Official announcements

All aboard ūüöÜ

Attention all contributors! As a reminder, the third and final milestone of this challenge ends today. Your contributions will continue to count for points up until midnight UTC. After that, no more points will be awarded for new contributions. So now is the time to get in any last commits! We'll follow up in the coming days with an announcement of final results and next steps for the project. Stay tuned!
2 months, 3 weeks ago

AWS Credits for Contributions to the Algorithms!

Thanks for all the awesome work so far! We're in a great place to use our remaining time improving the quality of our algorithms and how they integrate with the application. Towards that end, we want to make sure that contributors have the resources to train deep learning models and test their predictions. Our friends at AWS have generously provided us with credits to help with the cause. If you're planning to improve the algorithms, email info [at] drivendata [dot] org, and tell us what you're planning to work on, and why you need cloud compute resources to do that. We'll provide $1,000 in credits to the first 6 contributors that are planning on tackling big issues!
3 months, 4 weeks ago

Join us for a walkthrough of the current state of the application!

We'll be running a walkthrough of the application for all contributors on Monday, November 27 at 4pm UTC ([check your time zone](https://time.is/1600_27_Nov_2017_in_UTC)). This will be a great chance to spin up on the project and see the latest version of the end-to-end application. Thanks so much for all of your hardwork so far! Come see what we've accomplished and get motivated for what's next. Details for joining the webinar are below: #### Join from PC, Mac, Linux, iOS or Android: - https://zoom.us/j/378483110 #### Or iPhone one-tap: - US: +16465588656,,378483110# or +16699006833,,378483110# #### Or Telephone: - Dial(for higher quality, dial a number based on your current location)Ôľö - US: +1 646 558 8656 or +1 669 900 6833 (Meeting ID: 378 483 110) - [International numbers available here](https://zoom.us/zoomconference?m=iLlwfthYMV1QcYZrVsiHQY91HnZ5jVUj)
4 months, 3 weeks ago

MVP phase coming to a close!

Hi all, We're excited by the frenzy of progress being made on the project over the last few weeks. Some logistical notes for how this phase will get closed out: * PRs that are opened up until the MVP phase officially closes at midnight UTC tonight will be eligible for points on the MVP leaderboard. Naturally, that means the MVP leaderboard will be subject to change for a few days after the phase ends as the PRs are resolved. * We will treat these PRs normally and they will go through the customary review process ‚ÄĒ as usual, our policy within the open source side of the house is to pretend the points don't exist! ‚ÄĒ except that we expect contributors to be extremely responsive in addressing questions or requested changes so that we can finalize the MVP leaderboard in a timely fashion. * Once the phase ends, the milestone tag on Github will be switched over for issues that now properly belong in the feature building phase. This may not happen instantaneously, but no new points will be entered for the MVP phase after midnight UTC regardless of the Github milestone. For a reminder of how prizes will be allocated at the end of a milestone, see the ['Prizes' page](https://concepttoclinic.drivendata.org/prizes). Happy development! - The C2C team
5 months, 3 weeks ago

Private Docker Cloud repos for early contributors

Just to sweeten the deal: the first 100 submitters of accepted PRs will earn the submitter 5 private Docker Cloud repos thanks to the generous sponsorship of our friends at Docker! Check out the Docker Cloud [site](https://cloud.docker.com/) to learn more about what you can do with private repositories.
8 months, 1 week ago

Launch Announcement

The Concept to Clinic challenge is live! Thanks for joining in this community effort to bring advances in AI to the front lines of lung cancer detection. Read up on the challenge and Sign Up to start contributing today!
8 months, 2 weeks ago

Newsfeed

11 hours, 46 minutes ago
4 days, 11 hours ago
1 week, 1 day ago
1 week, 3 days ago
2 weeks, 5 days ago
2 weeks, 6 days ago
3 weeks, 1 day ago
1 month, 1 week ago
1 month, 1 week ago
1 month, 1 week ago
1 month, 2 weeks ago
1 month, 2 weeks ago
1 month, 2 weeks ago
1 month, 2 weeks ago

@lamby commented on PR #309: Improve various code styles

Any luck all? Would love to get this merged! :)
1 month, 2 weeks ago

@lamby commented on PR #309: Improve various code styles [WIP]

@isms I'm not blocked, more that I wouldn't want to be *introducing* any new issue. Indeed, this PR can be merged (renaming to match)
1 month, 2 weeks ago

