What are the best rich metadata fields for a Microscopy Image Database?
Locked
 
|
Praju Vikas Anekal |
What are the best rich metadata fields for a Microscopy Image Database?
|
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy Post images on http://www.imgur.com and include the link in your posting. ***** Dear Confocal Microscopy Listers, We are looking at setting up an Image repository Database here at the University of Auckland in conjunction with our Centre for eResearch (CeR). In addition to the detailed image and instrument metadata, we are also trying to capture more of the sematic and ontological metadata associated with the images. The objective in capturing this rich metadata is to make the Image DB a searchable resource. (“A Bioimage Data Integration and Publication Platform: The Image Data Resource. Williams et al. Nat Methods. 2017 Aug;14(8)” has been extremely helpful and informative in guiding us) Moreover, such a well-annotated image DB can act as a useful source for training datasets in machine learning efforts down the line. There appears to be a fine balance between capturing enough metadata versus capturing too much as to make it too arduous or too granular as to be counterproductive. In addition, there is also the question of how to collect this metadata. Having to enter lots of fields that need to be entered manually will be counterproductive. So we are also thinking of ideas on how to capture the metadata in the easiest way possible. We are setting up a discussion group among our researchers here to try and synthesis a balanced list of metadata fields that will prove useful (now and in future). I was hoping to also reach out to the (fantastic!) contributors to the Confocal Microscopy list to ‘crowd source’ ideas and suggestions about this. I’ll briefly list out the types of fields we have currently considered and I would appreciate any suggestions about these and others we may consider. Image Metadata • Unique image ID • Dimensional info (x/y/z pixel size, wavelength, time, bitrate etc) • Positional info (ROIs etc) • File info (compression, file formats etc) Equipment Metadata • Make /model/ Version • Instrument setting • Hardware configuration (filters, objectives, NA) User Metadata • User/PI/Dept • Linked Grant/Project/Papers Experimental Metadata • Project ID • Experiment type (disease vs control, knockdown vs control etc) • Cell/Tissue type • Fixation, thickness • Antibodies used? Concentrations Semantic/Ontological Metadata • Research area ontology • Phenotype Ontology • Image quality score (1-10 user scored) • Image processing steps • Linked Segmentation masks Thank you Yours sincerely, Praju Vikas Anekal. Ph.D. Biomed Imaging Microscopist/BioImage Analyst, Biomedical Imaging Research Unit. Faculty of Medical and Health Sciences, The University of Auckland. E-Mail : [hidden email] , Ext : 87831 |
|
|
Cammer, Michael-2 |
Re: What are the best rich metadata fields for a Microscopy Image Database?
|
|
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy Post images on http://www.imgur.com and include the link in your posting. ***** I recommend speaking with the designers of the Cell Image Library. They went very deep into data associated with images. http://www.cellimagelibrary.org/home Cheers- Michael Cammer -----Original Message----- From: Confocal Microscopy List <[hidden email]> On Behalf Of Praju Vikas Sent: Tuesday, May 4, 2021 11:27 PM To: [hidden email] Subject: What are the best rich metadata fields for a Microscopy Image Database? [EXTERNAL] ***** To join, leave or search the confocal microscopy listserv, go to: https://urldefense.com/v3/__http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJPI_OTk0$ Post images on https://urldefense.com/v3/__http://www.imgur.com__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJHHpMSY8$ and include the link in your posting. ***** Dear Confocal Microscopy Listers, We are looking at setting up an Image repository Database here at the University of Auckland in conjunction with our Centre for eResearch (CeR). In addition to the detailed image and instrument metadata, we are also trying to capture more of the sematic and ontological metadata associated with the images. The objective in capturing this rich metadata is to make the Image DB a searchable resource. (“A Bioimage Data Integration and Publication Platform: The Image Data Resource. Williams et al. Nat Methods. 