Development and validation of an AI-enabled digital breast cancer assay to predict early-stage breast cancer recurrence within 6 years.

BACKGROUND: Breast cancer (BC) grading plays a critical role in patient management despite the considerable inter- and intra-observer variability, highlighting the need for decision support tools to improve reproducibility and prognostic accuracy for use in clinical practice. The objective was to evaluate the ability of a digital artificial intelligence (AI) assay (PDxBr) to enrich BC grading and improve risk categorization for predicting recurrence. METHODS: In our population-based longitudinal clinical development and validation study, we enrolled 2075 patients from Mount Sinai Hospital with infiltrating ductal carcinoma of the breast. With 3:1 balanced training and validation cohorts, patients were retrospectively followed for a median of 6 years. The main outcome was to validate an automated BC phenotyping system combined with clinical features to produce a binomial risk score predicting BC recurrence at diagnosis. RESULTS: The PDxBr training model (n = 1559 patients) had a C-index of 0.78 (95% CI, 0.76-0.81) versus clinical 0.71 (95% CI, 0.67-0.74) and image feature models 0.72 (95% CI, 0.70-0.74). A risk score of 58 (scale 0-100) stratified patients as low or high risk, hazard ratio (HR) 5.5 (95% CI 4.19-7.2, p < 0.001), with a sensitivity 0.71, specificity 0.77, NPV 0.95, and PPV 0.32 for predicting BC recurrence within 6 years. In the validation cohort (n = 516), the C-index was 0.75 (95% CI, 0.72-0.79) versus clinical 0.71 (95% CI 0.66-0.75) versus image feature models 0.67 (95% CI, 0.63-071). The validation cohort had an HR of 4.4 (95% CI 2.7-7.1, p < 0.001), sensitivity of 0.60, specificity 0.77, NPV 0.94, and PPV 0.24 for predicting BC recurrence within 6 years. PDxBr also improved Oncotype Recurrence Score (RS) performance: RS 31 cutoff, C-index of 0.36 (95% CI 0.26-0.45), sensitivity 37%, specificity 48%, HR 0.48, p = 0.04 versus Oncotype RS plus AI-grade C-index 0.72 (95% CI 0.67-0.79), sensitivity 78%, specificity 49%, HR 4.6, p < 0.001 versus Oncotype RS plus PDxBr, C-index 0.76 (95% CI 0.70-0.82), sensitivity 67%, specificity 80%, HR 6.1, p < 0.001. CONCLUSIONS: PDxBr is a digital BC test combining automated AI-BC prognostic grade with clinical-pathologic features to predict the risk of early-stage BC recurrence. With future validation studies, we anticipate the PDxBr model will enrich current gene expression assays and enhance treatment decision-making.

Artificial intelligence in neuropathology: deep learning-based assessment of tauopathy.

Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and overlapping morphological phenotypes that obscure classification and quantitative assessments. Recently, powerful machine learning-based approaches have emerged, allowing the recognition and quantification of pathological changes from digital images. Here, we applied deep learning to the neuropathological assessment of NFT in postmortem human brain tissue to develop a classifier capable of recognizing and quantifying tau burden. The histopathological material was derived from 22 autopsy brains from patients with tauopathies. We used a custom web-based informatics platform integrated with an in-house information management system to manage whole slide images (WSI) and human expert annotations as ground truth. We utilized fully annotated regions to train a deep learning fully convolutional neural network (FCN) implemented in PyTorch against the human expert annotations. We found that the deep learning framework is capable of identifying and quantifying NFT with a range of staining intensities and diverse morphologies. With our FCN model, we achieved high precision and recall in naive WSI semantic segmentation, correctly identifying tangle objects using a SegNet model trained for 200 epochs. Our FCN is efficient and well suited for the practical application of WSIs with average processing times of 45 min per WSI per GPU, enabling reliable and reproducible large-scale detection of tangles. We measured performance on test data of 50 pre-annotated regions on eight naive WSI across various tauopathies, resulting in the recall, precision, and an F1 score of 0.92, 0.72, and 0.81, respectively. Machine learning is a useful tool for complex pathological assessment of AD and other tauopathies. Using deep learning classifiers, we have the potential to integrate cell- and region-specific annotations with clinical, genetic, and molecular data, providing unbiased data for clinicopathological correlations that will enhance our knowledge of the neurodegeneration.

