Project 3598: A. Porto, K. L. Voje. 2020. ML-morph: A fast, accurate and general approach for automated detection and landmarking of biological structures in images. Methods in Ecology and Evolution. (In Press)
Specimen: Pseudanthias (unvouchered)
View: Lateral

Abstract

1. Morphometrics has become an indispensable component of the statistical analysis of size and shape variation in biological structures. Morphometric data has traditionally been gathered through low-throughput manual landmark annotation, which represents a significant bottleneck for morphometric-based phenomics. Here we propose a machine-learning-based high-throughput pipeline to collect high-dimensional morphometric data in two dimensional images of semi rigid biological structures.

2. The proposed framework has four main strengths. First, it allows for dense phenotyping with minimal impact on specimens. Second, it presents landmarking accuracy comparable to manual annotators, when applied to standardized datasets. Third, it performs data collection at speeds several orders of magnitude higher than manual annotators. And finally, it is of general applicability (i.e., not tied to a specific study system).

3. State-of-the-art validation procedures show that the method achieves low error levels when applied to three morphometric datasets of increasing complexity, with error varying from 0.57% to 2.2% of the structure’s length in the automated placement of landmarks. As a benchmark for the speed of the entire automated landmarking pipeline, our framework places 23 landmarks on 13,686 objects (zooids) detected in 1684 pictures of fossil bryozoans in 3.12 minutes using a personal computer.

4. The proposed machine-learning-based phenotyping pipeline can greatly increase the scale, reproducibility and speed of data collection within biological research. To aid the use of the framework, we have developed a file conversion algorithm that can be used to leverage current morphometric datasets for automation, allowing the entire procedure, from model training all the way to prediction, to be performed in a matter of hours.



Read the article »

Project DOI: 10.7934/P3598, http://dx.doi.org/10.7934/P3598
This project contains
  • 861 Media
  • 3 Taxa
  • 3 Specimens
Total size of project's media files: 289.01M

Download Project SDD File
Currently Viewing:
MorphoBank Project 3598

    Authors' Institutions

    • University of Oslo



    Members

    member name taxa specimens media
    Arthur Porto
    Project Administrator
    33861


    Project has no matrices defined.



