You are here

ImageCLEFmed Caption

Welcome to the 3rd edition of the Caption Task!



Interpreting and summarizing the insights gained from medical images such as radiology output is a time-consuming task that involves highly trained experts and often represents a bottleneck in clinical diagnosis pipelines.

Consequently, there is a considerable need for automatic methods that can approximate this mapping from visual information to condensed textual descriptions. The more image characteristics are known, the more structured are the radiology scans and hence, the more efficient are the radiologists regarding interpretation. We work on the basis of a large-scale collection of figures from open access biomedical journal articles (PubMed Central). All images in the training data are accompanied by UMLS concepts extracted from the original image caption. 

Lessons learned:

  • In the first and second editions of this task, held at ImageCLEF 2017 and ImageCLEF 2018, participants noted a broad variety of content and situation among training images. For this year, the training data is reduced solely to radiology images
  • A large number of concepts was used in the previous years. This year, the captions are first processed before concept extraction, hence leading to a reduced number of concepts
  • As uncertainty regarding additional source was noted, we will clearly separate systems using exclusively the official training data from those that incorporate additional sources of evidence


  • 08.11.2018: website goes live
  • 25.11.2018: registration open at crowdAI
  • 15.01.2019: development data released at crowdAI
  • 18.03.2019: test data released at crowdAI
  • 29.04.2019: updated
  • 29.04.2019: submission deadline extended to 06.05.2019 (AoE)
  • 19.05.2019: working notes instruction added
  • 23.05.2019: working notes deadline extended to 27.05.2019 (UTC)

Task Description

Concept Detection Task

The first step to automatic image captioning and scene understanding is identifying the presence and location of relevant concepts in a large corpus of medical images. Based on the visual image content, this subtask provides the building blocks for the scene understanding step by identifying the individual components from which captions are composed. The concepts can be further applied for context-based image and information retrieval purposes.

Evaluation is conducted in terms of set coverage metrics such as precision, recall, and combinations thereof. This task will be run using a subset of the Radiology Objects in COntext (ROCO) dataset [1]. 


From the PubMed Open Access subset containing 1,828,575 archives, a total number of 6,031,814
image - caption pairs were extracted. To focus on radiology images and non-compound figures, automatic filtering with deep learning systems as well as manual revisions were applied, reducing the dataset to 70,786 radiology images of several medical imaging modalities.

NOTE: If the usage of an additional source for training is intended, it should not be a subset of PubMed Central Open Access (archiving date: 01.02.2018 - 01.02.2019), to avoid an overlap with the test data.

Evaluation Methodology

Evaluation is conducted in terms of F1 scores between system predicted and ground truth concepts, using the following methodology and parameters:

  • The default implementation of the Python scikit-learn (v0.17.1-2) F1 scoring method is used. It is documented here.
  • A Python (3.x) script loads the candidate run file, as well as the ground truth (GT) file, and processes each candidate-GT concept sets
  • For each candidate-GT concept set, the y_pred and y_true arrays are generated. They are binary arrays indicating for each concept contained in both candidate and GT set if it is present (1) or not (0).
  • The F1 score is then calculated. The default 'binary' averaging method is used.
  • All F1 scores are summed and averaged over the number of elements in the test set (10'000), giving the final score.

The ground truth for the test set was generated based on the UMLS Full Release 2017AB.

NOTE: The source code of the evaluation tool is available here. It must be executed using Python 3.x, on a system where the scikit-learn (>= v0.17.1-2) Python library is installed. The script should be run like this:

/path/to/python3 /path/to/candidate/file /path/to/ground-truth/file

Preliminary Schedule

  • 05.11.2018: Registration opens for all ImageCLEF tasks (until 26.04.2019)
  • 15.01.2019: Development data release starts
  • 18.03.2019: Test data release starts
  • 01.05.2019 06.05.2019: Deadline for submitting the participants runs
  • 13.05.2019: Release of the processed results by the task organizers
  • 24.05.2019 27.05.2019: Deadline for submission of working notes papers by the participants
  • 07.06.2019: Notification of acceptance of the working notes papers
  • 28.06.2019: Camera-ready working notes papers
  • 09-12.09.2019CLEF 2019, Lugano, Switzerland

Participant Registration

CrowdAI is shutting down and will move towards AICrowd. Please temporarily ignore the information below this paragraph. During the transition phase (until all challenges are migrated) we will have to provide the datasets and End User Agreement (EUA) handling ourselves. In order to get access to the dataset, please download the EUA at the bottom of this page and send a filled in and signed version to henning.mueller[at-character] Please refer to the ImageCLEF registration instructions to get some examples on how to fill in the EUA.