@reubano commented on PR #309: Improve various code styles [WIP]

> @reubano Can you dig in and find out what the source is? @isms @lamby can either of you verify that bug?
1 month, 2 weeks ago

@lamby commented on PR #309: Improve various code styles [WIP]

As in, I can't *quite* understand the error message!
1 month, 2 weeks ago
1 month, 2 weeks ago

@isms commented on PR #309: Improve various code styles [WIP]

@reubano Can you dig in and find out what the source is? @lamby I say we let that tests fail for a while while Reuben looks into it so you're not blocked :)
1 month, 3 weeks ago

@reubano commented on PR #309: Improve various code styles [WIP]

Possibly related to https://github.com/concept-to-clinic/concept-to-clinic/pull/305#issuecomment-363400089
1 month, 3 weeks ago

@lamby commented on PR #309: Improve various code styles [WIP]

@isms @reubano Any ideas on this testsuite failure? Can't seem to work out what is actually really broken, let alone how I could have done it...
1 month, 3 weeks ago

@lamby opened a new pull request: #309: Improve various code styles [WIP]

This PR includes a large number of code refactoring within the backend (Django) part of the application.
1 month, 3 weeks ago
1 month, 3 weeks ago
1 month, 3 weeks ago
1 month, 3 weeks ago
1 month, 3 weeks ago
1 month, 4 weeks ago

@reubano commented on PR #305: Refactor imagery loading EventEmitter into store value

When trying to run this on my own, I keep getting a blank screen at port `8080` and the following error: `GET http://34.227.31.217:8080/app.js net::ERR_CONTENT_LENGTH_MISMATCH` Anyone else? FYI, it's not specific to this PR either. I see even at [666fb1](https://github.com/concept-to-clinic/concept-to-clinic/commit/666fb1bf410cbd80eb7928d0e3c5f2433ad668a5).
2 months, 1 week ago
2 months, 2 weeks ago
2 months, 2 weeks ago

@vessemer commented on PR #306: Reactivate tests and clear models

All rebased after #307, tests have been passed.
2 months, 2 weeks ago
2 months, 2 weeks ago
2 months, 2 weeks ago
2 months, 2 weeks ago
2 months, 2 weeks ago

@louisgv commented on PR #305: Refactor imagery loading EventEmitter into store value

@lamby should I rebase/repush in order to make it pass Travis?
2 months, 2 weeks ago

@reubano commented on PR #307: Fix memory and styling errors

Yes, sorry for all the back and forth. The main idea was to just include the relevant (and minimum set of) code changes needed to get the travis test passing.
2 months, 2 weeks ago

@vessemer commented on PR #307: Fix memory and styling errors

@reubano, I'm a bit confused with all those permutations, but it seems to be done :)
2 months, 2 weeks ago

@reubano commented on PR #307: Chose an appropriate DATA_SHAPE

#306 should just be refactoring code. So it seems that the changes in `evaluate_detection.py` are actually fixes which would go here.
2 months, 2 weeks ago
2 months, 2 weeks ago

@reubano commented on PR #307: Chose an appropriate DATA_SHAPE

Tests look good now. Just the styling that is making travis unhappy ```bash $ flake8 prediction prediction/src/algorithms/evaluation/evaluate_detection.py:233:80: W291 trailing whitespace prediction/src/algorithms/evaluation/evaluate_detection.py:246:20: E999 SyntaxError: invalid syntax The command "flake8 prediction" exited with 1. $ pycodestyle prediction prediction/src/algorithms/evaluation/evaluate_detection.py:233:80: W291 trailing whitespace ```
2 months, 2 weeks ago

