Laser-camera data discrepancies and reliable perception in outdoor robotics

This work aims to promote integrity in autonomous perceptual systems, with a focus on outdoor unmanned ground vehicles equipped with a camera and a 2D laser range finder. A method to check for inconsistencies between the data provided by these two heterogeneous sensors is proposed and discussed. First, uncertainties in the estimated transformation between the laser and camera frames are evaluated and propagated up to the projection of the laser points onto the image. Then, for each pair of laser scan-camera image acquired, the information at corners of the laser scan is compared with the content of the image, resulting in a likelihood of correspondence. The result of this process is then used to validate segments of the laser scan that are found to be consistent with the image, while inconsistent segments are rejected. Experimental results illustrate how this technique can improve the reliability of perception in challenging environmental conditions, such as in the presence of airborne dust.

Prognostic impact of vascular and lymphovascular invasion in early lung cancer

Background The prognostic significance of vascular and lymphatic invasion in non-small-cell lung cancer is under continuous debate. We analyzed the effect of tumor aggressiveness (lymphatic and/or vessel invasion) on survival and relapse in stage I and II non-small-cell lung cancer. Methods We retrospectively analyzed prospectively collected data of 457 patients with stage I and II non-small-cell lung cancer from 1998 to 2008. Specimens were analyzed for intratumoral vascular invasion and lymphovascular space invasion. Overall survival and disease-free survival were estimated using the Kaplan-Meier method, and differences were determined by the logrank test. Cox regression analysis was performed to identify independent risk factors. Results: The incidence of intratumoral vascular invasion was 23.4%, and this correlated significantly with grade of differentiation, visceral pleural involvement, lymphovascular space invasion, and N status. The incidence of lymphovascular space invasion was 5.5%, and this correlated significantly with grade of differentiation, lymph nodes involved, and intratumoral vascular invasion. On multivariate analyses, intratumoral vascular invasion proved to be an significant independent risk factor for overall survival but not for disease-free survival. Lymphovascular space invasion was associated significantly with early tumor recurrence but not with overall survival. Conclusions: Vascular and lymphatic invasion can serve as independent prognostic factors in completely resected nonsmall- cell lung cancer. Intratumoral vascular invasion and lymphovascular space invasion in early stage non-small-cell lung cancer are important factors in overall survival and early tumor recurrence. Further large scale studies with more recent patient cohorts and refined histological techniques are warranted.

Genome stability pathways in head and neck cancers

Genomic instability underlies the transformation of host cells toward malignancy, promotes development of invasion and metastasis and shapes the response of established cancer to treatment. In this review, we discuss recent advances in our understanding of genomic stability in squamous cell carcinoma of the head and neck (HNSCC), with an emphasis on DNA repair pathways. HNSCC is characterized by distinct profiles in genome stability between similarly staged cancers that are reflected in risk, treatment response and outcomes. Defective DNA repair generates chromosomal derangement that can cause subsequent alterations in gene expression, and is a hallmark of progression toward carcinoma. Variable functionality of an increasing spectrum of repair gene polymorphisms is associated with increased cancer risk, while aetiological factors such as human papillomavirus, tobacco and alcohol induce significantly different behaviour in induced malignancy, underpinned by differences in genomic stability. Targeted inhibition of signalling receptors has proven to be a clinically-validated therapy, and protein expression of other DNA repair and signalling molecules associated with cancer behaviour could potentially provide a more refined clinical model for prognosis and treatment prediction. Development and expansion of current genomic stability models is furthering our understanding of HNSCC pathophysiology and uncovering new, promising treatment strategies. © 2013 Glenn Jenkins et al.

Towards training-free appearance-based localization : probabilistic models for whole-image descriptors

Whole image descriptors have recently been shown to be remarkably robust to perceptual change especially compared to local features. However, whole-image-based localization systems typically rely on heuristic methods for determining appropriate matching thresholds in a particular environment. These environment-specific tuning requirements and the lack of a meaningful interpretation of these arbitrary thresholds limits the general applicability of these systems. In this paper we present a Bayesian model of probability for whole-image descriptors that can be seamlessly integrated into localization systems designed for probabilistic visual input. We demonstrate this method using CAT-Graph, an appearance-based visual localization system originally designed for a FAB-MAP-style probabilistic input. We show that using whole-image descriptors as visual input extends CAT-Graph’s functionality to environments that experience a greater amount of perceptual change. We also present a method of estimating whole-image probability models in an online manner, removing the need for a prior training phase. We show that this online, automated training method can perform comparably to pre-trained, manually tuned local descriptor methods.