Training Nuclei Detection Algorithms with Simple Annotations

Abstract

Background:Generating good training datasets is essential for machine learning-based nuclei detection methods. However, creating exhaustive nuclei contour annotations, to derive optimal training data from, is often infeasible.Methods:We compared different approaches for training nuclei detection methods solely based on nucleus center markers. Such markers contain less accurate information, especially with regard to nuclear boundaries, but can be produced much easier and in greater quantities. The approaches use different automated sample extraction methods to derive image positions and class labels from nucleus center markers. In addition, the approaches use different automated sample selection methods to improve the detection quality of the classification algorithm and reduce the run time of the training process. We evaluated the approaches based on a previously published generic nuclei detection algorithm and a set of Ki-67-stained breast cancer images.Results:A Voronoi tessellation-based sample extraction method produced the best performing training sets. However, subsampling of the extracted training samples was crucial. Even simple class balancing improved the detection quality considerably. The incorporation of active learning led to a further increase in detection quality.Conclusions:With appropriate sample extraction and selection methods, nuclei detection algorithms trained on the basis of simple center marker annotations can produce comparable quality to algorithms trained on conventionally created training sets.

Deep Learning Accurately Quantifies Plasma Cell Percentages on CD138-Stained Bone Marrow Samples

“…Pairs were considered matching if they were within a distance threshold of 15px (3.75 μm), similar to other approaches. 15 , 16 …”

The human‐in‐the‐loop: an evaluation of pathologists’ interaction with artificial intelligence in clinical practice

“…To further investigate possible clinical impacts of Ki67 status, discordances were mapped as status error. The performance of the algorithm dropped in the clinical environment; this was a domain shift, as reflected by an F1 score for V2 of 0.68 as compared with the formerly reported 0.83 25 …”

“…The nuclear detection algorithm used in the system was based on machine learning, with pixel annotations in V1 24 and a larger number of point annotations in V2. 25 The cell detection F1 scores were 0.79 for V1 and 0.83 for V2, as determined from an image dataset described by Molin et al 26 The computeraided tool was fully integrated in the viewer.…”

“…The performance of the algorithm dropped in the clinical environment; this was a domain shift, as reflected by an F1 score for V2 of 0.68 as compared with the formerly reported 0.83. 25…”

Deep learning accurately quantifies plasma cell percentages on CD138-stained bone marrow samples

“…A linear sum assignment algorithm was used to match these point detections against the ground truth labels based on their proximities so that each detection could be placed in the three-class (CD138+, CD138−, and background) confusion matrix. Pairs were considered matching if they were within a distance threshold of 15px (3.75 µm), similar to other approaches [15,16].…”