Back

Cyclops: Cyclical Development Towards Operationalizing Ml Models For Health

Krishnan, A.; Subasri, V.; McKeen, K.; Kore, A.; Ogidi, F.; Alinoori, M.; Verma, A.; Razak, F.; Dhalla, A.; Lalani, N.; Pandya, D.; Dolatabadi, E.

2022-12-06 health informatics
10.1101/2022.12.02.22283021 medRxiv
Show abstract

AO_SCPLOWBSTRACTC_SCPLOWOpen source software that enable research and development of machine learning (ML) models for clinical use cases are fragmented, poorly maintained and fall short in functionality. CyclOps is a software framework designed to address this gap and help accelerate the development of ML models for health. In this paper, we describe the architecture, APIs and implementation details of CyclOps, while providing benchmarks on example clinical use cases. We emphasize that CyclOps is developed to be researcher friendly, while providing APIs for building end-to-end pipelines for model development as well as deployment. We adopt software engineering and ML operations (MLOps) best practices, while providing support for handling large volumes of health data. The design of the framework is centered around the notion of iterative and cyclical development of the overall ML system, which consists of data, model development and monitoring pipelines. The core CyclOps package can be installed through the Python Package Index (PyPI) and the source code is available at https://github.com/VectorInstitute/cyclops.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

1
BMC Medical Informatics and Decision Making
○ 43 papers in training set
Top 0.1%
17.6%
Similar papers in this journal
2
JAMIA Open
◐ 42 papers in training set
Top 0.1%
14.4%
Similar papers in this journal
3
PLOS Digital Health
● 106 papers in training set
Top 0.7%
9.2%
Similar papers in this journal
4
Journal of the American Medical Informatics Association
◐ 71 papers in training set
Top 0.4%
8.5%
Similar papers in this journal
5
PLOS ONE
● 5266 papers in training set
Top 22%
7.5%
Similar papers in this journal
50% of probability mass above
6
Journal of Biomedical Informatics
○ 47 papers in training set
Top 0.3%
5.3%
7
International Journal of Medical Informatics
○ 26 papers in training set
Top 0.4%
3.1%
8
Journal of Medical Internet Research
◐ 87 papers in training set
Top 0.9%
3.1%
9
Computer Methods and Programs in Biomedicine
○ 28 papers in training set
Top 0.2%
3.1%
10
JMIR Medical Informatics
◐ 18 papers in training set
Top 0.2%
2.5%
11
Patterns
○ 78 papers in training set
Top 0.9%
2.3%
12
Bioinformatics
◐ 1204 papers in training set
Top 6%
2.3%
13
GigaScience
◐ 212 papers in training set
Top 2%
2.0%
14
Frontiers in Digital Health
○ 24 papers in training set
Top 0.6%
1.6%
15
JCO Clinical Cancer Informatics
● 22 papers in training set
Top 0.4%
1.6%
16
npj Digital Medicine
○ 118 papers in training set
Top 2%
1.4%
17
IEEE Journal of Biomedical and Health Informatics
● 37 papers in training set
Top 0.7%
1.4%
18
Scientific Reports
○ 3612 papers in training set
Top 59%
1.4%
19
Frontiers in Artificial Intelligence
○ 20 papers in training set
Top 0.4%
1.3%
20
Artificial Intelligence in Medicine
○ 17 papers in training set
Top 0.5%
1.3%
21
PLOS Computational Biology
● 1863 papers in training set
Top 18%
1.1%
22
Computers in Biology and Medicine
○ 128 papers in training set
Top 5%
0.6%
23
Biology Methods and Protocols
◐ 61 papers in training set
Top 3%
0.6%
24
BMJ Health & Care Informatics
● 15 papers in training set
Top 1%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.