Analytical Validation of the PreciseDx Digital Prognostic Breast Cancer Test in Early-Stage Breast Cancer.

BACKGROUND: PreciseDx Breast (PDxBr) is a digital test that predicts early-stage breast cancer recurrence within 6-years of diagnosis. MATERIALS AND METHODS: Using hematoxylin and eosin-stained whole slide images of invasive breast cancer (IBC) and artificial intelligence-enabled morphology feature array, microanatomic features are generated. Morphometric attributes in combination with patient's age, tumor size, stage, and lymph node status predict disease free survival using a proprietary algorithm. Here, analytical validation of the automated annotation process and extracted histologic digital features of the PDxBr test, including impact of methodologic variability on the composite risk score is presented. Studies of precision, repeatability, reproducibility and interference were performed on morphology feature array-derived features. The final risk score was assessed over 20-days with 2-operators, 2-runs/day, and 2-replicates across 8-patients, allowing for calculation of within-run repeatability, between-run and within-laboratory reproducibility. RESULTS: Analytical validation of features derived from whole slide images demonstrated a high degree of precision for tumor segmentation (0.98, 0.98), lymphocyte detection (0.91, 0.93), and mitotic figures (0.85, 0.84). Correlation of variation of the assay risk score for both reproducibility and repeatability were less than 2%, and interference from variation in hematoxylin and eosin staining or tumor thickness was not observed demonstrating assay robustness across standard histopathology preparations. CONCLUSION: In summary, the analytical validation of the digital IBC risk assessment test demonstrated a strong performance across all features in the model and complimented the clinical validation of the assay previously shown to accurately predict recurrence within 6-years in early-stage invasive breast cancer patients.

Artificial intelligence-derived neurofibrillary tangle burden is associated with antemortem cognitive impairment.

Tauopathies are a category of neurodegenerative diseases characterized by the presence of abnormal tau protein-containing neurofibrillary tangles (NFTs). NFTs are universally observed in aging, occurring with or without the concomitant accumulation of amyloid-beta peptide (Aß) in plaques that typifies Alzheimer disease (AD), the most common tauopathy. Primary age-related tauopathy (PART) is an Aß-independent process that affects the medial temporal lobe in both cognitively normal and impaired subjects. Determinants of symptomology in subjects with PART are poorly understood and require clinicopathologic correlation; however, classical approaches to staging tau pathology have limited quantitative reproducibility. As such, there is a critical need for unbiased methods to quantitatively analyze tau pathology on the histological level. Artificial intelligence (AI)-based convolutional neural networks (CNNs) generate highly accurate and precise computer vision assessments of digitized pathology slides, yielding novel histology metrics at scale. Here, we performed a retrospective autopsy study of a large cohort (n = 706) of human post-mortem brain tissues from normal and cognitively impaired elderly individuals with mild or no Aß plaques (average age of death of 83.1 yr, range 55-110). We utilized a CNN trained to segment NFTs on hippocampus sections immunohistochemically stained with antisera recognizing abnormal hyperphosphorylated tau (p-tau), which yielded metrics of regional NFT counts, NFT positive pixel density, as well as a novel graph-theory based metric measuring the spatial distribution of NFTs. We found that several AI-derived NFT metrics significantly predicted the presence of cognitive impairment in both the hippocampus proper and entorhinal cortex (p < 0.0001). When controlling for age, AI-derived NFT counts still significantly predicted the presence of cognitive impairment (p = 0.04 in the entorhinal cortex; p = 0.04 overall). In contrast, Braak stage did not predict cognitive impairment in either age-adjusted or unadjusted models. These findings support the hypothesis that NFT burden correlates with cognitive impairment in PART. Furthermore, our analysis strongly suggests that AI-derived metrics of tau pathology provide a powerful tool that can deepen our understanding of the role of neurofibrillary degeneration in cognitive impairment.

