation via a single neural network. The contribution of this article is three-fold: (1) a novel scale-aware controller is proposed to generalize the dynamic neural network from single-scale to multi-scale; (2) semi-supervised consistency regularization of pseudo-labels is introduced to model the inter-scale correlation of unannotated tissue types into a single end-to-end learning paradigm; and (3) superior scale-aware generalization is evidenced by directly applying a model trained on human kidney images to mouse kidney images, without retraining. By learning from 150,000 human pathological image patches from six tissue types at three different resolutions, our approach achieved superior segmentation performance according to human visual assessment and evaluation of image-omics (i.e., spatial transcriptomics).

Tubular and Glomerular Size by Cortex Depth as Predictor of Progressive CKD after Radical Nephrectomy for Tumor

SIGNIFICANCE STATEMENT

Glomerular size differs by cortex depth. Larger nephrons are prognostic of progressive kidney disease, but it is unknown whether this risk differs by cortex depth or by glomeruli versus proximal or distal tubule size. We studied the average minor axis diameter in oval proximal and distal tubules separately and by cortex depth in patients who had radical nephrectomy to remove a tumor from 2019 to 2020. In adjusted analyses, larger glomerular volume in the middle and deep cortex predicted progressive kidney disease. Wider proximal tubular diameter did not predict progressive kidney disease independent of glomerular volume. Wider distal tubular diameter showed a gradient of strength of prediction of progressive kidney disease in the more superficial cortex than in the deep cortex.

BACKGROUND

Larger nephrons are prognostic of progressive kidney disease, but whether this risk differs by nephron segments or by depth in the cortex is unclear.

METHODS

We studied patients who underwent radical nephrectomy for a tumor between 2000 and 2019. Large wedge kidney sections were scanned into digital images. We estimated the diameters of proximal and distal tubules by the minor axis of oval tubular profiles and estimated glomerular volume with the Weibel-Gomez stereological model. Analyses were performed separately in the superficial, middle, and deep cortex. Cox proportional hazard models assessed the risk of progressive CKD (dialysis, kidney transplantation, sustained eGFR <10 ml/min per 1.73 m 2 , or a sustained 40% decline from the postnephrectomy baseline eGFR) with glomerular volume or tubule diameters. At each cortical depth, models were unadjusted, adjusted for glomerular volume or tubular diameter, and further adjusted for clinical characteristics (age, sex, body mass index, hypertension, diabetes, postnephrectomy baseline eGFR, and proteinuria).

RESULTS

Among 1367 patients were 62 progressive CKD events during a median follow-up of 4.5 years. Glomerular volume predicted CKD outcomes at all depths, but only in the middle and deep cortex after adjusted analyses. Proximal tubular diameter also predicted progressive CKD at any depth but not after adjusted analyses. Distal tubular diameter showed a gradient of more strongly predicting progressive CKD in the superficial than deep cortex, even in adjusted analysis.

CONCLUSIONS

Larger glomeruli are independent predictors of progressive CKD in the deeper cortex, whereas in the superficial cortex, wider distal tubular diameters are an independent predictor of progressive CKD.

Automated Deep Learning-Based Classification of Wilms Tumor Histopathology

(1) Background: Histopathological assessment of Wilms tumors (WT) is crucial for risk group classification to guide postoperative stratification in chemotherapy pre-treated WT cases. However, due to the heterogeneous nature of the tumor, significant interobserver variation between pathologists in WT diagnosis has been observed, potentially leading to misclassification and suboptimal treatment. We investigated whether artificial intelligence (AI) can contribute to accurate and reproducible histopathological assessment of WT through recognition of individual histopathological tumor components. (2) Methods: We assessed the performance of a deep learning-based AI system in quantifying WT components in hematoxylin and eosin-stained slides by calculating the Sørensen-Dice coefficient for fifteen predefined renal tissue components, including six tumor-related components. We trained the AI system using multiclass annotations from 72 whole-slide images of patients diagnosed with WT. (3) Results: The overall Dice coefficient for all fifteen tissue components was 0.85 and for the six tumor-related components was 0.79. Tumor segmentation worked best to reliably identify necrosis (Dice coefficient 0.98) and blastema (Dice coefficient 0.82). (4) Conclusions: Accurate histopathological classification of WT may be feasible using a digital pathology-based AI system in a national cohort of WT patients.

High-throughput image analysis with deep learning captures heterogeneity and spatial relationships after kidney injury

Recovery from acute kidney injury can vary widely in patients and in animal models. Immunofluorescence staining can provide spatial information about heterogeneous injury responses, but often only a fraction of stained tissue is analyzed. Deep learning can expand analysis to larger areas and sample numbers by substituting for time-intensive manual or semi-automated quantification techniques. Here we report one approach to leverage deep learning tools to quantify heterogenous responses to kidney injury that can be deployed without specialized equipment or programming expertise. We first demonstrated that deep learning models generated from small training sets accurately identified a range of stains and structures with performance similar to that of trained human observers. We then showed this approach accurately tracks the evolution of folic acid induced kidney injury in mice and highlights spatially clustered tubules that fail to repair. We then demonstrated that this approach captures the variation in recovery across a robust sample of kidneys after ischemic injury. Finally, we showed markers of failed repair after ischemic injury were correlated both spatially within and between animals and that failed repair was inversely correlated with peritubular capillary density. Combined, we demonstrate the utility and versatility of our approach to capture spatially heterogenous responses to kidney injury.