Glomerulus Detection Research Articles

Deep learning-based glomerulus detection and classification with generative morphology augmentation in renal pathology images

Glomerulus morphology on renal pathology images provides valuable diagnosis and outcome prediction information. To provide better care, an efficient, standardized, and scalable method is urgently needed to optimize the time-consuming and labor-intensive interpretation process by renal pathologists. This paper proposes a deep convolutional neural network (CNN)-based approach to automatically detect and classify glomeruli with different stains in renal pathology images. In the glomerulus detection stage, this paper proposes a flattened Xception with a feature pyramid network (FX-FPN). The FX-FPN is employed as a backbone in the framework of faster region-based CNN to improve glomerulus detection performance. In the classification stage, this paper considers classifications of five glomerulus morphologies using a flattened Xception classifier. To endow the classifier with higher discriminability, this paper proposes a generative data augmentation approach for patch-based glomerulus morphology augmentation. New glomerulus patches of different morphologies are generated for data augmentation through the cycle-consistent generative adversarial network (CycleGAN). The single detection model shows the F1 score up to 0.9524 in H&E and PAS stains. The classification result shows that the average sensitivity and specificity are 0.7077 and 0.9316, respectively, by using the flattened Xception with the original training data. The sensitivity and specificity increase to 0.7623 and 0.9443, respectively, by using the generative data augmentation. Comparisons with different deep CNN models show the effectiveness and superiority of the proposed approach.

Visualization of Renal Glomeruli in Human Native Kidneys With Sensing Ultrasound Localization Microscopy.

Kidney diseases significantly impact individuals' quality of life and strongly reduce life expectancy. Glomeruli play a crucial role in kidney function. Current imaging techniques cannot visualize them due to their small size. Sensing ultrasound localization microscopy (sULM) has shown promising results for visualizing in vivo the glomeruli of human kidney grafts. This study aimed to evaluate the ability of sULM to visualize glomeruli in vivo in native human kidneys despite their depth and a shorter duration of ultrasound acquisition limited by the period of the patient's apnea. Sensing ultrasound localization microscopy parameters in native kidneys and kidney grafts and their consequence regarding glomeruli detection were also compared. Exploration by sULM was conducted in 15 patients with native kidneys and 5 with kidney allografts. Glomeruli were counted using a normalized distance metric projected onto sULM density maps. The difference in the acquisition time, the kidney depth, and the frame rate between native kidneys and kidney grafts and their consequence regarding glomeruli detection were assessed. Glomerular visualization was achieved in 12 of 15 patients with native kidneys. It failed due to impossible breath-holding for 2 patients and a too-deep kidney for 1 patient. Sensing ultrasound localization microscopy found 16 glomeruli per square centimeter in the native kidneys (6-31) and 33 glomeruli per square centimeter in kidney transplant patients (18-55). This study demonstrated that sULM can visualize glomeruli in native human kidneys in vivo. The proposed method may have many hypothetical applications, including biomarker development, assisting biopsy, or potentially avoiding it. It establishes a framework for improving the detection of local microstructural pathology, influencing the evaluation of allografts, and facilitating disease monitoring in the native kidney.

Unsupervised Multi-Domain Progressive Stain Transfer Guided by Style Encoding Dictionary.

In histopathology, the tissue slides are usually stained by common H&E stain or special stains (MAS, PAS, and PASM, etc.) to clearly show specific tissue structures. The rapid development of deep learning provides a good solution to generate virtual staining images to significantly reduce the time and labor costs associated with histochemical staining. However, most existing methods need to train a special model for every two stains, which consumes a lot of computing resources with the increasing of staining types. To address this problem, we propose an unsupervised multi-domain stain transfer method, GramGAN, which realizes the progressive transfer through cascaded Style-Guided blocks. For each Style-Guided block, we design a style encoding dictionary to characterize and store all the staining style information. In addition, we propose a Rényi entropy-based regularization term to improve the discrimination ability of different styles. The experimental results show that our method can realize accurate transferring among multiple staining styles with better performance. Furthermore, we build and publish a special stained image dataset suitable for glomeruli segmentation (including H&E staining), where the accuracy of glomeruli detection and segmentation can be significantly improved after transferring H&E-stained images to PAS-stained and PASM-stained ones by our method. The code is publicly available at: https://github.com/xianchaoguan/GramGAN.

