No abstract

Two experiments investigated the neural mechanisms of Gestalt grouping by recording high-density event-related brain potentials (ERPs) during discrimination tasks. In Experiment 1, stimulus arrays contained luminance-defined local elements that were either evenly spaced or grouped into columns or rows based on either proximity or similarity of shape. Proximity grouping was indexed by a short-latency positivity (110-120 ms) over the medial occipital cortex and a subsequent right occipitoparietal negativity. Grouping by similarity was reflected only in a long-latency occipitotemporal negativity. In Experiment 2, proximity grouping was examined when local elements were defined by motion cues, and was again associated with a medial occipital positivity. However, the subsequent long-latency negativity was now enhanced over the left posterior areas. The implications of these results to the neural substrates subserving different grouping processes are discussed.

No abstract

Mammography is successfully used as an effective screening tool for cancer diagnosis. A calcification cluster on mammography is a primary sign of cancer. Early researches have proved the diagnostic value of the calcification, yet their performance is highly dependent on handcrafted image descriptors. Characterizing the calcification mammography in an automatic and robust way remains a challenge. In this paper, the calcification was characterized by descriptors obtained from deep learning and handcrafted descriptors. We compared the performances of different image feature sets on digital mammograms. The feature sets included the deep features alone, the handcrafted features, their combination, and the filtered deep features. Experimental results have demonstrated that the deep features outperform handcrafted features, but the handcrafted features can provide complementary information for deep features. We achieved a classification precision of 89.32% and sensitivity of 86.89% using the filtered deep features, which is the best performance among all the feature sets.