Please refer to the general ImageCLEF registration instructions

Submission Instructions

Please note that each group is allowed a maximum of 10 runs per subtask.

For the submission of the concept detection task we expect the following format:

  • <Figure-ID><TAB><Concept-ID-1>;<Concept-ID-2>;<Concept-ID-n>


  • ROCO_CLEF_41341 C0033785;C0035561
  • ROCO_CLEF_07563 C0043299;C1306645;C1548003;C1962945

You need to respect the following constraints:

  • The separator between the figure ID and the concepts has to be a tabular whitespace
  • The separator between the UMLS concepts has to be a semicolon (;)
  • Each figure ID of the test set must be included in the submitted file exactly once (even if there are not concepts)
  • The same concept cannot be specified more than once for a given figure ID
  • The maximum number of concepts per image is 100


Group Name Submission Run F1 Score Rank
AUEB NLP Group s2_results.csv 0.2823094 1
AUEB NLP Group ensemble_avg.csv 0.2792511 2
AUEB NLP Group s1_results.csv 0.2740204 3
damo test_cat_xi.txt 0.2655099 4
AUEB NLP Group s3_results.csv 0.2639952 5
damo test_results.txt 0.2613895 6
damo first_concepts_detection_result_check.txt 0.2316484 7
ImageSem F1TOP1.txt 0.2235690 8
ImageSem F1TOP2.txt 0.2227917 9
ImageSem F1TOP5_Pmax.txt 0.2216225 10
ImageSem F1TOP3.txt 0.2190201 11
ImageSem 07Comb_F1Top1.txt 0.2187337 12
ImageSem F1TOP5_Rmax.txt 0.2147437 13
damo test_tran_all.txt 0.2134523 14
damo test_cat.txt 0.2116252 15
UA.PT_Bioinformatics simplenet.csv 0.2058640 16
richard_ycli testing_result.txt 0.1952310 17
ImageSem 08Comb_Pmax.txt 0.1912173 18
UA.PT_Bioinformatics simplenet128x128.csv 0.1893430 19
UA.PT_Bioinformatics mix-1100-o0-2019-05-06_1311.csv 0.1825418 20
UA.PT_Bioinformatics aae-1100-o0-2019-05-02_1509.csv 0.1760092 21
Sam Maksoud TRIAL_1.txt 0.1749349 22
richard_ycli testing_result.txt 0.1737527 23
UA.PT_Bioinformatics ae-1100-o0-2019-05-02_1453.csv 0.1715210 24
UA.PT_Bioinformatics cedd-1100-o0-2019-05-03_0937-trim.csv 0.1667884 25
AI600 ai600_result_weighing_1557061479.txt 0.1656261 26
Sam Maksoud TRIAL_18.txt 0.1640647 27
richard_ycli testing_result_run4.txt 0.1633958 28
AI600 ai600_result_weighing_1557059794.txt 0.1628424 29
richard_ycli testing_result_run3.txt 0.1605645 30
AI600 ai600_result_weighing_1557107054.txt 0.1603341 31
AI600 ai600_result_weighing_1557062212.txt 0.1588862 32
AI600 ai600_result_weighing_1557062494.txt 0.1562828 33
AI600 ai600_result_weighing_1557107838.txt 0.1511505 34
richard_ycli testing_result_run2.txt 0.1467212 35
MacUni-CSIRO run1FinalOutput.txt 0.1435435 36
AI600 ai600_result_rgb_1556989393.txt 0.1345022 37
UA.PT_Bioinformatics simplenet64x64.csv 0.1279909 38
UA.PT_Bioinformatics resnet19-cnn.csv 0.1269521 39
ImageSem 09Comb_Rmax_new.txt 0.1121941 40
damo test_att_3_rl_best.txt 0.0590448 41
damo test_rl_5_result_check.txt 0.0584684 42
damo test_tran_rl_5.txt 0.0567311 43
damo test_tran_10.txt 0.0536554 44
pri2si17 submission_1.csv 0.0496821 45
AILAB results_v3.txt 0.0202243 46
AILAB results_v1.txt 0.0198960 47
AILAB results_v2.txt 0.0162458 48
pri2si17 submission_3.csv 0.0141422 49
AILAB results_v4.txt 0.0126845 50
LIST denseNet_pred_all_0.55.txt 0.0013269 51
ImageSem yu_1000_inception_v3_top6.csv 0.0009450 52
ImageSem yu_1000_resnet_152_top6.csv 0.0008925 53
LIST denseNet_pred_all_0.6.txt 0.0003665 54
LIST denseNet_pred_all.txt 0.0003400 55
LIST predictionBR(LR).txt 0.0002705 56
LIST denseNet_pred_all_0.6_50_0.04(max if null).txt 0.0002514 57
LIST predictionCC(LR).txt 0.0002494 58
AILAB results_v0.txt 0 -
pri2si17 submission_2.csv 0 -