Development of a multiple convolutional neural network-facilitated diagnostic screening program for immunofluorescence images of IgA nephropathy and idiopathic membranous nephropathy.

Immunoglobulin A nephropathy (IgAN) and idiopathic membranous nephropathy (IMN) are the most common glomerular diseases. Immunofluorescence (IF) tests of renal tissues are crucial for the diagnosis. We developed a multiple convolutional neural network (CNN)-facilitated diagnostic program to assist the IF diagnosis of IgAN and IMN. The diagnostic program consisted of four parts: a CNN trained as a glomeruli detection module, an IF intensity comparator, dual-CNN (D-CNN) trained as a deposition appearance and location classifier and a post-processing module. A total of 1573 glomerular IF images from 1009 patients with glomerular diseases were used for the training and validation of the diagnostic program. A total of 1610 images of 426 patients from different hospitals were used as test datasets. The performance of the diagnostic program was compared with nephropathologists. In >90% of the tested images, the glomerulus location module achieved an intersection over union >0.8. The accuracy of the D-CNN in recognizing irregular granular mesangial deposition and fine granular deposition along the glomerular basement membrane was 96.1% and 93.3%, respectively. As for the diagnostic program, the accuracy, sensitivity and specificity of diagnosing suspected IgAN were 97.6%, 94.4% and 96.0%, respectively. The accuracy, sensitivity and specificity of diagnosing suspected IMN were 91.7%, 88.9% and 95.8%, respectively. The corresponding areas under the curve (AUCs) were 0.983 and 0.935. When tested with images from the outside hospital, the diagnostic program showed stable performance. The AUCs for diagnosing suspected IgAN and IMN were 0.972 and 0.948, respectively. Compared with inexperienced nephropathologists, the program showed better performance. The proposed diagnostic program could assist the IF diagnosis of IgAN and IMN.

Artificial intelligence can accurately distinguish IgA nephropathy from diabetic nephropathy under Masson staining and becomes an important assistant for renal pathologists.

Hyperplasia of the mesangial area is common in IgA nephropathy (IgAN) and diabetic nephropathy (DN), and it is often difficult to distinguish them by light microscopy alone, especially in the absence of clinical data. At present, artificial intelligence (AI) is widely used in pathological diagnosis, but mainly in tumor pathology. The application of AI in renal pathological is still in its infancy. Patients diagnosed as IgAN or DN by renal biopsy in First Affiliated Hospital of Zhejiang Chinese Medicine University from September 1, 2020 to April 30, 2022 were selected as the training set, and patients who diagnosed from May 1, 2022 to June 30, 2022 were selected as the test set. We focused on the glomerulus and captured the field of the glomerulus in Masson staining WSI at 200x magnification, all in 1,000 × 1,000 pixels JPEG format. We augmented the data from training set through minor affine transformation, and then randomly split the training set into training and adjustment data according to 8:2. The training data and the Yolov5 6.1 algorithm were used to train the AI model with constant adjustment of parameters according to the adjusted data. Finally, we obtained the optimal model, tested this model with test set and compared it with renal pathologists. AI can accurately detect the glomeruli. The overall accuracy of AI glomerulus detection was 98.67% and the omission rate was only 1.30%. No Intact glomerulus was missed. The overall accuracy of AI reached 73.24%, among which the accuracy of IgAN reached 77.27% and DN reached 69.59%. The AUC of IgAN was 0.733 and that of DN was 0.627. In addition, compared with renal pathologists, AI can distinguish IgAN from DN more quickly and accurately, and has higher consistency. We constructed an AI model based on Masson staining images of renal tissue to distinguish IgAN from DN. This model has also been successfully deployed in the work of renal pathologists to assist them in their daily diagnosis and teaching work.

