Yottalook Update

New sites have been added to the Yottalook Web. In addition, Radiology Today Blog was added to the Yottalook Reader.

Search Yottalook in 33 Languages!

Today, we launched our beta version of Yottalook Worldwide, which will allow users to search in one of 33 different languages. People from over 160 countries have used Yottalook to date, and we wanted to improve user experience for whom English may not be their first language.

We are using Google Language API to provide translations of search term queries. When you type in a search term in one of 33 different languages supported, Google Language API translates the search string into English. This then gets sent to our Yottalook Images or Journals engine and the results are shown. Once translated, the search terms undergo same query expansion as Yottalook in English.

Following languages are supported:

• Arabic
• Bulgarian
• Chinese (Simplified and Traditional)
• Catalan
• Croatian
• Czech
• Danish
• Dutch
• English
• Finnish
• French
• German
• Greek
• Hebrew
• Hindi
• Indonesian
• Italian
• Japanese
• Korean
• Latvian
• Lithuanian
• Norwegian
• Polish
• Portuguese
• Romanian
• Russian
• Spanish
• Serbian
• Slovak
• Slovenian
• Swedish
• Ukrainian
• Vietnamese

You can type your search term in one of the 33 supported languages in the search box then select your language from the drop-down menu. Once you click on search, your query term(s) will be translated into English and search results will be shown. Your language preference will be saved for each online session. Once in the results page, you can search either in the language you selected previously or in English as the search box will understand both. For Spanish, French and Portuguese, we have implemented additional features. Look for the button; when you mouse over it, you will see translations of the caption appear (please give it a second or so to perform the translation; we are aware that the popup box with translation sometimes does not disappear, if you let it complete its translation process prior to moving your mouse that help preventing this. However, we are looking into having a better solution for this issue).

Same Yottalook Terms of Use and Disclaimer apply. Please note that this is currently in BETA. We are aware of limitations of the Google Language API and are looking for more ways to improve this engine. For example, that is why the button option has been only turned on for 3 languages so far. I have personally tested out Korean and found this feature to be not very helpful. If you are interested in testing out different languages, please contact us as we can generate special links that will allow you to test this feature.

Please help us improve this engine by providing us with your feedback. Despite limitations, we hope many of you will find this new interface useful.

Yottalook Local Exhibit Awarded RSNA 2008 Certificate of Merit

One of our exhibits ("Yottalook Local™: Demonstration of an Enterprise-wide Semantic Search Engine for Patient Information") was awarded Certificate of Merit from RSNA 2008.

New Images from Imaging Consult

We have just indexed more images to our search engine in anticipation of RSNA. Yottalook Images will now also show results from Elsevier's Imaging Consult. We are currently working to increase our index even more. Stay tuned for more updates.

Yottalook Integrated Within GE Centricity PACS-IW

GE and Yottalook have created a working prototype integration between GE's Centricity PACS-IW and Yottalook. Go to GE Healthcare IT (Booth #5529) at RSNA to check it out.

Yottalook's New Updated Images Engine

We have updated our Images engine to include more images.

If you look at the middle of the screen where we have our filters, you will find a brand new filter option called "No Subscription Mode: ON OFF". We have received many feedback requesting ability to filter out images where users don't have subscriptions to. We understand, that while many users from large academic centers may have access to many journals online, this is not the case for many of our users. By turning this option, Yottalook Images will attemp to maximize results that do not require subscription for viewing and bring those results near the top.

Our update includes more than increased number of images. We have also enabled media RSS capability. Download Cooliris, and click on "Start Slide Show" next to the bookmark button below the search box. Experience the new "3D Wall" that lets you browse through hundreds of images with ease and speed. Click on the image to view a larger version as well as caption.

Yottalook at RSNA 2008

Want to find out more about Yottalook? During RSNA this year, we will be having following sessions/presentations and exhibits.

Yottalook Local™: Demonstration of an Enterprise-wide Semantic Search Engine for Patient Information

This exhibit demonstrates the use and applications of an enterprise-wide semantic search engine to index radiology reports. The Yottalook Local search engine allows radiologists to search for patient information and reports quickly. By analyzing user queries in real time, the engine further enhances the user search experience by providing thumbnails, related terms using various ontologies, query-related special content, and modification of search results based on specific query terms used.

