Health Informatics Grand Rounds, November 22: Martin Hofmann-Apitius

I livetweeted this event. The emphasis in the marketing & early portion of the talk was about how to create bridges and facilitate communication across different scientific research communities and specialities. I found the approach taken to this end provocative and illuminating. As someone who attempts to follow the activities of the open-science / Science 2.0 communities, I was expecting something that would connect to that, something that focused on communication, collaboration, and tools to foster those types of social connections. Instead, as I should have perhaps expected, the talk focused on developing computational strategies and tools to create data and images that could potentially be used and comprehended by researchers in the different specialties. There seemed to be an “if we build it, they will come” mentality to the project — to paraphrase, perhaps something like: if we want you to understand what we are saying, then we will say our ‘stuff’ in terms or images you can understand, and then you will understand. Perhaps, I am misunderstanding? The tools looked very intriguing and useful, but I would really like to see this project embedded in a social context that facilitated communication across boundaries in an active rather than passive way.

Here are the tweets, with links.

November 22, 2010
2:09 pm pfanderson: Livetweeting Martin Hofmann-Apitius, Bioinformatics @ Fraunhofer “Integrating knowledge-drive & simulation-driven approaches #umgrndrds
2:10 pm pfanderson: Using information extraction to build bridges between people who usually apply equations & those who use statistics #umgrndrds
2:14 pm pfanderson: Data mining vs engineering simulation as a cultural issue. Need to be explicit about assumptions #umgrndrds
2:15 pm pfanderson: MHA describes how simulator folks sometimes believe it is unnecessary complexity to model dynamics of biological systems #umgrndrds
2:16 pm pfanderson: Examples @neurIST project #umgrndrds
2:17 pm pfanderson: MHA emphasizes importance of scientists in different domains communicating & collaborating #umgrndrds
2:18 pm pfanderson: “Mickey Mouse Slides” as how to tease abt the lovely simulation images that no one understands except the creator #umgrndrds
2:20 pm pfanderson: Intracranial aneurysms have high prevalence, low incidence. Gdlns fro Tx are diameter > 7, Tx + risk of rupture. Shape analysis? #umgrndrds
2:22 pm pfanderson: Rupture risk based on morphology (shape) seems 2improve decision making (but intro stratification/variables) >1d vs 2d analysis #umgrndrds
2:24 pm pfanderson: MHA describes SCAIview Article: #umgrndrds
2:24 pm pfanderson: Fraunhofer ProMiner tool #umgrndrds
2:27 pm pfanderson: ProMiner: Uses of textmining in professional literature to get at both sides of the literature #umgrndrds (Love of color to get grants?)
2:28 pm pfanderson: Application side/clinical = who cares about the tool unless it gives you the ifnormation you need? #umgrndrds
2:29 pm pfanderson: Oh I love this >> extracting information from chemistry patents #umgrndrds
2:30 pm pfanderson: Neat. Query: Find all patents that INvalidate proposed new patent. << what inventors / lawyers have always wanted #umgrndrds
2:32 pm pfanderson: More on the Fraunhofer patent search tests #umgrndrds
2:33 pm pfanderson: MHA is talking abt the I2B2 Challenge ’10, (mining health records) but I can’t find a link fast #umgrndrds
2:36 pm pfanderson: Brian Athey talking abt need for strong collaborations/partners between USA scientists & EU #umgrndrds
2:40 pm pfanderson: Measures: Concept identification task, statistical significance of performance, assertion classification task #umgrndrds
2:47 pm pfanderson: MHA: Direct use of information extraction fr scientific text for modeling and simulation in the life sciences #umgrndrds
2:47 pm pfanderson: W/ Model of disease state as network, what are the best network nodes to target for intervention? #umgrndrds
2:49 pm pfanderson: Direct Use of Information Extraction from Scientific Text for Modeling and Simulation in the Life Sciences #umgrndrds
2:50 pm pfanderson: “network of brain biomarker proteins in alzheimer’s” >> pretty picture 😉 #umgrndrds
2:51 pm pfanderson: Innovative Medicine Initiative #umgrndrds
2:55 pm pfanderson: PLIO – protein-ligand interaction ontology (can’t find a link) by Ivchenko #umgrndrds
2:57 pm pfanderson: PLIO search: Simulation vs experiment / confirming stmts vs. conflicting = information gain #umgrndrds
2:58 pm pfanderson: Can connecting semantics to mathematical terms help bridge the communication/comprehension divide between disciplines? #umgrndrds
3:00 pm pfanderson: All laugh @ MHA’s “Thank You” slide, debate erupts whether typewriter is IBM Selectric or Wheelwriter, illus avg age of audience #umgrndrds
3:01 pm pfanderson: What to do when best practice guidelines / Cochrane reviews contradict each other? Potential 4 recombination / textmining systems #umgrndrds
3:02 pm pfanderson: MHA: “Clinicians confronted w/ something they aren’t prepared to deal with – the informed patient!” #umgrndrds
3:03 pm pfanderson: MHA: Re emerging treatment options “It will come, and driven by the patients.” #umgrndrds
3:04 pm pfanderson: MHA: “How do we get medical doctors at least as informed as the patients are?” #umgrndrds
3:04 pm pfanderson: Q fr audience: How could we use textmining to link importance of case reports in detecting rare events? #umgrndrds
3:06 pm pfanderson: EU-ADR project on textmining in drug reactions (fyi >> @fda #fda #pharma) #umgrndrds
3:07 pm pfanderson: MHA proposes using textmining to compare individual patient path to best practices #umgrndrds < what abt clinician judgement?
3:08 pm pfanderson: The Q&A is often the best and most informative part of the talk. Clarifies the jargon & proposed implementation, ramifications #umgrndrds


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s