    Project views

    type number of views Individual items viewed (where applicable)
    Total project views27412
    Project overview723
    Media views24024Media search (1439 views); M684301 (147 views); M684295 (148 views); M684330 (120 views); M684376 (115 views); M684322 (120 views); M684298 (156 views); M684348 (114 views); M684312 (141 views); M684287 (155 views); M684285 (149 views); M684325 (124 views); M684328 (113 views); M684292 (148 views); M684377 (115 views); M684323 (127 views); M684281 (145 views); M684360 (112 views); M684304 (158 views); M684357 (110 views); M684289 (150 views); M684310 (152 views); M684368 (110 views); M684347 (121 views); M684338 (113 views); M684288 (154 views); M684361 (121 views); M684336 (115 views); M684315 (143 views); M684308 (146 views); M684327 (120 views); M684305 (147 views); M684319 (147 views); M684282 (146 views); M684362 (121 views); M684297 (146 views); M684317 (155 views); M684354 (135 views); M684355 (115 views); M684353 (126 views); M684379 (113 views); M684337 (113 views); M684326 (128 views); M684344 (113 views); M684346 (115 views); M684320 (151 views); M684293 (159 views); M684296 (147 views); M684286 (152 views); M684363 (124 views); M684284 (155 views); M684367 (123 views); M684352 (111 views); M684350 (114 views); M684316 (157 views); M684307 (145 views); M684313 (135 views); M684340 (115 views); M684311 (148 views); M684341 (112 views); M684378 (113 views); M684371 (120 views); M684303 (147 views); M684318 (152 views); M684335 (118 views); M684290 (153 views); M684373 (114 views); M684302 (146 views); M684366 (108 views); M684333 (108 views); M684356 (114 views); M684370 (109 views); M684375 (117 views); M684300 (155 views); M684324 (124 views); M684358 (111 views); M684283 (144 views); M684364 (111 views); M684339 (115 views); M684365 (110 views); M684359 (119 views); M684342 (116 views); M684374 (118 views); M684334 (113 views); M684343 (114 views); M684351 (113 views); M684309 (140 views); M684280 (153 views); M684306 (142 views); M684372 (114 views); M684314 (146 views); M684299 (144 views); M684291 (152 views); M684332 (114 views); M684345 (114 views); M684369 (114 views); M684294 (159 views); M684349 (122 views); M684331 (110 views); M684321 (144 views); M684329 (116 views); M684499 (72 views); M684493 (77 views); M684433 (77 views); M684471 (78 views); M684467 (77 views); M684461 (74 views); M684460 (78 views); M684383 (74 views); M684422 (80 views); M684380 (77 views); M684474 (83 views); M684381 (77 views); M684476 (72 views); M684446 (76 views); M684384 (80 views); M684387 (84 views); M684490 (79 views); M684473 (75 views); M684455 (76 views); M684464 (73 views); M684410 (83 views); M684486 (73 views); M684423 (77 views); M684430 (83 views); M684418 (76 views); M684420 (69 views); M684504 (78 views); M684386 (76 views); M684449 (81 views); M684421 (82 views); M684427 (81 views); M684500 (75 views); M684479 (71 views); M684462 (78 views); M684425 (77 views); M684469 (76 views); M684439 (76 views); M684442 (75 views); M684393 (75 views); M684482 (78 views); M684406 (75 views); M684435 (76 views); M684495 (76 views); M684401 (79 views); M684390 (74 views); M684456 (82 views); M684431 (78 views); M684415 (75 views); M684388 (76 views); M684502 (74 views); M684399 (73 views); M684441 (77 views); M684440 (75 views); M684503 (72 views); M684480 (71 views); M684489 (73 views); M684472 (79 views); M684438 (74 views); M684397 (74 views); M684395 (83 views); M684466 (81 views); M684400 (75 views); M684497 (74 views); M684496 (72 views); M684414 (82 views); M684402 (74 views); M684416 (77 views); M684426 (78 views); M684413 (76 views); M684477 (75 views); M684494 (65 views); M684470 (76 views); M684483 (74 views); M684443 (74 views); M684398 (84 views); M684463 (73 views); M684488 (73 views); M684405 (80 views); M684444 (75 views); M684407 (76 views); M684475 (74 views); M684450 (75 views); M684465 (78 views); M684403 (84 views); M684458 (80 views); M684478 (76 views); M684505 (74 views); M684417 (76 views); M684457 (85 views); M684409 (76 views); M684487 (72 views); M684448 (73 views); M684492 (73 views); M684436 (74 views); M684491 (75 views); M684392 (76 views); M684412 (76 views); M684428 (76 views); M684419 (75 views); M684468 (81 views); M684408 (83 views); M684447 (79 views); M684454 (76 views); M684389 (77 views); M684429 (78 views); M684394 (79 views); M684411 (78 views); M684452 (78 views); M684382 (78 views); M684385 (81 views); M684391 (76 views); M684396 (75 views); M684404 (75 views); M684424 (79 views); M684432 (80 views); M684434 (77 views); M684437 (76 views); M684445 (80 views); M684451 (75 views); M684453 (77 views); M684481 (79 views); M684484 (78 views); M684485 (78 views); M684498 (78 views); M684501 (78 views); M684517 (1 view); M684711 (1 view); M684782 (1 view);
    Views for media list537
    Taxon list826
    Bibliography221
    Specimen list1081




    Project downloads

    type number of downloads Individual items downloaded (where applicable)
    Total downloads from project125
    Project downloads112
    Media downloads13M684288 (1 download); M684289 (1 download); M684290 (1 download); M684292 (1 download); M684293 (1 download); M684294 (1 download); M684295 (1 download); M684296 (1 download); M684298 (1 download); M684299 (1 download); M684303 (1 download); M684481 (1 download); M684281 (1 download);