@vessemer commented on PR #307: Chose an appropriate DATA_SHAPE

``` ============================= test session starts ============================== platform linux -- Python 3.6.3, pytest-3.1.3, py-1.5.2, pluggy-0.4.0 rootdir: /app, inifile: collected 63 items src/tests/test_classification.py ... src/tests/test_classification_3dlrcnn.py ... src/tests/test_cropping.py ... src/tests/test_endpoints.py ..x..... src/tests/test_generators.py ......... src/tests/test_grt123_preprocess.py ... src/tests/test_identification.py xxx src/tests/test_improved_segmentation.py .. src/tests/test_loading.py ....... src/tests/test_preprocess_dicom.py .... src/tests/test_segment_evaluate.py ........ src/tests/test_segmentation.py ....xxx src/tests/test_volume_calculation.py ... =========================== short test summary info ============================ XFAIL src/tests/test_endpoints.py::test_identify Test was stopped after timeout XFAIL src/tests/test_identification.py::test_identify_nodules_001 Test was stopped after timeout XFAIL src/tests/test_identification.py::test_identify_nodules_003 Test was stopped after timeout XFAIL src/tests/test_identification.py::test_identify_luna Test was stopped after timeout XFAIL src/tests/test_segmentation.py::test_nodule_segmentation Test was stopped after timeout XFAIL src/tests/test_segmentation.py::test_lung_segmentation Test was stopped after timeout XFAIL src/tests/test_segmentation.py::test_stop_timeout Test was stopped after timeout =============================== warnings summary =============================== src/tests/test_endpoints.py::test_identify /app/src/tests/../../src/preprocess/extract_lungs.py:34: RuntimeWarning: invalid value encountered in less truncate=2.0) < intensity_th src/tests/test_identification.py::test_identify_nodules_001 /app/src/tests/../../src/preprocess/extract_lungs.py:34: RuntimeWarning: invalid value encountered in less truncate=2.0) < intensity_th src/tests/test_identification.py::test_identify_nodules_003 /app/src/tests/../../src/preprocess/extract_lungs.py:34: RuntimeWarning: invalid value encountered in less truncate=2.0) < intensity_th -- Docs: http://doc.pytest.org/en/latest/warnings.html ============== 56 passed, 7 xfailed, 3 warnings in 991.87 seconds ============== ```
2 months, 2 weeks ago

@vessemer commented on PR #307: Chose an appropriate DATA_SHAPE

@reubano, done, but I've added all code from #308, 'cause in other cases, it'll lead to wrong tests for a new `calculate_volumes` which now works with real-world units.
2 months, 2 weeks ago

@reubano commented on PR #307: Chose an appropriate DATA_SHAPE

Based on [this comment](https://github.com/concept-to-clinic/concept-to-clinic/pull/306#discussion_r165175097) it seems like best option is to add to here, just the bits of code from #308 that will fix the current error... agree?
2 months, 2 weeks ago
2 months, 2 weeks ago

@reubano commented on PR #307: Chose an appropriate DATA_SHAPE

Failing. But at least not due to `memory error` :) ```bash =================================== FAILURES =================================== ______________________________ test_segment_luna _______________________________ metaimage_path = '/images/LUNA-0001/1.3.6.1.4.1.14519.5.2.1.6279.6001.102133688497886810253331438797/1.3.6.1.4.1.14519.5.2.1.6279.6001.102133688497886810253331438797.mhd' luna_nodule = {'x': 0, 'y': 100, 'z': 556} def test_segment_luna(metaimage_path, luna_nodule): > predicted = predict(metaimage_path, [luna_nodule]) src/tests/test_segmentation.py:35: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ src/algorithms/segment/trained_model.py:51: in predict volumes = calculate_volume(segment_path, centroids) src/algorithms/segment/trained_model.py:78: in calculate_volume labels = [mask[centroid['x'], centroid['y'], centroid['z']] for centroid in centroids] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .0 = <list_iterator object at 0x7fab5f819fd0> > labels = [mask[centroid['x'], centroid['y'], centroid['z']] for centroid in centroids] E IndexError: index 556 is out of bounds for axis 2 with size 512 src/algorithms/segment/trained_model.py:78: IndexError ```
2 months, 2 weeks ago
2 months, 2 weeks ago

@vessemer commented on PR #304: Chose an appropriate DATA_SHAPE

@reubano, done by [`Rfactoring`](https://github.com/concept-to-clinic/concept-to-clinic/pull/306), [`Chose an appropriate DATA_SHAPE`](https://github.com/concept-to-clinic/concept-to-clinic/pull/307) and [`Allows to calculate_volume in mm`](https://github.com/concept-to-clinic/concept-to-clinic/pull/308).
2 months, 2 weeks ago

@reubano commented on PR #304: Chose an appropriate DATA_SHAPE

Just saw your new PRs. Let's see how travis likes them :).