Glomerulus Detection Using Segmentation Neural Networks.

Digital pathology is vital for the correct diagnosis of kidney before transplantation or kidney disease identification. One of the key challenges in kidney diagnosis is glomerulus detection in kidney tissue segments. In this study, we propose a deep learning-based method for glomerulus detection from digitized kidney slide segments. The proposed method applies models based on convolutional neural networks to detect image segments containing the glomerulus region. We employ various networks such as ResNets, UNet, LinkNet, and EfficientNet to train the models. In our experiments on a network trained on the NIH HuBMAP kidney whole slide image dataset, the proposed method achieves the highest scores with Dice coefficient of 0.942.

Glo-In-One: holistic glomerular detection, segmentation, and lesion characterization with large-scale web image mining.

Purpose: The quantitative detection, segmentation, and characterization of glomeruli from high-resolution whole slide imaging (WSI) play essential roles in the computer-assisted diagnosis and scientific research in digital renal pathology. Historically, such comprehensive quantification requires extensive programming skills to be able to handle heterogeneous and customized computational tools. To bridge the gap of performing glomerular quantification for non-technical users, we develop the Glo-In-One toolkit to achieve holistic glomerular detection, segmentation, and characterization via a single line of command. Additionally, we release a large-scale collection of 30,000 unlabeled glomerular images to further facilitate the algorithmic development of self-supervised deep learning. Approach: The inputs of the Glo-In-One toolkit are WSIs, while the outputs are (1)WSI-level multi-class circle glomerular detection results (which can be directly manipulated with ImageScope), (2)glomerular image patches with segmentation masks, and (3)different lesion types. In the current version, the fine-grained global glomerulosclerosis (GGS) characterization is provided, including assessed-solidified-GSS (associated with hypertension-related injury), disappearing-GSS (a further end result of the SGGS becoming contiguous with fibrotic interstitium), and obsolescent-GSS (nonspecific GGS increasing with aging) glomeruli. To leverage the performance of the Glo-In-One toolkit, we introduce self-supervised deep learning to glomerular quantification via large-scale web image mining. Results: The GGS fine-grained classification model achieved a decent performance compared with baseline supervised methods while only using 10% of the annotated data. The glomerular detection achieved an average precision of 0.627 with circle representations, while the glomerular segmentation achieved a 0.955 patch-wise Dice dimilarity coefficient. Conclusion: We develop and release an open-source Glo-In-One toolkit, a software with holistic glomerular detection, segmentation, and lesion characterization. This toolkit is user-friendly to non-technical users via a single line of command. The toolbox and the 30,000 web mined glomerular images have been made publicly available at https://github.com/hrlblab/Glo-In-One.

POS-020 ARTIFICIAL INTELLIGENCE ASSISTED QUANTIFICATION OF COMPLEMENT CONVERTASES IN GLOMERULONEPHRITIS

POS-020 ARTIFICIAL INTELLIGENCE ASSISTED QUANTIFICATION OF COMPLEMENT CONVERTASES IN GLOMERULONEPHRITIS

MO055: Multi-Class Segmentation of Kidney Tissues using Convolutional Neuronal Networks (CNNS)