2017 Aug;14(8)” has been extremely helpful and informative in guiding us) Moreover, such a well-annotated image DB can act as a useful source for training datasets in machine learning efforts down the line. There appears to be a fine balance between capturing enough metadata versus capturing too much as to make it too arduous or too granular as to be counterproductive. In addition, there is also the question of how to collect this metadata. Having to enter lots of fields that need to be entered manually will be counterproductive. So we are also thinking of ideas on how to capture the metadata in the easiest way possible. We are setting up a discussion group among our researchers here to try and synthesis a balanced list of metadata fields that will prove useful (now and in future). I was hoping to also reach out to the (fantastic!) contributors to the Confocal Microscopy list to ‘crowd source’ ideas and suggestions about this. I’ll briefly list out the types of fields we have currently considered and I would appreciate any suggestions about these and others we may consider. Image Metadata • Unique image ID • Dimensional info (x/y/z pixel size, wavelength, time, bitrate etc) • Positional info (ROIs etc) • File info (compression, file formats etc) Equipment Metadata • Make /model/ Version • Instrument setting • Hardware configuration (filters, objectives, NA) User Metadata • User/PI/Dept • Linked Grant/Project/Papers Experimental Metadata • Project ID • Experiment type (disease vs control, knockdown vs control etc) • Cell/Tissue type • Fixation, thickness • Antibodies used? Concentrations Semantic/Ontological Metadata • Research area ontology • Phenotype Ontology • Image quality score (1-10 user scored) • Image processing steps • Linked Segmentation masks Thank you Yours sincerely, Praju Vikas Anekal. Ph.D. Biomed Imaging Microscopist/BioImage Analyst, Biomedical Imaging Research Unit. Faculty of Medical and Health Sciences, The University of Auckland. E-Mail : [hidden email] , Ext : 87831 |
|
|
Claire Brown |
Re: What are the best rich metadata fields for a Microscopy Image Database?
|
|
In reply to this post by Praju Vikas Anekal
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy Post images on http://www.imgur.com and include the link in your posting. ***** You timing is great! You must look at this work lead by Caterina Strambio De Castillia as part of BINA, 4DN and QUAREP. https://www.biorxiv.org/content/10.1101/2021.04.25.441198v2 Tools are being built to collect MetaData as well. I'm sure Caterina and others would be happy to talk with you. Claire |
|
|
Stefanie WP |
AW: What are the best rich metadata fields for a Microscopy Image Database?
|
|
In reply to this post by Cammer, Michael-2
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy Post images on http://www.imgur.com and include the link in your posting. ***** Indeed, great timing for this question. A lot of activity in this area for many good reasons! Please also look here: https://arxiv.org/abs/2103.02942 cheers, Stefanie Dr. Stefanie Weidtkamp-Peters Center for Advanced Imaging German BioImaging – GMB e.V. Heinrich-Heine University Duesseldorf - Chair of Board - Building 26.24.00.011 www.germanbioimaging.org Universitaetsstr. 1, 40225 Duesseldorf Phone: +49-211-81-11682 Email: [hidden email] www.cai.hhu.de -----Ursprüngliche Nachricht----- Von: Confocal Microscopy List <[hidden email]> Im Auftrag von Cammer, Michael Gesendet: Mittwoch, 5. Mai 2021 15:51 An: [hidden email] Betreff: Re: What are the best rich metadata fields for a Microscopy Image Database? ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy Post images on http://www.imgur.com and include the link in your posting. ***** I recommend speaking with the designers of the Cell Image Library. They went very deep into data associated with images. http://www.cellimagelibrary.org/home Cheers- Michael Cammer -----Original Message----- From: Confocal Microscopy List <[hidden email]> On Behalf Of Praju Vikas Sent: Tuesday, May 4, 2021 11:27 PM To: [hidden email] Subject: What are the best rich metadata fields for a Microscopy Image Database? [EXTERNAL] ***** To join, leave or search the confocal microscopy listserv, go to: https://urldefense.com/v3/__http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJPI_OTk0$ Post images on https://urldefense.com/v3/__http://www.imgur.com__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJHHpMSY8$ and include the link in your posting. ***** Dear Confocal Microscopy Listers, We are looking at setting up an Image repository Database here at the University of Auckland in conjunction with our Centre for eResearch (CeR). In addition to the detailed image and instrument metadata, we are also trying to capture more of the sematic and ontological metadata associated with the images. The objective in capturing this rich metadata is to make the Image DB a searchable resource. (“A Bioimage Data Integration and Publication Platform: The Image Data Resource. Williams et al. Nat Methods. 