2 months, 2 weeks ago

@vessemer opened a new pull request: #308: Allows to calculate_volume in mm

[calculate_volume](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/files#diff-d685e97b90047b1be62100544969f8ac) function now accept centroids in real-world units (mm) if meta was provided, which is logical and also leads to renewal of the [tests](https://github.com/concept-to-clinic/concept-to-clinic/compare/master...vessemer:calculate_volume?expand=1#diff-4dac9e3a237cb8ddd0ca837da32061ce) The swap in tests is due to the `xyz` way to store volumetric data pushed by the PR #44. ## CLA - [X] I have signed the CLA; if other committers are in the commit history, they have signed the CLA as well
2 months, 2 weeks ago

@vessemer opened a new pull request: #307: Chose an appropriate DATA_SHAPE

Previous behaviour leads to various `memory error`s. It happened time to time 'cause of unreasonably large choice of the [`DATA_SHAPE`](https://github.com/concept-to-clinic/concept-to-clinic/commit/06117fb72b1bf03ad8b522d2f9e42af1b6711a2c#diff-638bbca55ea01b2d5eeb28b7e52321eeR21). The average maximum height of lungs for a grown man, according to [this research](https://www.researchgate.net/profile/Kevin_Capello/publication/221873867_Linear_dimensions_and_volumes_of_human_lungs_obtained_from_CT_images/links/5a1d79daaca2726120b2c4b5/Linear-dimensions-and-volumes-of-human-lungs-obtained-from-CT-images.pdf) (Table 2), is 282mm with std equals to 22mm and median 274mm (Table 2). Assuming normal distribution, [3 sigmas](https://en.wikipedia.org/wiki/68%E2%80%9395%E2%80%9399.7_rule) aside the mean will lead to the probability of case `h` belong to the interval `P{|282 - h| <= 66} ~= 0.9974`. Taking into an account the fact that all algorithms which were described (not only implemented ones) works with spacing on z-axis >= 0.9, therefore, having an upper bound value of `(282 + 66) / 0.9 = 386` (voxels) for z-axis should be enough for all cases. I've made `DATA_SHAPE` to be `(512, 512, 512)` which should fit in memory while having reserved space for the z-axis. ## CLA - [x] I have signed the CLA; if other committers are in the commit history, they have signed the CLA as well
2 months, 2 weeks ago

@vessemer opened a new pull request: #306: Refactoring

Refactoring of [`evaluate_detection.py`](https://github.com/concept-to-clinic/concept-to-clinic/compare/master...vessemer:refactoring?expand=1#diff-4764e224c7f674ff2dc226aeafcfedb9) and uncommented tests in [`test_classification.py`](https://github.com/concept-to-clinic/concept-to-clinic/compare/master...vessemer:refactoring?expand=1#diff-923cd917610b38983b2ce7907638acc4), as was mentioned in [this](https://github.com/concept-to-clinic/concept-to-clinic/pull/298#pullrequestreview-92612801) comment. ## CLA - [ x ] I have signed the CLA; if other committers are in the commit history, they have signed the CLA as well
2 months, 2 weeks ago

@vessemer commented on PR #304: Chose an appropriate DATA_SHAPE

> Does this mean it wasn't previously being caught by the code linter that travis runs? Not all of them, 'cause code linter indicate `E999 SyntaxError: invalid syntax` and not proceed after that. > This just sounds like a refactoring, so the current test set should (hopefully) be able to confirm nothing broke That's true. > This one looks like an enhancement. I noticed that the x and z values were swapped. Is that something specific to using mm units? Or was it a latent bug that you caught? The swap is done due to the `xyz` way to store volumetric data pushed by the PR #44.
2 months, 2 weeks ago

@reubano commented on PR #304: Chose an appropriate DATA_SHAPE

> Codestyle of evaluate_detection.py commit is merely about that some thing didn't look right, Does this mean it wasn't previously being caught by the code linter that travis runs? > Remove redundant self.scale_factor which is neither dependend on DATA_SHAPE nor on self.input_shape by this and this. This just sounds like a refactoring, so the current test set should (hopefully) be able to confirm nothing broke > calculate_volume function now accept centroids in real world units (mm) if meta was provided, which is logical and also leads to renewal of the tests This one looks like an enhancement. I noticed that the `x` and `z` values were swapped. Is that something specific to using mm units? Or was it a latent bug that you caught? -- Again, overall great work! And I think we may need to split this up into at least 2 individual PRs. One addressing the `memory error` and commented out tests. This 1st PR could also address the codestyle as long as the current linters properly identified the problem. If not, then a 2nd PR would fix the codestyle **and** the linter. If the `x` and `z` values swapping isn't specifically related to mm units, that should probably go in a sep PR as a bug fix. Otherwise, it can just be included in the last PR to address the mm units and the related test adjustments. Does that make sense?
