ImageCLEF 2018 VQA-Med

Important note:

The ImageCLEF 2018 VQA-Med challenge has officially ended and we would like to thank everyone for their participation. The official results are already emailed to corresponding participants.

Post-challenge submissions and the leaderboard will remain enabled for a few weeks so you will still be able to submit result files and have them continuously evaluated during a limited period. Please consider that in order to see the version of the leaderboard with the post-challenge submissions integrated, you have to turn on the switch Show post-challenge submission right below the leaderboard.

At the same time we’d like to encourage you to submit a CLEF Working notes paper until the end of May.

Please also note that participants registering from now on will not be automatically registered with CLEF anymore.

Note: Do not forget to read the Rules section on this page

Motivation

With the increasing interest in artificial intelligence (AI) to support clinical decision making and improve patient engagement, opportunities to generate and leverage algorithms for automated medical image interpretation are currently being explored. Since patients may now access structured and unstructured data related to their healthcare utilization via patient portals, such access also motivates the need to help them better understand their conditions in line their available data, including medical images.

Clinicians’ confidence in interpreting complex medical images can be significantly enhanced by “second opinion” provided by an automated system. In addition, patients may be interested in the morphology/physiology and disease-status of anatomical structures around a lesion that has been well characterized by their healthcare providers – and they may not necessarily be willing to pay significant amounts for a separate office- or hospital visit just to address such questions. Although patients often turn to search engines (e.g. Google) to disambiguate complex terms or obtain answers to confusing aspects of the medical image, results from search engines may be nonspecific, erroneous and misleading, or overwhelming in terms of the volume of information.

Challenge description

Visual Question Answering is a new and exciting problem that combines natural language processing and computer vision techniques. Inspired by the recent success of visual question answering in the general domain , we propose a pilot task this year to focus on visual question answering in the medical domain. Given a medical image accompanied with a set of clinically relevant questions, participating systems are tasked with answering the questions based on the visual image content.

Data

The data will tentatively include a training set (5K) and a validation set (0.5K) with medical images accompanied with question-answer pairs, and a test set (0.5K) of images with questions only. To create the datasets for the proposed task, we would consider the medical domain images extracted from PubMed articles (essentially a subset of the ImageCLEF 2017 caption prediction task).

Submission instructions

As soon as the submission is open, you will find a “Create Submission” button on this page (just next to the tabs)

-Each team is allowed to submit a maximum of 5 runs.

-We expect the following format for the result submission file: <QA-ID><TAB><Image-ID><TAB><Answer>

For example:

1 rjv03401 answer of the first question in one single line
2 AIAN-14-313-g002 answer of the second question
3 wjem-11-76f3 answer of the third question

-You need to respect the following constraints:

• The separator between <QA-ID>, <Image-ID> and <Answer> has to be a tabular white space (tab).
• Each <QA-ID> of the test set must be included in the run file exactly once.
• You should not include special characters in the <Answer> field.
• All 500 <QA-ID> and <Image-ID> pairs must be present in a participant’s run file in the same order as the VQAMed2018Test-QA.csv file.

-Participants are allowed to use other resources asides from the official training/validation datasets, however the use of the additional resources must have to be explicitly stated. For meaningful comparison, we will separately group systems who exclusively use the official training data and who incorporate additional sources.

Please provide the necessary information and select a submission file. As soon as a submission file is selected the form is submitted automatically. After the submission of the form the grading process will be initiated where an external grader validates/evaluates the submitted file and eventually returns the score back to CrowdAI. Depending on the file size, the evaluation algorithm and the total grading workload this could take a while. You can see the status of your submission in the “Submissions” tab of this challenge’s page, where you will redirected to automatically after having submitted. In case the evaluation failed, the “Status” field shows “failed” and an error message in the “Message” field is displayed.

Acknowledgements

PubMed Central

Evaluation criteria

Information will be posted soon.

Resources

Contact us

• Technical issues : https://gitter.im/crowdAI/imageclef-2018-vqa-med
• Discussion Forum : https://www.crowdai.org/challenges/imageclef-2018-vqa-med/topics

We strongly encourage you to use the public channels mentioned above for communications between the participants and the organizers. In extreme cases, if there are any queries or comments that you would like to make using a private communication channel, then you can send us an email at :

• Sharada Prasanna Mohanty: sharada.mohanty@epfl.ch
• Sadid Hasan: sadid[DOT]hasan[AT]philips[DOT]com
• Yuan Ling: yuan[DOT]ling[AT]philips[DOT]com
• Henning Müller: henning[DOT]mueller[AT]hevs[DOT]ch
• Ivan Eggel: ivan[DOT]eggel[AT]hevs[DOT]ch

More information

You can find additional information on the challenge here: http://imageclef.org/2018/VQA-Med

Prizes

ImageCLEF 2018 is an evaluation campaign that is being organized as part of the CLEF initiative labs. The campaign offers several research tasks that welcome participation from teams around the world. The results of the campaign appear in the working notes proceedings, published by CEUR Workshop Proceedings (CEUR-WS.org). Selected contributions among the participants, will be invited for publication in the following year in the Springer Lecture Notes in Computer Science (LNCS) together with the annual lab overviews.

Datasets License