Antemortem detection of Parkinson's disease pathology in peripheral biopsies using artificial intelligence.

The diagnosis of Parkinson's disease (PD) is challenging at all stages due to variable symptomatology, comorbidities, and mimicking conditions. Postmortem assessment remains the gold standard for a definitive diagnosis. While it is well recognized that PD manifests pathologically in the central nervous system with aggregation of α-synuclein as Lewy bodies and neurites, similar Lewy-type synucleinopathy (LTS) is additionally found in the peripheral nervous system that may be useful as an antemortem biomarker. We have previously found that detection of LTS in submandibular gland (SMG) biopsies is sensitive and specific for advanced PD; however, the sensitivity is suboptimal especially for early-stage disease. Further, visual microscopic assessment of biopsies by a neuropathologist to identify LTS is impractical for large-scale adoption. Here, we trained and validated a convolutional neural network (CNN) for detection of LTS on 283 digital whole slide images (WSI) from 95 unique SMG biopsies. A total of 8,450 LTS and 35,066 background objects were annotated following an inter-rater reliability study with Fleiss Kappa = 0.72. We used transfer learning to train a CNN model to classify image patches (151 × 151 pixels at 20× magnification) with and without the presence of LTS objects. The trained CNN model showed the following performance on image patches: sensitivity: 0.99, specificity: 0.99, precision: 0.81, accuracy: 0.99, and F-1 score: 0.89. We further tested the trained network on 1230 naïve WSI from the same cohort of research subjects comprising 42 PD patients and 14 controls. Logistic regression models trained on features engineered from the CNN predictions on the WSI resulted in sensitivity: 0.71, specificity: 0.65, precision: 0.86, accuracy: 0.69, and F-1 score: 0.76 in predicting clinical PD status, and 0.64 accuracy in predicting PD stage, outperforming expert neuropathologist LTS density scoring in terms of sensitivity but not specificity. These findings demonstrate the practical utility of a CNN detector in screening for LTS, which can translate into a computational tool to facilitate the antemortem tissue-based diagnosis of PD in clinical settings.

PAIP 2019: Liver cancer segmentation challenge.

Pathology Artificial Intelligence Platform (PAIP) is a free research platform in support of pathological artificial intelligence (AI). The main goal of the platform is to construct a high-quality pathology learning data set that will allow greater accessibility. The PAIP Liver Cancer Segmentation Challenge, organized in conjunction with the Medical Image Computing and Computer Assisted Intervention Society (MICCAI 2019), is the first image analysis challenge to apply PAIP datasets. The goal of the challenge was to evaluate new and existing algorithms for automated detection of liver cancer in whole-slide images (WSIs). Additionally, the PAIP of this year attempted to address potential future problems of AI applicability in clinical settings. In the challenge, participants were asked to use analytical data and statistical metrics to evaluate the performance of automated algorithms in two different tasks. The participants were given the two different tasks: Task 1 involved investigating Liver Cancer Segmentation and Task 2 involved investigating Viable Tumor Burden Estimation. There was a strong correlation between high performance of teams on both tasks, in which teams that performed well on Task 1 also performed well on Task 2. After evaluation, we summarized the top 11 team's algorithms. We then gave pathological implications on the easily predicted images for cancer segmentation and the challenging images for viable tumor burden estimation. Out of the 231 participants of the PAIP challenge datasets, a total of 64 were submitted from 28 team participants. The submitted algorithms predicted the automatic segmentation on the liver cancer with WSIs to an accuracy of a score estimation of 0.78. The PAIP challenge was created in an effort to combat the lack of research that has been done to address Liver cancer using digital pathology. It remains unclear of how the applicability of AI algorithms created during the challenge can affect clinical diagnoses. However, the results of this dataset and evaluation metric provided has the potential to aid the development and benchmarking of cancer diagnosis and segmentation.

BACH: Grand challenge on breast cancer histology images.