2 months, 2 weeks ago
2 months, 2 weeks ago

@vessemer commented on PR #304: Chose an appropriate DATA_SHAPE

Thanks, @reubano! Let me describe in a bit more details: commit [All tests acitve](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/commits/32e002d854c9633bdc4538a89933ab60ecbe9708) is just to uncomment the commented tests, then [Codestyle of evaluate_detection.py](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/commits/e4ba764af6c6cb570fdb86efc2632eb59cbf7008) commit is merely about that some thing didn't look right, though in the last commit I've included (additional to the DATA_SHAPE things): - Remove redundant [`self.scale_factor`](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/files#diff-7ebde87c389dab64f0c9e8a81e6b6391L40) which is neither dependend on DATA_SHAPE nor on [`self.input_shape`](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/files#diff-7ebde87c389dab64f0c9e8a81e6b6391L39) by [this](https://github.com/vessemer/concept-to-clinic/blame/e4ba764af6c6cb570fdb86efc2632eb59cbf7008/prediction/src/algorithms/segment/src/models/simple_3d_model.py#L64-L73) and [this](https://github.com/vessemer/concept-to-clinic/blame/e4ba764af6c6cb570fdb86efc2632eb59cbf7008/prediction/src/algorithms/segment/src/models/simple_3d_model.py#L48-L55). - [calculate_volume](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/files#diff-d685e97b90047b1be62100544969f8ac) function now accept centroids in real world units (mm) if meta was provided, which is logical and also leads to renewal of the [tests](https://github.com/concept-to-clinic/concept-to-clinic/pull/304/files#diff-4dac9e3a237cb8ddd0ca837da32061ce)
2 months, 2 weeks ago

@reubano commented on PR #304: Chose an appropriate DATA_SHAPE

Great work @vessemer! I think the main issue is the inclusion of "other fixes" unrelated to `DATA_SHAPE` things. Did you just happen to notice some thing didn't look right (like the `print` statements)? Or were there some tests that were still failing?
2 months, 2 weeks ago

@vessemer commented on PR #304: Chose an appropriate DATA_SHAPE

@lamby, initially I've created this PR to uncomment (activate) all tests, which I forget to uncomment in my previous PR #298, but then I've decided to [figure out an appropriate `DATA_SHAPE`](https://github.com/concept-to-clinic/concept-to-clinic/pull/304#issuecomment-361691349), along with other fixes, thus I've renamed this PR.
2 months, 2 weeks ago
2 months, 2 weeks ago

@lamby commented on PR #305: [WIP] Refactor imagery loading EventEmitter into store value

Please rename this PR (namely, remove "[WIP] " suffix!) when we should review in more detail, or ping here for review ahead of that :)
2 months, 2 weeks ago

@lamby commented on PR #304: All tests active

Can you clarify what you mean by "active" here? :)
2 months, 2 weeks ago

@louisgv opened a new pull request: #305: [WIP] Refactor imagery loading EventEmitter into store value

<!--- Provide a general summary of your changes in the Title above --> ## Description <!--- Describe your changes in detail --> I refactored the event emitter inside the TreeView to send it to the global store. This way if any other component need access to the imagery, it can simply grab it from the store. ## Reference to official issue <!--- If fixing a bug, there should be an existing issue describing it with steps to reproduce --> <!--- Please link to the issue here: --> #302 ## How Has This Been Tested? <!--- Please describe in detail how you tested your changes. --> <!--- Include details of your testing environment, and the tests you ran to --> <!--- see how your change affects other areas of the code, etc. --> Manual testing ## CLA - [ ] I have signed the CLA; if other committers are in the commit history, they have signed the CLA as well
2 months, 2 weeks ago
2 months, 2 weeks ago
2 months, 2 weeks ago

@vessemer commented on PR #304: All tests acitve

I've made `DATA_SHAPE` to be `(512, 512, 512)` which should fit in memory while having reserved space for the z-axis.
2 months, 2 weeks ago

@vessemer commented on PR #304: All tests acitve

