International Business Machines Corporation

Tal Tlusty Shapiro

Simona Rabinovici-Cohen

Dana Levanony

Efrat Hexter

Shaked Perek

Ami Abutbul

Ella Barkan

Rami Ben-Ari

Guy Amit

An embodiment for generating balanced train-test splits for machine learning analysis. The embodiment may automatically extract low-level features and high-level features from a series of received datasets. The embodiment may automatically determine a series of impactful features for each of the received datasets correlating to a corresponding label. The embodiment may automatically select subsets of impactful features The embodiment may automatically cluster the received datasets to generate series of clusters, each of the generated series of clusters corresponding to one of the selected subsets of impactful features. The embodiment may automatically generate train-test split versions using datasets from each cluster in each of the generated series of clusters. The embodiment may automatically score each of the generated train-test split versions and select a highest-scoring train-test split version.

Generating balanced train-test splits for machine learning

Embodiments herein disclose computer-implemented methods, computer program products and computer systems for annotating magnetic resonance imaging (MRI) images. The method may include receiving mammogram (MG) image data representing annotated MG images of a patient breast, the annotated MG images being one of either a craniocaudal view or of a mediolateral oblique view. The method may include identifying annotations representing an abnormality at a first location in the annotated MG images; receiving MRI image data representing MRI images of the patient breast; generating annotated MRI image data using the MRI image data and the annotations identified in the annotated MG images, the annotated MRI image data including MRI annotations at a second location based at least in part on the first location, the MRI annotations in the annotated MRI image data representing the abnormality; and storing the annotated MRI image data in a database.

MRI annotation

An approach for improving determining a significant slice associated with a tumor from a volume of medical images is disclosed. The approach is based on the annotation of tumor range and the slice index in which the tumor appears to have the largest area. The approach infer a tumor growth classifier on sliding window of the volume slices and creates a discrete integral function out of the classifier predictions. The approach applies post processing on the discrete integral function which can include a smoothing function and a bias correction. The approach selects the slice index of maximum value from the post processing step.

Medical object detection and identification

Techniques for training machine learning models for improved accuracy at classifying medical imaging data sets by trimming ambiguous samples from training data sets are described herein. In some embodiments, a machine learning model is trained using a data set, where a subset of the data set comprises data with a conflict between a first label based on an expert opinion and a second label based on a ground truth based on a medical examination. During some epochs of training the machine learning model, loss values for each data sample in the epoch are compared against a loss threshold, with data samples with corresponding loss values that exceed the loss threshold that also belong to the subclass trimmed from the data set for subsequent epochs of training. The loss threshold for the next epoch is then updated based on loss values of the trimmed data set.

Training machine learning models to exclude ambiguous data samples

A system for analyzing image data can include a processor to detect a set of medical images and partition each medical image into a plurality of non-overlapping or overlapping patches. The processor can also generate an ordered set of patch clusters derived from the medical images, wherein the patch clusters are sorted based on a type of tissue identified in each of the medical images. Additionally, the processor can generate a digital representation of a percentage of the type of tissue calculated for each of the medical images and transmit an alert in response to detecting that the percentage of the type of tissue for an image exceeds a predetermined threshold.

Growing community of inventors

Zichon Yaacov, Israel

Tal Tlusty Shapiro