2017 Aug;14(8)” has been extremely helpful and informative in guiding us) Moreover, such a well-annotated image DB can act as a useful source for training datasets in machine learning efforts down the line. There appears to be a fine balance between capturing enough metadata versus capturing too much as to make it too arduous or too granular as to be counterproductive. In addition, there is also the question of how to collect this metadata. Having to enter lots of fields that need to be entered manually will be counterproductive. So we are also thinking of ideas on how to capture the metadata in the easiest way possible. We are setting up a discussion group among our researchers here to try and synthesis a balanced list of metadata fields that will prove useful (now and in future). I was hoping to also reach out to the (fantastic!) contributors to the Confocal Microscopy list to ‘crowd source’ ideas and suggestions about this. I’ll briefly list out the types of fields we have currently considered and I would appreciate any suggestions about these and others we may consider. Image Metadata • Unique image ID • Dimensional info (x/y/z pixel size, wavelength, time, bitrate etc) • Positional info (ROIs etc) • File info (compression, file formats etc) Equipment Metadata • Make /model/ Version • Instrument setting • Hardware configuration (filters, objectives, NA) User Metadata • User/PI/Dept • Linked Grant/Project/Papers Experimental Metadata • Project ID • Experiment type (disease vs control, knockdown vs control etc) • Cell/Tissue type • Fixation, thickness • Antibodies used? Concentrations Semantic/Ontological Metadata • Research area ontology • Phenotype Ontology • Image quality score (1-10 user scored) • Image processing steps • Linked Segmentation masks Thank you Yours sincerely, Praju Vikas Anekal. Ph.D. Biomed Imaging Microscopist/BioImage Analyst, Biomedical Imaging Research Unit. Faculty of Medical and Health Sciences, The University of Auckland. E-Mail : [hidden email] , Ext : 87831 |
|
|
Strambio, Caterina |
Re: What are the best rich metadata fields for a Microscopy Image Database?
|
|
Praju Vikas Anekal |
Re: What are the best rich metadata fields for a Microscopy Image Database?
|
|
In reply to this post by Stefanie WP
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy Post images on http://www.imgur.com and include the link in your posting. ***** Hi Confocal Microscopy Listers, Thank you all for the valuable feedback and the kind offers to connect and discuss further. It is greatly appreciated and I’ll definitely follow up on this moving forward. With the wealth of material, our group will definitely have plenty to work with! There were a few specific points I wanted delve into further 1. One of the future goals of this (or any other DB) will be to act as ‘annotated’ training data for machine learning algorithms. Are there any metadata consideration for this aspect that we should pay particular attention to? Personally, I think this is where the rich semantic and ontological metadata will become useful. 2. The DB will attempt to incorporate both new and ‘legacy’ data. With legacy data, it will be nigh impossible to obtain all the metadata to populate these fields. Any thoughts on how to deal with ‘legacy’ data? 3. Also potentially some datasets may not able to provide all the required fields. We thought of having a metadata grading system…A for when all the metadata fields are populated, B when most are and C when only few are and so on… (that way, future search can gate metadata quality as well) Thank you again for your help! Yours sincerely, Praju Vikas Anekal. Ph.D. Biomed Imaging Microscopist/BioImage Analyst, Biomedical Imaging Research Unit. Faculty of Medical and Health Sciences, The University of Auckland. E-Mail : [hidden email]<mailto:[hidden email]> , Ext : 87831 From: Confocal Microscopy List <[hidden email]> On Behalf Of Weidtkamp-Peters, Stefanie Sent: Thursday, 6 May 2021 6:41 AM To: [hidden email] Subject: AW: What are the best rich metadata fields for a Microscopy Image Database? ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy> Post images on http://www.imgur.com<http://www.imgur.com> and include the link in your posting. ***** Indeed, great timing for this question. A lot of activity in this area for many good reasons! Please also look here: https://arxiv.org/abs/2103.02942<https://arxiv.org/abs/2103.02942> cheers, Stefanie Dr. Stefanie Weidtkamp-Peters Center for Advanced Imaging German BioImaging – GMB e.V. Heinrich-Heine University Duesseldorf - Chair of Board - Building 26.24.00.011 www.germanbioimaging.org<http://www.germanbioimaging.org> Universitaetsstr. 