Breast cancer is the most common invasive cancer in women, affecting more than 10% of women worldwide. Microscopic analysis of a biopsy remains one of the most important methods to diagnose the type of breast cancer. This requires specialized analysis by pathologists, in a task that i) is highly time- and cost-consuming and ii) often leads to nonconsensual results. The relevance and potential of automatic classification algorithms using hematoxylin-eosin stained histopathological images has already been demonstrated, but the reported results are still sub-optimal for clinical use. With the goal of advancing the state-of-the-art in automatic classification, the Grand Challenge on BreAst Cancer Histology images (BACH) was organized in conjunction with the 15th International Conference on Image Analysis and Recognition (ICIAR 2018). BACH aimed at the classification and localization of clinically relevant histopathological classes in microscopy and whole-slide images from a large annotated dataset, specifically compiled and made publicly available for the challenge. Following a positive response from the scientific community, a total of 64 submissions, out of 677 registrations, effectively entered the competition. The submitted algorithms improved the state-of-the-art in automatic classification of breast cancer with microscopy images to an accuracy of 87%. Convolutional neuronal networks were the most successful methodology in the BACH challenge. Detailed analysis of the collective results allowed the identification of remaining challenges in the field and recommendations for future developments. The BACH dataset remains publicly available as to promote further improvements to the field of automatic classification in digital pathology.

Development and validation of a novel automated Gleason grade and molecular profile that define a highly predictive prostate cancer progression algorithm-based test.

BACKGROUND: Postoperative risk assessment remains an important variable in the effective treatment of prostate cancer. There is an unmet clinical need for a test with the potential to enhance the Gleason grading system with novel features that more accurately reflect a personalized prediction of clinical failure. METHODS: A prospectively designed retrospective study utilizing 892 patients, post radical prostatectomy, followed for a median of 8 years. In training, using digital image analysis to combine microscopic pattern analysis/machine learning with biomarkers, we evaluated Precise Post-op model results to predict clinical failure in 446 patients. The derived prognostic score was validated in 446 patients. Eligible subjects required complete clinical-pathologic variables and were excluded if they had received neoadjuvant treatment including androgen deprivation, radiation or chemotherapy prior to surgery. No patients were enrolled with metastatic disease prior to surgery. Evaluate the assay using time to event concordance index (C-index), Kaplan-Meier, and hazards ratio. RESULTS: In the training cohort (n = 306), the Precise Post-op test predicted significant clinical failure with a C-index of 0.82, [95% CI: 0.76-0.86], HR:6.7, [95% CI: 3.59-12.45], p < 0.00001. Results were confirmed in validation (n = 284) with a C-index 0.77 [95% CI: 0.72-0.81], HR = 5.4, [95% CI: 2.74-10.52], p < 0.00001. By comparison, a clinical feature base model had a C-index of 0.70 with a HR = 3.7. The Post-Op test also re-classified 58% of CAPRA-S intermediate risk patients as low risk for clinical failure. CONCLUSIONS: Precise Post-op tissue-based test discriminates low from intermediate high risk prostate cancer disease progression in the postoperative setting. Guided by machine learning, the test enhances traditional Gleason grading with novel features that accurately reflect the biology of personalized risk assignment.

Gland ring morphometry for prostate cancer prognosis in multispectral immunofluorescence images.

Morphometric features characterizing the fusion and fragmentation of the glandular architecture of advanced prostate cancer have not previously been based upon the automated segmentation of discrete gland rings, due in part to the difficulty of extracting these structures from the H&E stained tissues. We present a novel approach for segmenting gland rings in multi-spectral immunofluorescence (IF) images and demonstrate the utility of the resultant features in predicting cancer recurrence in a cohort of 1956 images of prostate biopsies and prostatectomies from 679 patients. The proposed approach is evaluated for prediction of actual clinical outcomes of interest to physicians in comparison with previously published gland-unit features, yielding a concordance index (CI) of 0.67. This compares favorably to the CI of 0.66 obtained using a semi-automated segmentation of the corresponding H&E images from the same patients. This work presents the first algorithms for segmentation of gland rings lacking a central lumen, and for separation of touching epithelial units, and introduces new gland adjacency features for predicting prostate cancer clinical progression across both biopsy and prostatectomy images.