en
0.25
0.5
0.75
1.25
1.5
1.75
2
Unsupervised Learning of Disease Progression Models
Published on Oct 08, 20142617 Views
Chronic diseases, such as Alzheimer's Disease, Diabetes, and Chronic Obstructive Pulmonary Disease, usually progress slowly over a long period of time, causing increasing burden to the patients, their
Related categories
Chapter list
The burden of chronic diseases00:00
COPD diagnosis & progression00:43
Our contribution01:59
Learn from Electronic Health Records (EHR)03:14
Challenges of disease progression modeling03:53
Our overall model04:39
Markov Jump Process05:18
Model for data at single point in Ime: Noisy-‐OR network - 106:21
Model for data at single point in Ime: Noisy-‐OR network - 206:45
Model for data at single point in Ime: Noisy-‐OR network - 307:34
Model for data at single point in Ime: Noisy-‐OR network - 407:54
Anchored noisy-‐OR network08:11
Model of comorbidiIes across Ime08:49
Inference09:25
CustomizaIons for COPD10:20
Experimental evaluaIon10:36
Specifying the latent variables11:12
Which edges are learned?11:43
Edges learned for kidney disease - 111:50
Edges learned for kidney disease - 212:01
Edges learned for kidney disease - 312:22
Edges learned for lung cancer - 112:38
Edges learned for lung cancer - 212:41
Edges learned for lung cancer - 312:46
Edges learned for lung infection12:53
Prevalence of comorbidiIes across stages (Lung cancer and Obesity)12:54
Prevalence of comorbidiIes across stages (Kidney disease)13:32
Prevalence of comorbidiIes across stages (Lung infecIons)13:39
Prevalence of comorbidiIes across stages (Diabetes & Musculoskeletal disorders)13:49
Prevalence of comorbidiIes across stages (Cardiovascular disease) - 113:54
Prevalence of comorbidiIes across stages (Cardiovascular disease) - 214:03
Inference of disease progression - 114:20
Inference of disease progression - 214:44
Conclusion and future work14:55