CEUR Working Notes

  • All participating teams with atleast one graded submission, regardless of F1 score, should submit a CEUR working notes paper.
  • The working notes paper should be submitted using this link:
    Click on "enter as an author", then select track "ImageCLEF - Multimedia Retrieval in CLEF".
    Add author information, paper title/abstract, keywords, select "Task 3 - ImageCLEFmedical" and upload your working notes paper as pdf.


When referring to the ImageCLEFmed 2019 concept detection task general goals, general results, etc. please cite the following publication which will be published by September 2019:

  • Obioma Pelka, Christoph M. Friedrich, Alba García Seco de Herrera and Henning Müller. Overview of the ImageCLEFmed 2019 Concept Detection Task, CLEF working notes, CEUR, 2019.
  • BibTex:
    author = {Pelka, Obioma and Friedrich, Christoph M and Garc\'ia Seco de Herrera, Alba and M\"uller, Henning},
    title = {Overview of the {ImageCLEFmed} 2019 Concept Prediction Task},
    booktitle = {CLEF2019 Working Notes},
    series = {{CEUR} Workshop Proceedings},
    year = {2019},
    volume = {},
    publisher = { $$},
    pages = {},
    month = {September 09-12},
    address = {Lugano, Switzerland},

When referring to the ImageCLEF 2019 task in general, please cite the following publication to be published by September 2019:

  • BibTex:
    author = {Bogdan Ionescu and Henning M\"uller and Renaud P\'{e}teri
    and Yashin Dicente Cid and Vitali Liauchuk and Vassili Kovalev and
    Dzmitri Klimuk and Aleh Tarasau and Asma Ben Abacha and Sadid A. Hasan
    and Vivek Datla and Joey Liu and Dina Demner-Fushman and Duc-Tien
    Dang-Nguyen and Luca Piras and Michael Riegler and Minh-Triet Tran and
    Mathias Lux and Cathal Gurrin and Obioma Pelka and Christoph M.
    Friedrich and Alba Garc\'ia Seco de Herrera and Narciso Garcia and
    Ergina Kavallieratou and Carlos Roberto del Blanco and Carlos Cuevas
    Rodr\'{i}guez and Nikos Vasillopoulos and Konstantinos Karampidis and
    Jon Chamberlain and Adrian Clark and Antonio Campello},
    title={{ImageCLEF 2019}: Multimedia Retrieval in Medicine, Lifelogging, Security and Nature},
    booktitle={Experimental IR Meets Multilinguality, Multimodality, and Interaction},
    series = {Proceedings of the Tenth International Conference of the CLEF Association (CLEF 2019)},
    year = {2019},
    volume = {},
    publisher = {{LNCS} Lecture Notes in Computer Science, Springer},
    pages = {},
    month = {September 09-12},
    address = {Lugano, Switzerland},}


  • Obioma Pelka <obioma.pelka(at)>, University of Applied Sciences and Arts Dortmund, Germany
  • Christoph M. Friedrich <christoph.friedrich(at)>, University of Applied Sciences and Arts Dortmund, Germany
  • Alba García Seco de Herrera <alba.garcia(at)>,University of Essex, UK
  • Henning Müller <henning.mueller(at)>, University of Applied Sciences Western Switzerland, Sierre, Switzerland

Join our mailing list:
Follow @imageclef


[1] O. Pelka, S. Koitka, J. Rückert, F. Nensa und C. M. Friedrich „Radiology Objects in COntext (ROCO): A Multimodal Image Dataset“, Proceedings of the MICCAI Workshop on Large-scale Annotation of Biomedical data and Expert Label Synthesis (MICCAI LABELS 2018), Granada, Spain, September 16, 2018, Lecture Notes in Computer Science (LNCS) Volume 11043, Page 180-189, DOI: 10.1007/978-3-030-01364-6_20, Springer Verlag, 2018.