A positive/unlabeled approach for the segmentation of medical sequences using point-wise supervision.

The ability to quickly annotate medical imaging data plays a critical role in training deep learning frameworks for segmentation. Doing so for image volumes or video sequences is even more pressing as annotating these is particularly burdensome. To alleviate this problem, this work proposes a new method to efficiently segment medical imaging volumes or videos using point-wise annotations only. This allows annotations to be collected extremely quickly and remains applicable to numerous segmentation tasks. Our approach trains a deep learning model using an appropriate Positive/Unlabeled objective function using sparse point-wise annotations. While most methods of this kind assume that the proportion of positive samples in the data is known a-priori, we introduce a novel self-supervised method to estimate this prior efficiently by combining a Bayesian estimation framework and new stopping criteria. Our method iteratively estimates appropriate class priors and yields high segmentation quality for a variety of object types and imaging modalities. In addition, by leveraging a spatio-temporal tracking framework, we regularize our predictions by leveraging the complete data volume. We show experimentally that our approach outperforms state-of-the-art methods tailored to the same problem.

Iterative multi-path tracking for video and volume segmentation with sparse point supervision.

Recent machine learning strategies for segmentation tasks have shown great ability when trained on large pixel-wise annotated image datasets. It remains a major challenge however to aggregate such datasets, as the time and monetary cost associated with collecting extensive annotations is extremely high. This is particularly the case for generating precise pixel-wise annotations in video and volumetric image data. To this end, this work presents a novel framework to produce pixel-wise segmentations using minimal supervision. Our method relies on 2D point supervision, whereby a single 2D location within an object of interest is provided on each image of the data. Our method then estimates the object appearance in a semi-supervised fashion by learning object-image-specific features and by using these in a semi-supervised learning framework. Our object model is then used in a graph-based optimization problem that takes into account all provided locations and the image data in order to infer the complete pixel-wise segmentation. In practice, we solve this optimally as a tracking problem using a K-shortest path approach. Both the object model and segmentation are then refined iteratively to further improve the final segmentation. We show that by collecting 2D locations using a gaze tracker, our approach can provide state-of-the-art segmentations on a range of objects and image modalities (video and 3D volumes), and that these can then be used to train supervised machine learning classifiers.

Incidence and nature of karate injuries.

OBJECTIVES: To determine the incidence and nature of karate injuries sustained in karate clubs and to identify risk factors for injuries. METHODS: One hundred eighty-six individuals from three karate clubs in Brest, France, were entered in a retrospective study extending from September 2002 to June 2003. Each athlete was asked to complete a questionnaire on karate injuries sustained during the previous year (type, location, mechanism, exercise during which the injury occurred, number of days off training and work, and medical care). Injury types were described by number of injuries and risk factors per number of injured athletes. RESULTS: Forty-eight (28.8%) of the 186 athletes sustained 83 injuries (63 while training and 20 while competing). The annual injury rate was 44.6 per 100 athletes. Incidence rates were similar in males and females and across the three clubs but increased with age, time spent training (3.6+/-1.7 vs. 2.9+/-1.5 h/week; P=0.001), rank (lower ranks vs. brown and black belts, P=0.015), and years of practice (7.3+/-5.5 years in athletes with injuries vs. 5.1+/-4.8 in those without injuries; P=0.03). Injuries consisted of 43 (53%) hematomas, 16 (19%) sprains, seven (7%) muscle lesions, six (7%) fractures, four (5%) malaise episodes, and seven (7%) miscellaneous lesions. Time off training occurred for 26 (31.3%) injuries and ranged from 8 to >30 days. The body region involved was the head in 22 (26.5%) injuries, the torso in eight injuries (9.6%), the upper limb in 24 (28.9%) injuries, and the lower limb in 29 (35%) injuries. CONCLUSION: Karate injuries are fairly common but usually minor. They are more likely to occur during competitions than while training. The head and limbs are the main regions involved. Longer training times per week and higher rank are associated with an increased risk of injury. Prevention seems crucial.