EXHIBIT HOURS:
Sun (Nov 30) 8:00 AM - 6:00 PM
Mon (Dec 1) - Thurs (Dec 4) 7:00 AM - 10:00 PM
Fri (Dec 5) 7:00 AM - 12:45 PM
LOCATION: Lakeside Learning Center
Click here for more info

Multimedia RSS for Immersive Display of Radiology Image Search Results

Multimedia RSS (MRSS) can be used to display radiology image search results from the web in a highly optimized and immersive 3D environment which permits the user to zoom, pan, and "fly-though" their search results. We will discuss similarities and difference between standard MRSS and RSS and how to use various online and client RSS readers to subscribe to MRSS feeds. We will also demonstrate how to use MRSS to distribute other multimedia content including radiology resident conference video.

EXHIBIT HOURS:
Sun (Nov 30) 8:00 AM - 6:00 PM
Mon (Dec 1) - Thurs (Dec 4) 7:00 AM - 10:00 PM
Fri (Dec 5) 7:00 AM - 12:45 PM
LOCATION: Lakeside Learning Center
Click here for more info

SESSION: What are Radiologists Searching For? - Obtaining Quality Information at the Desktop, Radiology Specific Search Engines and Beyond
Finding the right information at the time of patient care: Yottalook case study

The web search engine has become the preferred decision support tool by Radiologists, however, overall satisfaction levels for websearch remain relatively low. Multiple factors contribute to this dissatisfaction including: (1) content returned by a search often does not adequately address the question being asked, (2) the content is not presented in a way that it can be "digested" easily, (3) multiple searches or search refinements are frequently required to achieve success. The purpose of this session is to familiarize attendees with a new generation of Radiology specific search and decision support tools that are geared towards providing high quality Radiology-centric search results to the desktop with a minimum of effort by the user. Novel presentation methods which aggregate information from multiple sources (ie. mashups) in a single portal will also be featured. This format allows users to consume content more efficiently than through conventional means.

DATE: Sunday, November 30, 2008
START TIME: 12:15 PM
END TIME: 01:45 PM
LOCATION: S501ABC
Click here for more info

Moving toward a Hospital-wide HL7-based Semantic Search Engine for Patient Information

Our hospital-wide patient information search engine was successful in indexing radiology reports as well as extracting semantic metadata from the content. Such Web-based tools can provide decision support at the point of care and offer robust capabilities for data-mining and retrospective data retrieval.

DATE: Monday, December 01 2008
START TIME: 11:30 AM
END TIME: 11:40 AM
LOCATION: S403A
Click here for more info

SESSION: Continuing Medical Education in the Clinical Practice (Basic Imaging Informatics)
YottaLook™: How to Avoid Searching for Content in All the Wrong Places?

Yottalook is a Web search engine that allows imaging specialists to obtain radiology-specific content and images from both the Internet and within the enterprise (www.yottalook.com). In this refresher course we will demonstrates the use and applications of a medical imaging search engine that identifies and retrieves online peer-reviewed content (including images) not indexed by generic search engines. Users will also learn how Yottalook enhances the search experience by analyzing user queries and modifying them in real time with RadLex-driven ontologies.

DATE: Thursday, December 04, 2008
START TIME: 08:30 AM
END TIME: 10:00 AM
LOCATION: S403A
Click here for more info

New Yottalook Journals Engine

With the newest update, we have introduced beta version of a new search engine, "Journals". Our new Journals engine allows users to search within radiology journals. Similar to our other engines, we have allowed options to filter by imaging modalities. Additional special filters exist, including "Review Articles", "Related Articles" and "Repeat Search in This Journal Only" options. We have used same query expansion technique employed in other Yottalook engines to increase relevance of search results. The "More..." button allows users ability to view the entire abstract without having to click on the search result.

What if you do not want to limit your search just to the radiology journals? Right underneath the search box, you will notice "Radiology Only" as well as "Search All" options. Radiology Only is selected by default, but one can click on "Search All" to expand search to include non-radiology journals as well.

We hope users will find this new search engine useful. Please feel free to provide your feedback as we want to constantly improve this beta engine.

New Revised Yottalook!

I haven't blogged for a while, but we have been very busy this year working on many new features; the updates throughout the year have been small. Last night, we did a major update to include many new features we have been working on this year.

The first thing you will notice is our new design. We have returned to a more simple layout.

You will also notice that we have added a brand new engine: "Journals". This is a new engine that will allow users to search journal articles. It will give you options to search either just the radiology journals or all the other journals. Think of it as our Google Scholar-like engine. However, this engine is not based on Google, allowing us more control over the search results. I will blog about this new engine separately later.

