An artificial intelligence model that automatically labels roux-en-Y gastric bypasses, a comparison to trained surgeon annotators.

INTRODUCTION: Artificial intelligence (AI) can automate certain tasks to improve data collection. Models have been created to annotate the steps of Roux-en-Y Gastric Bypass (RYGB). However, model performance has not been compared with individual surgeon annotator performance. We developed a model that automatically labels RYGB steps and compares its performance to surgeons. METHODS AND PROCEDURES: 545 videos (17 surgeons) of laparoscopic RYGB procedures were collected. An annotation guide (12 steps, 52 tasks) was developed. Steps were annotated by 11 surgeons. Each video was annotated by two surgeons and a third reconciled the differences. A convolutional AI model was trained to identify steps and compared with manual annotation. For modeling, we used 390 videos for training, 95 for validation, and 60 for testing. The performance comparison between AI model versus manual annotation was performed using ANOVA (Analysis of Variance) in a subset of 60 testing videos. We assessed the performance of the model at each step and poor performance was defined (F1-score < 80%). RESULTS: The convolutional model identified 12 steps in the RYGB architecture. Model performance varied at each step [F1 > 90% for 7, and > 80% for 2]. The reconciled manual annotation data (F1 > 80% for > 5 steps) performed better than trainee's (F1 > 80% for 2-5 steps for 4 annotators, and < 2 steps for 4 annotators). In testing subset, certain steps had low performance, indicating potential ambiguities in surgical landmarks. Additionally, some videos were easier to annotate than others, suggesting variability. After controlling for variability, the AI algorithm was comparable to the manual (p < 0.0001). CONCLUSION: AI can be used to identify surgical landmarks in RYGB comparable to the manual process. AI was more accurate to recognize some landmarks more accurately than surgeons. This technology has the potential to improve surgical training by assessing the learning curves of surgeons at scale.

Automated classification of brain tumor type in whole-slide digital pathology images using local representative tiles.

Computerized analysis of digital pathology images offers the potential of improving clinical care (e.g. automated diagnosis) and catalyzing research (e.g. discovering disease subtypes). There are two key challenges thwarting computerized analysis of digital pathology images: first, whole slide pathology images are massive, making computerized analysis inefficient, and second, diverse tissue regions in whole slide images that are not directly relevant to the disease may mislead computerized diagnosis algorithms. We propose a method to overcome both of these challenges that utilizes a coarse-to-fine analysis of the localized characteristics in pathology images. An initial surveying stage analyzes the diversity of coarse regions in the whole slide image. This includes extraction of spatially localized features of shape, color and texture from tiled regions covering the slide. Dimensionality reduction of the features assesses the image diversity in the tiled regions and clustering creates representative groups. A second stage provides a detailed analysis of a single representative tile from each group. An Elastic Net classifier produces a diagnostic decision value for each representative tile. A weighted voting scheme aggregates the decision values from these tiles to obtain a diagnosis at the whole slide level. We evaluated our method by automatically classifying 302 brain cancer cases into two possible diagnoses (glioblastoma multiforme (N = 182) versus lower grade glioma (N = 120)) with an accuracy of 93.1% (p << 0.001). We also evaluated our method in the dataset provided for the 2014 MICCAI Pathology Classification Challenge, in which our method, trained and tested using 5-fold cross validation, produced a classification accuracy of 100% (p << 0.001). Our method showed high stability and robustness to parameter variation, with accuracy varying between 95.5% and 100% when evaluated for a wide range of parameters. Our approach may be useful to automatically differentiate between the two cancer subtypes.

A method for normalizing pathology images to improve feature extraction for quantitative pathology.

PURPOSE: With the advent of digital slide scanning technologies and the potential proliferation of large repositories of digital pathology images, many research studies can leverage these data for biomedical discovery and to develop clinical applications. However, quantitative analysis of digital pathology images is impeded by batch effects generated by varied staining protocols and staining conditions of pathological slides. METHODS: To overcome this problem, this paper proposes a novel, fully automated stain normalization method to reduce batch effects and thus aid research in digital pathology applications. Their method, intensity centering and histogram equalization (ICHE), normalizes a diverse set of pathology images by first scaling the centroids of the intensity histograms to a common point and then applying a modified version of contrast-limited adaptive histogram equalization. Normalization was performed on two datasets of digitized hematoxylin and eosin (H&E) slides of different tissue slices from the same lung tumor, and one immunohistochemistry dataset of digitized slides created by restaining one of the H&E datasets. RESULTS: The ICHE method was evaluated based on image intensity values, quantitative features, and the effect on downstream applications, such as a computer aided diagnosis. For comparison, three methods from the literature were reimplemented and evaluated using the same criteria. The authors found that ICHE not only improved performance compared with un-normalized images, but in most cases showed improvement compared with previous methods for correcting batch effects in the literature. CONCLUSIONS: ICHE may be a useful preprocessing step a digital pathology image processing pipeline.