Abstract BACKGROUND AND AIMS Routine pathological diagnostics in kidneys are mainly based on semi-quantitative eyeballing. In own former studies, we showed predictive value of precise immune cell quantification in allografts using digital semi-automated techniques. We now aim to achieve fully automated segmentation workflow with CNNs. METHOD Standard routine stains (immuno/histochemistry, immunofluorescence) were digitized (20×) with Metafer, a commercial scanning/imaging platform. Diagnostically relevant anatomical compartments (cortex, medulla, glomeruli, tubuli [proximal/distal/collecting duct], glomerular/peritubular capillaries and nuclei) were manually annotated by use of immunomarkers to generate large data sets on human renal biopsies and nephrectomies. Data were used to train multi-class semantic segmentation CNNs with broad data augmentation to achieve a robustness against staining variances. RESULTS Using Jones-HE stains for multi-class segmentation, a cortex-medulla-extrarenal CNN revealed pixel based hit rates above 97.9%, detection of glomeruli had a pixel based hit rate above 99%, a multi-class CNN for tubules, tubular membranes and peritubular capillaries resulted in a hit rate of 91.5%, and nuclear-based cell detection shows pixel based hit rates above 98%. Identification of cell location in interstitium, tubuli, glomeruli, peritubular and glomerular capillaries reached very high hit rates: Glomerular endothelial cells actually result in 83% true positives, 13% false negatives and 4% false positives. Additionally, a tubulus classifier (proximal tubulus, distal tubulus, collecting duct and atrophic tubulus) with an accuracy >90% was developed. CONCLUSION Automated structure segmentation by CNNs can complement and specify classical nephropathological diagnostics, especially for spatial risk marker evaluation in early transplant biopsies.

Dense Multi-Object 3D Glomerular Reconstruction and Quantification on 2D Serial Section Whole Slide Images.

There has been a long pursuit for precise and reproducible glomerular quantification in the field of renal pathology in both research and clinical practice. Currently, 3D glomerular identification and reconstruction of large-scale glomeruli are labor-intensive tasks, and time-consuming by manual analysis on whole slide imaging (WSI) in 2D serial sectioning representation. The accuracy of serial section analysis is also limited in the 2D serial context. Moreover, there are no approaches to present 3D glomerular visualization for human examination (volume calculation, 3D phenotype analysis, etc.). In this paper, we introduce an end-to-end holistic deep-learning-based method that achieves automatic detection, segmentation and multi-object tracking (MOT) of individual glomeruli with large-scale glomerular-registered assessment in a 3D context on WSIs. The high-resolution WSIs are the inputs, while the outputs are the 3D glomerular reconstruction and volume estimation. This pipeline achieves 81.8 in IDF1 and 69.1 in MOTA as MOT performance, while the proposed volume estimation achieves 0.84 Spearman correlation coefficient with manual annotation. The end-to-end MAP3D+ pipeline provides an approach for extensive 3D glomerular reconstruction and volume quantification from 2D serial section WSIs.

Holistic fine-grained global glomerulosclerosis characterization: from detection to unbalanced classification.

Purpose: Recent studies have demonstrated the diagnostic and prognostic values of global glomerulosclerosis (GGS) in IgA nephropathy, aging, and end-stage renal disease. However, the fine-grained quantitative analysis of multiple GGS subtypes (e.g., obsolescent, solidified, and disappearing glomerulosclerosis) is typically a resource extensive manual process. Very few automatic methods, if any, have been developed to bridge this gap for such analytics. We present a holistic pipeline to quantify GGS (with both detection and classification) from a whole slide image in a fully automatic manner. In addition, we conduct the fine-grained classification for the subtypes of GGS. Our study releases the open-source quantitative analytical tool for fine-grained GGS characterization while tackling the technical challenges in unbalanced classification and integrating detection and classification. Approach: We present a deep learning-based framework to perform fine-grained detection and classification of GGS, with a hierarchical two-stage design. Moreover, we incorporate the state-of-the-art transfer learning techniques to achieve a more generalizable deep learning model for tackling the imbalanced distribution of our dataset. This way, we build a highly efficient WSI-to-results GGS characterization pipeline. Meanwhile, we investigated the largest fine-grained GGS cohort as of yet with 11,462 glomeruli and 10,619 nonglomeruli, which include 7841 globally sclerotic glomeruli of three distinct categories. With these data, we apply deep learning techniques to achieve (1) fine-grained GGS characterization, (2) GGS versus non-GGS classification, and (3) improved glomeruli detection results. Results: For fine-grained GGS characterization, when pretrained on the larger dataset, our model can achieve a 0.778-macro- score, compared to a 0.746-macro- score when using the regular ImageNet-pretrained weights. On the external dataset, our best model achieves an area under the curve (AUC) score of 0.994 when tasked with differentiating GGS from normal glomeruli. Using our dataset, we are able to build algorithms that allow for fine-grained classification of glomeruli lesions and are robust to distribution shifts. Conclusion: Our study showed that the proposed methods consistently improve the detection and fine-grained classification performance through both cross validation and external validation. Our code and pretrained models have been released for public use at https://github.com/luyuzhe111/glomeruli.

