1
|
Habart D, Koza A, Leontovyc I, Kosinova L, Berkova Z, Kriz J, Zacharovova K, Brinkhof B, Cornelissen DJ, Magrane N, Bittenglova K, Capek M, Valecka J, Habartova A, Saudek F. IsletSwipe, a mobile platform for expert opinion exchange on islet graft images. Islets 2023; 15:2189873. [PMID: 36987915 PMCID: PMC10064927 DOI: 10.1080/19382014.2023.2189873] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 03/30/2023] Open
Abstract
We previously developed a deep learning-based web service (IsletNet) for an automated counting of isolated pancreatic islets. The neural network training is limited by the absent consensus on the ground truth annotations. Here, we present a platform (IsletSwipe) for an exchange of graphical opinions among experts to facilitate the consensus formation. The platform consists of a web interface and a mobile application. In a small pilot study, we demonstrate the functionalities and the use case scenarios of the platform. Nine experts from three centers validated the drawing tools, tested precision and consistency of the expert contour drawing, and evaluated user experience. Eight experts from two centers proceeded to evaluate additional images to demonstrate the following two use case scenarios. The Validation scenario involves an automated selection of images and islets for the expert scrutiny. It is scalable (more experts, images, and islets may readily be added) and can be applied to independent validation of islet contours from various sources. The Inquiry scenario serves the ground truth generating expert in seeking assistance from peers to achieve consensus on challenging cases during the preparation for IsletNet training. This scenario is limited to a small number of manually selected images and islets. The experts gained an opportunity to influence IsletNet training and to compare other experts' opinions with their own. The ground truth-generating expert obtained feedback for future IsletNet training. IsletSwipe is a suitable tool for the consensus finding. Experts from additional centers are welcome to participate.
Collapse
Affiliation(s)
- David Habart
- Laboratory of Pancreatic Islets, Center of Experimental Medicine, Institute for Clinical and Experimental Medicine (IKEM), Prague, Czech Republic
- CONTACT David Habart Laboratory of pancreatic islets, Center of Experimental Medicine, Institute for Clinical and Experimental Medicine, Videnska 1958/9, Prague 4, 140 21, Czech Republic
| | - Adam Koza
- Dino School & Novy PORG, Prague, Czech Republic
| | - Ivan Leontovyc
- Laboratory of Pancreatic Islets, Center of Experimental Medicine, Institute for Clinical and Experimental Medicine (IKEM), Prague, Czech Republic
| | - Lucie Kosinova
- Laboratory of Pancreatic Islets, Center of Experimental Medicine, Institute for Clinical and Experimental Medicine (IKEM), Prague, Czech Republic
| | - Zuzana Berkova
- Laboratory of Pancreatic Islets, Center of Experimental Medicine, Institute for Clinical and Experimental Medicine (IKEM), Prague, Czech Republic
| | - Jan Kriz
- Diabetes Center, Institute for Clinical and Experimental Medicine, Prague, Czech Republic
| | - Klara Zacharovova
- Laboratory of Pancreatic Islets, Center of Experimental Medicine, Institute for Clinical and Experimental Medicine (IKEM), Prague, Czech Republic
| | - Bas Brinkhof
- Department of Internal Medicine, Leiden University Medical Center (LUMC), Leiden, Netheralnds
| | - Dirk-Jan Cornelissen
- Department of Internal Medicine, Leiden University Medical Center (LUMC), Leiden, Netheralnds
| | - Nicholas Magrane
- Nuffield department of surgical sciences, Oxford Consortium for Islet transplantation, Oxford, UK
| | - Katerina Bittenglova
- Diabetes Center, Institute for Clinical and Experimental Medicine, Prague, Czech Republic
| | - Martin Capek
- Light Microscopy Laboratory, Institute of Molecular Genetics of the Czech Academy of Sciences, Prague, Czech Republic
- Laboratory of Biomathematics, Institute of Physiology of the Czech Academy of Sciences, Prague, Czech Republic
| | - Jan Valecka
- Laboratory of Biomathematics, Institute of Physiology of the Czech Academy of Sciences, Prague, Czech Republic
| | - Alena Habartova
- Redox Photochemistry Lab, Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Prague, Czech Republic
| | - František Saudek
- Diabetes Center, Institute for Clinical and Experimental Medicine, Prague, Czech Republic
| |
Collapse
|
2
|
Girman P, Berkova Z, Dobolilova E, Saudek F. How to use image analysis for islet counting. Rev Diabet Stud 2008; 5:38-46. [PMID: 18548169 PMCID: PMC2517167 DOI: 10.1900/rds.2008.5.38] [Citation(s) in RCA: 14] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/15/2008] [Revised: 05/22/2008] [Accepted: 05/25/2008] [Indexed: 11/03/2022] Open
Abstract
AIM Assessment of islet mass before islet transplantation requires a reliable technique to enable exact analysis of islet volume. This study aimed to test the applicability of digital image analysis (DIA) for evaluation of samples of purified and non-purified islets. METHODS Pancreatic islets were isolated from 10 Lewis rats. Samples of purified (n = 10) and non-purified islets (n = 30) were counted conventionally and by using a computerized method. The equipment for the computerized counting consisted of a digital camera installed on a stereomicroscope and connected to a personal computer. Images of 2272x1704 pixels were processed using a previously described non-commercial program originally developed for this purpose. Islets were converted to equivalents using globe and ellipsoid models. The insulin content of purified islets was assessed using radioimmunoassay and was correlated to the absolute and standardized islet number. RESULTS Mean absolute numbers of purified islets +/- SD were 908 +/- 130 and 1049 +/- 230 (manually and DIA respectively). Mean insulin content +/- SD obtained from purified islets was 161 +/- 45 mU. The mean equivalents of purified islets (1589 +/- 555 for globe and 1219 +/- 452 for ellipsoid) significantly correlated with insulin content. However, this correlation was not significant when absolute islet numbers were used, counted using either method. There was no significant difference in absolute non-purified islet numbers assessed by manual and computerized methods (average +/- SD in 50 microl samples; 12.6 +/- 4.1 and 13.3 +/- 5.3 respectively; p = 0.22). The manual method showed a significantly higher yield of islet equivalents (IE; p < 0.001 for both globe and ellipsoid). CONCLUSION The computer-based system for islet counting correlated better to insulin content than conventional islet estimation and prevented overestimation. Reproducibility and ease of assessment make it potentially applicable to clinical islet transplantation.
Collapse
Affiliation(s)
- Peter Girman
- Diabetes Center, Institute for Clinical and Experimental Medicine, Prague, Czech Republic
| | | | | | | |
Collapse
|