We present a study of multiple sclerosis segmentation algorithms conducted at the international MICCAI 2016 challenge. This challenge was operated using a new open-science computing infrastructure. This allowed for the automatic and independent evaluation of a large range of algorithms in a fair and completely automatic manner. This computing infrastructure was used to evaluate thirteen methods of MS lesions segmentation, exploring a broad range of state-of-theart algorithms, against a high-quality database of 53 MS cases coming from four centers following a common definition of the acquisition protocol. Each case was annotated manually by an unprecedented number of seven different experts. Results of the challenge highlighted that automatic algorithms, including the recent machine learning methods (random forests, deep learning, …), are still trailing human expertise on both detection and delineation criteria. In addition, we demonstrate that computing a statistically robust consensus of the algorithms performs closer to human expertise on one score (segmentation) although still trailing on detection scores.

Email address: Olivier.Commowick@inria.fr (Olivier Commowick) Preprint submitted to Nature Scientific Reports July 12, 2018 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint . http://dx.doi.org/10.1101/367557 doi: bioRxiv preprint first posted online Jul. 13, 2018; We present a study of multiple sclerosis segmentation algorithms conducted at the international MICCAI 2016 challenge. This challenge was operated using a new open-science computing infrastructure. This allowed for the automatic and independent evaluation of a large range of algorithms in a fair and completely automatic manner. This computing infrastructure was used to evaluate thirteen methods of MS lesions segmentation, exploring a broad range of state-of-theart algorithms, against a high-quality database of 53 MS cases coming from four centers following a common definition of the acquisition protocol. Each case was annotated manually by an unprecedented number of seven different experts. Results of the challenge highlighted that automatic algorithms, including the recent machine learning methods (random forests, deep learning, . . . ), are still trailing human expertise on both detection and delineation criteria.In addition, we demonstrate that computing a statistically robust consensus of the algorithms performs closer to human expertise on one score (segmentation) although still trailing on detection scores.

<p>We present a study of multiple sclerosis segmentation algorithms conducted at the international MICCAI 2016 challenge. This challenge was operated using a new open-science computing infrastructure. This allowed for the automatic and independent evaluation of a large range of algorithms in a fair and completely automatic manner. This computing infrastructure was used to evaluate thirteen methods of MS lesions segmentation, exploring a broad range of state-of-theart algorithms, against a high-quality database of 53 MS cases coming from four centers following a common definition of the acquisition protocol. Each case was annotated manually by an unprecedented number of seven different experts. Results of the challenge highlighted that automatic algorithms, including the recent machine learning methods (random forests, deep learning, …), are still trailing human expertise on both detection and delineation criteria. In addition, we demonstrate that computing a statistically robust consensus of the algorithms performs closer to human expertise on one score (segmentation) although still trailing on detection scores.</p> <p><br></p>

a b s t r a c tSynthetic amorphous silica (SAS) like NM-200 is used in a wide variety of technological applications and consumer products. Although SAS has been widely investigated the available reproductive toxicity studies are old and do not cover all requirements of current OECD Guidelines. As part of a CEFIC-LRI project, NM-200 was tested in a two-generation reproduction toxicity study according to OECD guideline 416. Male and female rats were treated by oral gavage with NM-200 at dose levels of 0, 100, 300 and 1000 mg/kg bw/day for two generations. Body weight and food consumption were measured throughout the study. Reproductive and developmental parameters were measured and at sacrifice (reproductive) organs and tissues were sampled for histopathological analysis. Oral administration of NM-200 up to 1000 mg/kg bw/day had no adverse effects on the reproductive performance of rats or on the growth and development of the offspring into adulthood for two consecutive generations. The NOAEL was 1000 mg/kg body weight per day.

No abstract

No abstract

pMut+, n¼32) with Clinical Dementia Rating (CDR) scale global score of 0, and symptomatic mutation carriers (sMut+, n¼41) with CDR>0. Whole brain, ventricle, and hippocampal regions were delineated from each image, and measures of atrophy were calculated over these regions using the groupwise Boundary Shift Integral (BSI). Results: Demographics and rates of change are shown in the Table. The sMut+ group had higher rates of atrophy than both NMC and pMut+. While no differences were observed between NMC and pMut+ groups, a significant (p¼0.049) positive correlation between brain atrophy and expected onset was observed (p¼0.049, see Figure). When using the trial eligible population, preliminary sample size point estimates of 748 (brain), 709 (ventricle), 1317 (left hippocampus), and 902 (right hippocampus) per arm are needed to detect a 25% change in atrophy over one year (corrected for ageing) with 95% significance and 80% power. Conclusions: Brain atrophy using longitudinal MR scans show changes in key regions of affected carriers compared to non-carriers. The inclusion criteria proposed for these trials appear to be sensible if using atrophy as one of the disease modification endpoints.

No abstract