Glomerular disease classification and lesion identification by machine learning

BackgroundClassification of glomerular diseases and identification of glomerular lesions require careful morphological examination by experienced nephropathologists, which is labor-intensive, time-consuming, and prone to interobserver variability. In this regard, recent advance in machine learning-based image analysis is promising. MethodsWe combined Mask Region-based Convolutional Neural Networks (Mask R–CNN) with an additional classification step to build a glomerulus detection model using human kidney biopsy samples. A Long Short-Term Memory (LSTM) recurrent neural network was applied for glomerular disease classification, and another two-stage model using ResNeXt-101 was constructed for glomerular lesion identification in cases of lupus nephritis. ResultsThe detection model showed state-of-the-art performance on variedly stained slides with F1 scores up to 0.944. The disease classification model showed good accuracies up to 0.940 on recognizing different glomerular diseases based on H&E whole slide images. The lesion identification model demonstrated high discriminating power with area under the receiver operating characteristic curve up to 0.947 for various glomerular lesions. Models showed good generalization on external testing datasets. ConclusionThis study is the first-of-its-kind showing how each step of kidney biopsy interpretation carried out by nephropathologists can be captured and simulated by machine learning models. The models were integrated into a whole slide image viewing and annotating platform to enable nephropathologists to review, correct, and confirm the inference results. Further improvement on model performances and incorporating inputs from immunofluorescence, electron microscopy, and clinical data might realize actual clinical use.

Automated identification of glomeruli and synchronised review of special stains in renal biopsies by machine learning and slide registration: a cross-institutional study.

Machine learning in digital pathology can improve efficiency and accuracy via prescreening with automated feature identification. Studies using uniform histological material have shown promise. Generalised application requires validation on slides from multiple institutions. We used machine learning to identify glomeruli on renal biopsies and compared performance between single and multiple institutions. Randomly selected, adequately sampled renal core biopsy cases (71) consisting of four stains each (haematoxylin and eosin, trichrome, silver, periodic acid Schiff)from three institutions were digitisedat ×40. Glomeruli were manually annotated by three renal pathologists using a digitaltool. Cases were divided into training/validation (n=52) and evaluation (n=19) cohorts. An algorithm was trained to develop three convolutional neural network (CNN) models which tested case cohorts intra- and inter-institutionally. Raw CNN search data from each of the four slides per case were merged into composite regions of interest containing putative glomeruli. The sensitivity and modified specificity of glomerulus detection (versus annotated truth) were calculated for each model/cohort. Intra-institutional (3) sensitivity ranged from 90 to 93%, with modified specificity from 86 to 98%. Interinstitutional (1) sensitivity was 77%, with modified specificity 97%. Combined intra- and inter-institutional (1) sensitivity was 86%, with modified specificity 92%. Feature detection sensitivity degrades when training and test material originate from different sites. Training using a combined set of digital slides from three institutions improves performance. Differing histology methods probably account for algorithm performance contrasts. Our data highlight the need for diverse training sets for the development of generalisable machine learning histology algorithms.