1, 40225 Duesseldorf Phone: +49-211-81-11682 Email: [hidden email]<mailto:[hidden email]> www.cai.hhu.de<http://www.cai.hhu.de> -----Ursprüngliche Nachricht----- Von: Confocal Microscopy List <[hidden email]<mailto:[hidden email]>> Im Auftrag von Cammer, Michael Gesendet: Mittwoch, 5. Mai 2021 15:51 An: [hidden email]<mailto:[hidden email]> Betreff: Re: What are the best rich metadata fields for a Microscopy Image Database? ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy> Post images on http://www.imgur.com<http://www.imgur.com> and include the link in your posting. ***** I recommend speaking with the designers of the Cell Image Library. They went very deep into data associated with images. http://www.cellimagelibrary.org/home<http://www.cellimagelibrary.org/home> Cheers- Michael Cammer -----Original Message----- From: Confocal Microscopy List <[hidden email]<mailto:[hidden email]>> On Behalf Of Praju Vikas Sent: Tuesday, May 4, 2021 11:27 PM To: [hidden email]<mailto:[hidden email]> Subject: What are the best rich metadata fields for a Microscopy Image Database? [EXTERNAL] ***** To join, leave or search the confocal microscopy listserv, go to: https://urldefense.com/v3/__http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJPI_OTk0$<https://urldefense.com/v3/__http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJPI_OTk0$> Post images on https://urldefense.com/v3/__http://www.imgur.com__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJHHpMSY8$<https://urldefense.com/v3/__http://www.imgur.com__;!!MXfaZl3l!PqgWpRTIU8NGSkoru5VpSU5Iv1axY1bD_Cy4Ctr8EZT-zRydv0pxzvdS4_8ed_pJHHpMSY8$> and include the link in your posting. ***** Dear Confocal Microscopy Listers, We are looking at setting up an Image repository Database here at the University of Auckland in conjunction with our Centre for eResearch (CeR). In addition to the detailed image and instrument metadata, we are also trying to capture more of the sematic and ontological metadata associated with the images. The objective in capturing this rich metadata is to make the Image DB a searchable resource. (“A Bioimage Data Integration and Publication Platform: The Image Data Resource. Williams et al. Nat Methods. 2017 Aug;14(8)” has been extremely helpful and informative in guiding us) Moreover, such a well-annotated image DB can act as a useful source for training datasets in machine learning efforts down the line. There appears to be a fine balance between capturing enough metadata versus capturing too much as to make it too arduous or too granular as to be counterproductive. In addition, there is also the question of how to collect this metadata. Having to enter lots of fields that need to be entered manually will be counterproductive. So we are also thinking of ideas on how to capture the metadata in the easiest way possible. We are setting up a discussion group among our researchers here to try and synthesis a balanced list of metadata fields that will prove useful (now and in future). I was hoping to also reach out to the (fantastic!) contributors to the Confocal Microscopy list to ‘crowd source’ ideas and suggestions about this. I’ll briefly list out the types of fields we have currently considered and I would appreciate any suggestions about these and others we may consider. Image Metadata • Unique image ID • Dimensional info (x/y/z pixel size, wavelength, time, bitrate etc) • Positional info (ROIs etc) • File info (compression, file formats etc) Equipment Metadata • Make /model/ Version • Instrument setting • Hardware configuration (filters, objectives, NA) User Metadata • User/PI/Dept • Linked Grant/Project/Papers Experimental Metadata • Project ID • Experiment type (disease vs control, knockdown vs control etc) • Cell/Tissue type • Fixation, thickness • Antibodies used? Concentrations Semantic/Ontological Metadata • Research area ontology • Phenotype Ontology • Image quality score (1-10 user scored) • Image processing steps • Linked Segmentation masks Thank you Yours sincerely, Praju Vikas Anekal. Ph.D. Biomed Imaging Microscopist/BioImage Analyst, Biomedical Imaging Research Unit. Faculty of Medical and Health Sciences, The University of Auckland. E-Mail : [hidden email]<mailto:[hidden email]> , Ext : 87831 |
«
Return to Confocal Microscopy List
|
1 view|%1 views
| Free forum by Nabble | Edit this page |