Added sections on embeddings for medical ontologies and causal inference #339
Conversation
…dinal Lab Tests in Electronic Helath Records data set.
…rning, so I don't know if it should be included, but it is relevant ot the section.
|
Thanks @DaveDeCaprio. I'll look at the Treat section update this afternoon. @cgreene can you please review the Categorize updates? @DaveDeCaprio can you also inspect the Travis CI build failure? I believe the new references are missing |
|
Will take a look at the build failure now. Sorry about those, I couldn’t build locally and had to run out right after the pull.
Dave
From: Anthony Gitter [mailto:notifications@github.com]
Sent: Sunday, April 23, 2017 7:49 AM
To: greenelab/deep-review <deep-review@noreply.github.com>
Cc: David DeCaprio <daved@alum.mit.edu>; Mention <mention@noreply.github.com>
Subject: Re: [greenelab/deep-review] Added sections on embeddings for medical ontologies and causal inference (#339)
Thanks @DaveDeCaprio <https://github.com/DaveDeCaprio> . I'll look at the Treat section update this afternoon. @cgreene <https://github.com/cgreene> can you please review the Categorize updates?
@DaveDeCaprio <https://github.com/DaveDeCaprio> can you also inspect the Travis CI build failure? I believe the new references are missing @ or @doi: in some cases.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub <#339 (comment)> , or mute the thread <https://github.com/notifications/unsubscribe-auth/AAzVupgCLr7EQO-Lhp0UwaxEkzJHJa5vks5ry0iugaJpZM4NFQmx> . <https://github.com/notifications/beacon/AAzVutdvBuWRQotQ31rStnPlWMyEhoTEks5ry0iugaJpZM4NFQmx.gif>
|
| was first used on free text medical notes by De Vine et al. | ||
| [@doi:10.1145/2661829.2661974] with results at or better than traditional methods. | ||
| Y. Choi et al.[@doi:10.1145/2567948.2577348] built embeddings of standardized | ||
| terminologies, such as ICD and NDC, used in widely available administrative |
There was a problem hiding this comment.
just want to make sure for when we have to look for the first occurrence.
There was a problem hiding this comment.
Yes, that is correct for NDC
There was a problem hiding this comment.
Ok - when we go through and check acronyms + define at first occurrence I'll know now. Thanks! I think the DOI below is my only requested change.
| results support the potential of unsupervised feature construction in this | ||
| domain. However, numerous challenges including data integration (patient | ||
| feature construction methods [@doi:10.1038/srep26094]. Razavian et al. | ||
| [@arxiv:1608.00647] used a set of 18 common lab tests to predict disease onset |
There was a problem hiding this comment.
Good addition + touches on different architectures.
sections/03_categorize.md
Outdated
| @@ -236,7 +253,9 @@ making methodological choices that either reduce the need for labeled examples | |||
| or that use transformations to training data to increase the number of times it | |||
| can be used before overfitting occurs. For example, the unsupervised and | |||
| semi-supervised methods that we've discussed reduce the need for labeled | |||
| examples [@doi:10.1016/j.jbi.2016.10.007]. The adversarial training example | |||
| examples [@doi:10.1016/j.jbi.2016.10.007]. The anchor and learn framework | |||
| [@pmid:27107443] uses expert knowledge to identify high confidence | |||
There was a problem hiding this comment.
Can you make this a @doi tag so we can automatically snag the reference information?
I think the DOI is: 10.1093/jamia/ocw011
There was a problem hiding this comment.
I just changed it to the doi. The build flagged this as possibly formatted wrong, but I think it is correct. The link works.
There was a problem hiding this comment.
Very nice addition. I only have these minor comments and then will merge.
@blengerich this builds on what you wrote so I'm tagging you in case you have comments before or after the merge.
sections/05_treat.md
Outdated
| [@doi:10.1037/h0037350]. Johansson [@arxiv:1605.03661] use deep neural networks | ||
| to create representation models for covariates that capture nonlinear effects | ||
| and show significant performance improvements over existing models. In a less | ||
| formal approach, Kale et al [@pmid:PMC4765623] first create a deep neural |
There was a problem hiding this comment.
Can you please convert the PMC reference to [@pmid:26958203]? It looks like it built okay, but this will help keep things slightly more standardized.
There was a problem hiding this comment.
I can't find a doi for this. I used https://www.ncbi.nlm.nih.gov/pmc/pmctopmid/ and looked around but the only identifier I can find is the pmc id.
There was a problem hiding this comment.
hmm - found the google scholar identifier and I also don't see a DOI on any of these versions:
https://scholar.google.com/scholar?oi=bibs&hl=en&cluster=4098913113177273200
Odd.
There was a problem hiding this comment.
I found that PMC id on the PubMed page https://www.ncbi.nlm.nih.gov/pubmed/26958203, which is where I came up with [@pmid:26958203]. Is that the correct reference?
There was a problem hiding this comment.
Ahh - missed that it was PMC! Good eyes @agitter!
sections/05_treat.md
Outdated
| A critical challenge in moving from prediction to treatment recommendations | ||
| is the necessity to establish a causal relationship for a recommendation. | ||
| Causal inference is often framed in terms of counterfactual question | ||
| [@doi:10.1037/h0037350]. Johansson [@arxiv:1605.03661] use deep neural networks |
|
Since there doesn't appear to be a DOI for that article, can you make this (#339 (comment)) change? Then we'll get this merged. |
|
I restarted the Travis build and it passed this time. I'm merging. |
…nce (#339) This build is based on 22c54f0. This commit was created by the following Travis CI build and job: https://travis-ci.org/greenelab/deep-review/builds/225555590 https://travis-ci.org/greenelab/deep-review/jobs/225555591 [ci skip] The full commit message that triggered this build is copied below: Added sections on embeddings for medical ontologies and causal inference (#339) * Add reference to Multi-task Prediction of Disease Onsets from Longitudinal Lab Tests in Electronic Helath Records data set. * Added in the anchor and learn framework. This isn't strictly deep learning, so I don't know if it should be included, but it is relevant ot the section. * Added in references to neural embeddings in medical coding * Added causal inference references * Changed pmid to doi * Changed PMC id to regular PubMed id.
…nce (#339) This build is based on 22c54f0. This commit was created by the following Travis CI build and job: https://travis-ci.org/greenelab/deep-review/builds/225555590 https://travis-ci.org/greenelab/deep-review/jobs/225555591 [ci skip] The full commit message that triggered this build is copied below: Added sections on embeddings for medical ontologies and causal inference (#339) * Add reference to Multi-task Prediction of Disease Onsets from Longitudinal Lab Tests in Electronic Helath Records data set. * Added in the anchor and learn framework. This isn't strictly deep learning, so I don't know if it should be included, but it is relevant ot the section. * Added in references to neural embeddings in medical coding * Added causal inference references * Changed pmid to doi * Changed PMC id to regular PubMed id.
I've made several suggested edits. Each logical change is in a separate commit. There were a couple cases where I added an individual reference that I thought was significant that wasn't already present. In two cases I added an additional paragraph to cover a topic that wasn't included.
I've tried to follow the existing style as much as possible. Let me know if don't like it and I can make any edits.
Thanks,
Dave