Recognition and Classification of Glomerular Pathological Images Based on Deep Learning

<p indent=0mm>The identification and classification of glomeruli in pathological sections is the key to diagnosing the degree and type of renal lesions. In order to solve the problem of glomerular recognition and classification, a complete glomerular detection and classification framework based on deep learning is designed. Glomeruli are detected and classified in the entire slice image. The framework includes four stages of glomerular recognition. In the first stage of scanning window generation, a new network framework, RGNet, is designed to initially determine the possible location of glomeruli. In the second stage of detection and coarse classification, Faster R-CNN is improved for glomerular data. In the third stage, the NMS-Lite algorithm is designed based on the NMS algorithm to merge the detected glomeruli. In the fourth stage of fine classification, two neural networks are trained using data augmentation to classify the degree of glomerular lesions. The experimental results has show that the glomerulus detection method proposed in this paper has achieved comparable accuracy on the test set with similar methods, and to a certain extent solves the problem that similar types of glomeruli are difficult to distinguish.

Image Analysis Pipeline for Renal Allograft Evaluation and Fibrosis Quantification

IntroductionDigital pathology improves the standardization and reproducibility of kidney biopsy specimen assessment. We developed a pipeline allowing the analysis of many images without requiring human preprocessing and illustrate its use with a simple algorithm for quantification of interstitial fibrosis on a large dataset of kidney allograft biopsy specimens.MethodsMasson trichrome–stained images from kidney allograft biopsy specimens were used to train and validate a glomeruli detection algorithm using a VGG19 convolutional neural network and an automatic cortical region of interest (ROI) selection algorithm including cortical regions containing all predicted glomeruli. A positive-pixel count algorithm was used to quantify interstitial fibrosis on the ROIs and the association between automatic fibrosis and pathologist evaluation, estimated glomerular filtration rate (GFR) and allograft survival was assessed.ResultsThe glomeruli detection (F1 score of 0.87) and ROIs selection (F1 score 0.83 [SD 0.13]) algorithms displayed high accuracy. The correlation between the automatic fibrosis quantification on manually and automatically selected ROIs was high (r = 1.00 [0.99–1.00]). Automatic fibrosis quantification was only moderately correlated with pathologists’ assessment and was not significantly associated with eGFR or allograft survival.ConclusionThis pipeline can automatically and accurately detect glomeruli and select cortical ROIs that can easily be used to develop, validate, and apply image analysis algorithms.

Cycle-consistent GAN-based stain translation of renal pathology images with glomerulus detection application

Motivation: Renal biopsy is an irreplaceable diagnostic tool for kidney diseases. Glomeruli provide important information for an accurate disease diagnosis. This paper applies deep learning techniques to automate translation of renal pathology images and glomerulus detection to improve the efficiency and accuracy on pathological diagnoses. Methods: This paper first proposes a new method for automatic translation of different renal pathology staining styles using the cycle-consistent Generative Adversarial Network (GAN). This paper then proposes the combination of faster region-based convolutional neural network (R-CNN) with an aspect ratio filter to detect glomeruli in light microscopy images processed with four different stains at various optical magnifications. Finally, this paper improves glomerulus detection at different stains by using translated image stains from the CycleGAN. Results: To show the effectiveness of the translation and detection methods, in addition to quantitative analysis of the results, the involvement of assessment from four physicians is also performed. Experimental results show that the physicians fail to differentiate real and translated stains and the automatic glomerulus detection method outperforms that manually labeled by the physicians. Conclusion: The proposed method works well and improves the efficiency of renal pathological diagnosis. This work contributes in the area of automated medical diagnosis.

Orbit Image Analysis: An open-source whole slide image analysis tool.

We describe Orbit Image Analysis, an open-source whole slide image analysis tool. The tool consists of a generic tile-processing engine which allows the execution of various image analysis algorithms provided by either Orbit itself or from other open-source platforms using a tile-based map-reduce execution framework. Orbit Image Analysis is capable of sophisticated whole slide imaging analyses due to several key features. First, Orbit has machine-learning capabilities. This deep learning segmentation can be integrated with complex object detection for analysis of intricate tissues. In addition, Orbit can run locally as standalone or connect to the open-source image server OMERO. Another important characteristic is its scale-out functionality, using the Apache Spark framework for distributed computing. In this paper, we describe the use of Orbit in three different real-world applications: quantification of idiopathic lung fibrosis, nerve fibre density quantification, and glomeruli detection in the kidney.

