0:00:00hello my name is my case some tiny and on the order of the article
0:00:04titled disease progression with cd nist in mc i any subjects
0:00:11the choice of the c d n s one or disease progression analysis in a
0:00:15patient population close to the answer to dimension
0:00:18was guided by addison can be addressed guidance undoubtedly grows or the outsiders disease
0:00:23the draft guidance of course the concept and ridge in clinical trials
0:00:26the patient's most likely to progress to moreover dementia using what clinical and but are
0:00:32basic idea
0:00:33interestingly what patients with this
0:00:36the draft guidance kinds of std lsp as an example of or was a disease
0:00:40progression
0:00:41and as a candidate for a single whatever the efficacy n y
0:00:45that combines assessment of all mission and function
0:00:50this is production has coming to describe using linear exponential and logistic structure models in
0:00:55the literature
0:00:57in this analysis we formally compared to structure models
0:01:00the three bottom adjusting model describes an and shape disease progression curve and follows a
0:01:05nonlinear and such a verbal trajectory that stays within the boundaries of the scale
0:01:10this analysis indicated that the three but i'm logistic model best describes your data
0:01:18a problem can be what's is a very but also the regression israel or at
0:01:23low and high c d n is these scores with human scores of about and
0:01:28exhibiting faster rates are duration
0:01:30this involvement use you have relationship is a cat interest a okay and ship progression
0:01:34curve described by the t parameter logistic model which is also highlighted here
0:01:39the model predicts that the production rate approaches you don't as the scores of course
0:01:43the boundaries of the scale
0:01:45within the range of the scale the production rate is proportional to the current seriously
0:01:49score one parameter that i just e subject compression rate
0:01:54and advantage of the lsp is that less and two percent of the observation reside
0:01:59on the boundaries of the scale for this population and most of the zero one
0:02:02observations will present the subject that we're by the market negative
0:02:06moreover a comparison of models with different probability distributions suggested that the larger normal distribution
0:02:13capture the behavior but assume error appropriately
0:02:16this is a letter model has the advantage of the model predictions even after accounting
0:02:21but as to error stay within the boundaries of the scan
0:02:25a mixture model was to the baseline by marker data
0:02:28and i shows the density goes based on the mixture model with the estimated threshold
0:02:33of zero point one or seven what we are living tissue
0:02:37you don't densities and i don't be sure that cognitively normal and in beaumont the
0:02:42distributions are quite a state with low and high recalibrated issues respectively
0:02:48in contrast the and then see i subjects represent a mixture some of the methodology
0:02:53and some in l
0:02:54the of the baseline by the mother data on me has regression because it ellen's
0:02:59i and my e d is shown by l c and d
0:03:03subjects the baseline be immediate initial below the zero point one four seven to ensure
0:03:08exhibit a scroll data collection was as those subjects about the first four we show
0:03:13how much faster progression
0:03:17equal syllable models that you improvement in more performance
0:03:21additional task using a bimodal analysis
0:03:25the first this study assumption
0:03:26whether the progresses
0:03:28which to ration represented not regressors this as often was found to be true
0:03:33this suggests that the csi biomarkers distinguish progressive still not regressors
0:03:38secondly it was found the best error or not progresses was greater than their vocal
0:03:42service
0:03:43this suggests that study nancy i subjects
0:03:46i do not biomarkers has the potential to reduce errors in a remote controls
0:03:53additional covariate analysis indicated that episodic memory as measured by the national remember a skin
0:03:58related logical memory to an executive functioning well also was also you have a regression
0:04:04the results also suggest that a prosodic boundary is
0:04:07more predictable progression in latency identity or that it was obvious to be true what
0:04:12exactly functioning
0:04:15as part of our analysis it was found that can be does and will have
0:04:19lower scores as compared to that one complete study
0:04:22this was done for by the rubber model which suggested that the likelihood of subjects
0:04:26database missing was related to this score right really about probably now
0:04:32this was really project was part of the model classification
0:04:36the rows represent latency i subjects
0:04:39while the bottom panels
0:04:40sure my at subjects
0:04:42because of the lab
0:04:43after subjects that of a marker negative
0:04:46where the graph on the right
0:04:47or something that i by one
0:04:49the b c suggest that the model described stimulus progression reasoning about in both in
0:04:55the and they then cfs subjects and by market positive and negative subgroups
0:05:01in summary csi biomarkers have the ability to discriminate subjects in the progress snr regressors
0:05:07and that was either one and reading m c i clinical trials
0:05:10this means of presentation and we hope that you will find out analysis useful