The Images engine has been updated to include now more than 750,000 images. It contains new radiology images upto November of this year. Due to the update, you will notice that a lot of the results will say "Subscription May Be Required". This is our warning to the user that the journal content shown likely requires a subscription to view the full content. If you are using this engine and either have a subscription to that particular journal or are coming from an academic site, you shouldn't have issue accessing the content. However, many of you do not have subscriptions to these journals. While we want to give higher ranking to more recent content, we are sensitive to this reality, and as a result, we have implemented a new filter called, "No Subscription Mode". By turning this on, the engine will try to maximize results that do not require subscription first; you will still see some but the result sets should be more optimal for those desiring this option. More changes have been made to the Images engine, but again I will blog about this separately later.

Yottalook Anatomy: New Changes & Tips

This past month, we have been continuing to improve the current beta Yottalook Anatomy engine. You can access this engine either in our Yottalook Labs or directly by typing in http://anatomy.yottalook.com.

One of the common search queries for radiologists is related to anatomy. This engine was designed to provide anatomy-related information. Below is a set of revised refinements.

The engine will return results first from various online atlases. Hence, if search for "knee" and then select the "AP/PA" refinement, you will get a result that would show you a labeled radiograph image of a frontal view of the knee. By default, the online atlas results will show up first.

However, some of you may be looking for radiology journal articles on anatomy. Then, you can try to refine your search further by selecting "Search Radiology Journals >>". For example, type in "deep transverse metatarsal ligament" and then select "Search Radiology Journals >>" and then click on "Images Only" to get a drawing of it in a radiology journal.

Or, if you are just looking for detailed anatomy information from a classic anatomy textbook like Gray's Anatomy, simply click on "Search Gray's Anatomy".

Give it a try today and give us feedback on how to improve this engine.

Yottalook Images Improvements

I know we at Yottalook have been rather quiet this past month in terms of blog entries, but it is because we have been extremely busy working on new features. While many of the features are still being tested internally, I wanted share with you some of the new features that have been added this past month. This particular blog entry will describe some of the new features added to the Images engine.

First, we have expanded the exclusion options to exclude not only "Charts/Graphs" but also "Histo" for histology as this has been a feature that has been requested by some of the users.

Second, now the users have two options of displaying their results. You will notice the new set of options below the search box called "Individual Fig" and "Grouped Fig". The first option is selected by default. The second option "Grouped Fig" is a new option when selected returns not one but all the images from a particular article. This option is useful if you are looking for a series of images from a particular article so that you don't have to click on individual result links. Below is a graphical representation of how the returned results would differ. You can see this by clicking on the "What's this?" link next to the options. This feature is also in response to various users' request for a more grouped results. We hope you will find these features useful and potentially speed your browsing. If you have any suggestions, please provide your feedback.

Yottalook with New Interface

We have updated our Yottalook interface. The "References" engine has been renamed as "Web" and has been made the default engine. This change was made as what is now the "Web" engine has a broader coverage of radiology-related web content than the "Images" engine. You can still access the "Images" engine through the hyperlinked-interface both on the main page as well as the results pages.

If you click on the "Images" engine without entering a search query, it will take you to the "Images" main page. If you click on the "Images" engine after you type in the query term, you will be taken directly to the "Images" results page. The benefit of the new hyperlinked-interface is that it allows users to switch between the two engines without having to retype in the search term and quickly view the results from different engines.

Exclude Charts and Graphs on Yottalook Images

Today we added a feature to exclude charts and graphs from the Yottalook Images results. By using a certain set of terms, checking the “Charts&Graphs” box will attempt to exclude charts and graphs of various types. It is a temporary solution while we work on a more robust method of excluding certain set of images.

Yottalook with Improved User Interface

While radio buttons work well for selection of which search engine to use from the Yottalook homepage, its use and function on the results page have been less than optimal with the Reference engine always being selected by default regardless of the engine used to get to the results page. Today, we announce improved user interface that addresses this issue. Now, you will be directed to the specific results page for the search engine used. In addition, with the other engine hyperlinked on top of the search box, the user can easily switch back and forth between the Images and References engines and see the results change without having to retype in the search term(s) or pressing Enter/Search button following selection of the alternate engine.

In addition, we have launched a separate advanced search option for the Images engine. From the main page, selection of "Advanced Search" will open up different Advanced Search page depending on which engine has been selected (via radio button).

Yottalook - Wikipedia, the free encyclopedia

Yottalook - Wikipedia, the free encyclopedia

Woojin created a Wikipedia entry for Yottalook today. Our focus is to educate the community about the need for vertical search engine for specific health care domain and also address some of the information needs for radiologists and trainees in clinical practice today.

Help us refine the description. I hope we can make the Wikipedia entry look more like an independ neutral review rather than an advertisment.