Improved small blob detection in 3D images using jointly constrained deep learning and Hessian analysis

Imaging biomarkers are being rapidly developed for early diagnosis and staging of disease. The development of these biomarkers requires advances in both image acquisition and analysis. Detecting and segmenting objects from images are often the first steps in quantitative measurement of these biomarkers. The challenges of detecting objects in images, particularly small objects known as blobs, include low image resolution, image noise and overlap between the blobs. The Difference of Gaussian (DoG) detector has been used to overcome these challenges in blob detection. However, the DoG detector is susceptible to over-detection and must be refined for robust, reproducible detection in a wide range of medical images. In this research, we propose a joint constraint blob detector from U-Net, a deep learning model, and Hessian analysis, to overcome these problems and identify true blobs from noisy medical images. We evaluate this approach, UH-DoG, using a public 2D fluorescent dataset for cell nucleus detection and a 3D kidney magnetic resonance imaging dataset for glomerulus detection. We then compare this approach to methods in the literature. While comparable to the other four comparing methods on recall, the UH-DoG outperforms them on both precision and F-score.

Using Image Registration and Machine Learning to Develop a Workstation Tool for Rapid Analysis of Glomeruli in Medical Renal Biopsies

Background: Prescreening of biopsies has the potential to improve pathologists’ workflow. Tools that identify features and display results in a visually thoughtful manner can enhance efficiency, accuracy, and reproducibility. Machine learning for detection of glomeruli ensures comprehensive assessment and registration of four different stains allows for simultaneous navigation and viewing. Methods: Medical renal core biopsies (4 stains each) were digitized using a Leica SCN400 at *40 and loaded into the Corista Quantum research platform. Glomeruli were manually annotated by pathologists. The tissue on the 4 stains was registered using a combination of keypoint- and intensity-based algorithms, and a 4-panel simultaneous viewing display was created. Using a training cohort, machine learning convolutional neural net (CNN) models were created to identify glomeruli in all stains, and merged into composite fields of views (FOVs). The sensitivity and specificity of glomerulus detection, and FOV area for each detection were calculated. Results: Forty-one biopsies were used for training (28) and same-batch evaluation (6). Seven additional biopsies from a temporally different batch were also evaluated. A variant of AlexNet CNN, used for object recognition, showed the best result for the detection of glomeruli with same-batch and different-batch evaluation: Same-batch sensitivity 92%, “modified” specificity 89%, average FOV size represented 0.8% of the total slide area; different-batch sensitivity 90%, “modified” specificity 98% and average FOV size 1.6% of the total slide area. Conclusions: Glomerulus detection in the best CNN model shows that machine learning algorithms may be accurate for this task. The added benefit of biopsy registration with simultaneous display and navigation allows reviewers to move from one machine-generated FOV to the next in all 4 stains. Together these features could increase both efficiency and accuracy in the review process.

CircleNet: Anchor-free Glomerulus Detection with Circle Representation.

Object detection networks are powerful in computer vision, but not necessarily optimized for biomedical object detection. In this work, we propose CircleNet, a simple anchor-free detection method with circle representation for detection of the ball-shaped glomerulus. Different from the traditional bounding box based detection method, the bounding circle (1) reduces the degrees of freedom of detection representation, (2) is naturally rotation invariant, (3) and optimized for ball-shaped objects. The key innovation to enable this representation is the anchor-free framework with the circle detection head. We evaluate CircleNet in the context of detection of glomerulus. CircleNet increases average precision of the glomerulus detection from 0.598 to 0.647. Another key advantage is that CircleNet achieves better rotation consistency compared with bounding box representations.