Distance measures for image segmentation evaluation

The task considered in this paper is performance evaluation of region segmentation algorithms in the ground-truth-based paradigm. Given a machine segmentation and a ground-truth segmentation, performance measures are needed. We propose to consider the image segmentation problem as one of data clustering and, as a consequence, to use measures for comparing clusterings developed in statistics and machine learning. By doing so, we obtain a variety of performance measures which have not been used before in image processing. In particular, some of these measures have the highly desired property of being a metric. Experimental results are reported on both synthetic and real data to validate the measures and compare them with others.

Aortic rupture in high-speed skiing crashes

BACKGROUND: New equipment and techniques in winter sports, such as carving skis and snowboards, have brought up new trauma patterns into the spectrum of leisure trauma. The injuries resemble high-energy trauma known from road crashes. The aim of the present study was to assess the incidence of acute traumatic descending aortic rupture in recreational skiing-crashes. MATERIAL: Between January 1995 and December 2004, 22 patients were admitted to our hospital for aortic rupture. Four patients had skiing crashes (18.2%). Mean age was 31 years, all patients were male. In two cases, aortic rupture was associated with fractures of the upper and lower extremities. One patient additionally had a cerebral contusion with an initial Glasgow Coma Scale score of 13. In two patients, isolated aortic rupture was diagnosed. RESULTS: Two patients were treated by graft interposition, and one by endograft. One patient arrived under mechanical resuscitation without blood pressure. He died at admission. He had been observed for 5 hours in another hospital, complaining of severe intrascapular back pain, before transport to our trauma unit for unknown bleeding. In the other three cases, treatment was successful. CONCLUSION: Rescue services and paramedics should be aware of this new type of injury. Acute aortic rupture has to be considered as possible injury in high velocity skiing crashes.

Fitting of Mixtures with Unspecified Number of Components Using Cross Validation Distance Estimate

Estimation of the number of mixture components (k) is an unsolved problem. Available methods for estimation of k include bootstrapping the likelihood ratio test statistics and optimizing a variety of validity functionals such as AIC, BIC/MDL, and ICOMP. We investigate the minimization of distance between fitted mixture model and the true density as a method for estimating k. The distances considered are Kullback-Leibler (KL) and “L sub 2”. We estimate these distances using cross validation. A reliable estimate of k is obtained by voting of B estimates of k corresponding to B cross validation estimates of distance. This estimation methods with KL distance is very similar to Monte Carlo cross validated likelihood methods discussed by Smyth (2000). With focus on univariate normal mixtures, we present simulation studies that compare the cross validated distance method with AIC, BIC/MDL, and ICOMP. We also apply the cross validation estimate of distance approach along with AIC, BIC/MDL and ICOMP approach, to data from an osteoporosis drug trial in order to find groups that differentially respond to treatment.