Yottalook Goes Mobile

A lot of users have been asking for compatibility of Yottalook with mobile devices and PDAs, so we are releasing an alpha version of Yottalook Mobile (http://mobile.yottalook.com) for user feedback and comments.

So go ahead and give us comments, report bugs or any other issues that you may have with this. Remember, when you click on a result, you are directed away from Yottalook.com and we have no control over the look of the resulting site.

Please report any sites that you feel completely fail on mobile devices. Point your PDAs and mobile devices to http://mobile.yottalook.com and enjoy.

Yottalook Anatomy - Revised Refinements

Today, we are releasing new refinements for the Yottalook Anatomy search engine (www.yottalook.com/labs/yottalook_anat.html). First, you will notice a new refinement called, "Atlas >>". Selection of this "label" will direct the user to another set of label hierarchy, which includes modality and axis plane. What's the purpose? One of the original purposes of creating a radiology-centric anatomy search engine was to serve as a source of various online atlases. We wanted to separate out the atlases from other anatomy-related sites and journal articles. Now, selection of "Atlas >>" will only link you to online atlases. For example, type in "knee" and do a search. Then, click on "Atlas >>". Once you have been directed to the next level of labels, select "AP/PA" and click on any of the returned results.

In addition, similar to the current refinements in the Yottalook References engine, you will find "Google Scholar" and "Not enough results?" labels, which will search through Google Scholar and the entire web, respectively. Also, similar to the current References engine, we have implemented a checkbox option for returning only articles in PDF format.

Yottalook Anatomy Update - Get Radiological Anatomy Content From Select Journals

Following the release of Yottalook Labs this past week, I had one of the residents test out the Yottalook Anatomy engine (www.yottalook.com/labs/yottalook_anat.html). The first anatomical structure he wanted to look for was "Artery of Adamkiewicz". After couple attempts at finding the correct spelling, we found out that the engine failed to return any results. Surprised, I broadened the search by looking up "Artery of Adamkiewicz" on Google Images and snap. I still was not satisfied with the search results. When I did a web search for the term, I got results from sites such as Wikipedia and Who Named It. As a radiologist, when I type in "Artery of Adamkiewicz", I expect to find either information about that anatomical structure and/or an image of it. This little experience has led to revising the engine. Since this past week, we have tweaked our engine to search only for anatomy-related information from select radiology journals. The reasons for limiting the number of journals covered were two fold: 1) to maximize relevancy for those in the field of radiology and 2) to test out this new algorithm prior to increasing the number of journals covered. As a result, now when you type in "Adamkiwiecz", you will get results on this structure that is (hopefully) more relevant to radiologists.

Yottametta Adds 7 More Journals

Yottametta is one of the Yottalook Labs projects (www.yottalook.com/labs/yottametta.html). This week we added following journals in its coverage: Chest and AHA journals, which include Arteriosclerosis, Thrombosis and Vascular Biology, Circulation, Circulation Research, Hypertension, and Stroke. As of today, the Chest journal contains over 11,000 and AHA journals all together contain over 169,000 figures and tables. Not all the images are radiology-related images, but addition of these journals gives radiologists access to even more images from peer-reviewed sources from a single site. The users should understand one difference with these journals when used with Yottametta. When you click on the tab after entering the search term(s), you will notice that a new browser window opens up with results. This is not a bug but rather complying with the restrictions set by the journal sites.

Along with the new additions, we have fixed the restriction that forced the user to press the button to initiate the search. Now the search will initiate by either pressing "Enter" or by pressing the "Search" button following entering the search term(s).

Yottalook Images Search Engine Expands Coverage

Yottalook has expanded its journal coverage. When we first began our beta testing of the Yottalook Images search engine (www.yottalook.com) back in mid-December, we initially started out by searching through radiology journals only. We have been gradually increasing the number of radiology journals covered. Since about three weeks ago, we have started to include non-radiology journals as well. However, the algorithm for this is trickier as typically non-radiology journals contain many figures that are not directly relevant to the field of radiology. We decided to test out our new algorithm by searching through The New England Journal of Medicine (NEJM). We silently rolled out this new addition little less than 2 weeks ago. If you would like to see a demonstration of its implementation, type in "ectopic calcinosis" into our Images search engine.

After we were satisfied with the results we were getting with the NEJM, we have since increased our coverage, including journals like JAMA and Annals of Internal Medicine, among several others. To test out its implementation, go ahead and type in terms like "AIDS brain", "anthrax", and "duodenal gastrinoma". Unfortunately, not all radiological images from these various journals have been indexed by Google as of date. As a result, the search results through these journals are less than complete, but we do provide those radiology-related figures that have been indexed by Google and will provide more as they become available.