@reubano, It happens time to time 'cause of unreasonably large choice of the [`DATA_SHAPE`](https://github.com/concept-to-clinic/concept-to-clinic/commit/06117fb72b1bf03ad8b522d2f9e42af1b6711a2c#diff-638bbca55ea01b2d5eeb28b7e52321eeR21). The average maximum height of lungs for a grown man, according to [this research](https://www.researchgate.net/profile/Kevin_Capello/publication/221873867_Linear_dimensions_and_volumes_of_human_lungs_obtained_from_CT_images/links/5a1d79daaca2726120b2c4b5/Linear-dimensions-and-volumes-of-human-lungs-obtained-from-CT-images.pdf) (Table 2), is 282mm with std equals to 22mm and median 274mm (Table 2). Assuming normal distribution, [3 sigmas](https://en.wikipedia.org/wiki/68%E2%80%9395%E2%80%9399.7_rule) aside the mean will lead to the probability of case `h` belong to the interval `P{|282 - h| <= 66} ~= 0.9974`. Taking into an account the fact that all algorithms which were described (not only implemented ones) works with spacing on z-axis >= 0.9, therefore, having an upper bound value of `(282 + 66) / 0.9 = 386` for z-axis should be enough for all cases.
2 months, 2 weeks ago

@reubano commented on PR #304: All tests acitve

Various `memory error`s. Restarted the build to see if that helps.
2 months, 2 weeks ago

@louisgv commented on issue #302: Refactor EventEmitter to using store observer

Since #288 is closed, I will start working on this issue.
2 months, 2 weeks ago

@vessemer opened a new pull request: #304: All tests acitve

Uncommented tests, as was mentioned in [this](https://github.com/concept-to-clinic/concept-to-clinic/pull/298#pullrequestreview-92612801) comment. ## CLA - [x] I have signed the CLA; if other committers are in the commit history, they have signed the CLA as well
2 months, 2 weeks ago
2 months, 2 weeks ago
2 months, 2 weeks ago

@isms commented on PR #300: Add Training Process for Nodule Detection and Classification - added customized datasets

@swarm-ai You are more than welcome to keep working on the PR if you'd like but at this point it won't result in additional points. Feel free to email us directly if you have questions or concerns.
2 months, 2 weeks ago

@swarm-ai commented on PR #300: Add Training Process for Nodule Detection and Classification - added customized datasets

Hi @isms Can you give me 1-2 days to work on resolving these issues and only just saw these comments?
2 months, 2 weeks ago

@isms commented on PR #300: Add Training Process for Nodule Detection and Classification - added customized datasets

We've discussed internally, and have concluded that both of the following points are true: 1. There is some really interesting and potentially helpful stuff in this PR. 2. We can't accept the PR as-is and there is no apparent roadmap to acceptance. We're going to close the PR but we encourage participants to use this as a resource to help inform model training and potentially other pieces of the application. The submission will be recognized for this aspect of contribution under the "Community" heading.
2 months, 2 weeks ago

@isms commented on PR #288: [WIP] Wizard flow

@louisgv I'm going to go ahead and close this, but we'll still recognize the effort. Thanks for working on it.
2 months, 3 weeks ago

@isms commented on PR #300: Add Training Process for Nodule Detection and Classification - added customized datasets

@swarm-ai We'll need quite a bit more context for the PR. This is a big PR with very little reference to any of the pieces of the existing project. > The data should have lesion box annotations What are these? Where do they come from? Could they be expected to come from new CT imagery without hand labeling? > An exmaple using a CT scan data set from a Taiwan-based clinic is included. This is extremely interesting, but is hard to envision how to integrate this when it comes right before the end of the last phase.
2 months, 3 weeks ago

@00rebe commented on issue #200: Do CasePleuralSpaces.{left.right}_pleural_space need to be nullable?

Hey @lamby, I'm interested in taking a look at this if it's still open. If we get rid of `null=True` on CasePleuralSpaces, the implications are: 1. Tidier and cleaner 2. Both left and right lung information must be given. It's not possible to have a case about a left_pleural_space or a right_pleural_space alone. + Suggestion: it should be possible to provide information for either lung or both lungs but not no information of either i.e. at max, 1 of `left_pleural_space`, `right_pleural_space` can be null. + I'm new to django, if this makes sense do you have any suggestions as to how I could implement this? Thanks, 00rebe
2 months, 3 weeks ago

@louisgv commented on PR #288: [WIP] Wizard flow

@isms I see! If that's the case, I imagine this wizard would be more of a horizontal element that vertical? If so, we might be able to utilize the vue wizard dependency (tho, some of its UI might still not make sense for our app), or I can create a common Flow component extending what I have that will observe the global cross-page steps.
2 months, 3 weeks ago