Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Aug 30;13(17):2813.
doi: 10.3390/diagnostics13172813.

Integrating Artificial Intelligence Tools in the Clinical Research Setting: The Ovarian Cancer Use Case

Lorena Escudero Sanchez  1   2   3 , Thomas Buddenkotte  1   4   5   6 , Mohammad Al Sa'd  3   7 , Cathal McCague  1   2   8 , James Darcy  3   9 , Leonardo Rundo  1   2   10 , Alex Samoshkin  11 , Martin J Graves  1   8 , Victoria Hollamby  12 , Paul Browne  13 , Mireia Crispin-Ortuzar  2   14 , Ramona Woitek  1   2   15 , Evis Sala  1   2   3   8   16   17 , Carola-Bibiane Schönlieb  4 , Simon J Doran  3   9 , Ozan Öktem  18
Affiliations

Integrating Artificial Intelligence Tools in the Clinical Research Setting: The Ovarian Cancer Use Case

Lorena Escudero Sanchez et al. Diagnostics (Basel). .

Abstract

Artificial intelligence (AI) methods applied to healthcare problems have shown enormous potential to alleviate the burden of health services worldwide and to improve the accuracy and reproducibility of predictions. In particular, developments in computer vision are creating a paradigm shift in the analysis of radiological images, where AI tools are already capable of automatically detecting and precisely delineating tumours. However, such tools are generally developed in technical departments that continue to be siloed from where the real benefit would be achieved with their usage. Significant effort still needs to be made to make these advancements available, first in academic clinical research and ultimately in the clinical setting. In this paper, we demonstrate a prototype pipeline based entirely on open-source software and free of cost to bridge this gap, simplifying the integration of tools and models developed within the AI community into the clinical research setting, ensuring an accessible platform with visualisation applications that allow end-users such as radiologists to view and interact with the outcome of these AI tools.

Keywords: artificial intelligence; cancer research; clinical integration; imaging; radiomics.

PubMed Disclaimer

Conflict of interest statement

E.S: Lucida Medical (co-founder and shareholder), GE HealthCare (research support, speakers’ bureau), Canon (research support, speakers’ bureau). The rest of the authors declare no potential conflict of interest.

Figures

Figure 1
Figure 1
Example script using XNATPy to establish a connection with the XNAT database, open stored DICOM files and perform an action on them.
Figure 2
Figure 2
Example custom inference implemented for the ovarian cancer segmentation to integrate in the AIAA framework, showing the I/O logic necessary before and after running the algorithm.
Figure 3
Figure 3
Simple configuration file showing the I/O elements necessary for a custom model with two possible outcomes (treatment response and non-response) to interface with AIAA/MONAI, keeping all the logic in the custom inference (CustomInference_om_crs.CustomInference).
Figure 4
Figure 4
Schematic view of the display of a test project in XNAT.
Figure 5
Figure 5
View of a CT imaging session in the OHIF viewer plugin to XNAT with the mask menu on the right-hand side presenting the connection to the AIAA server established and the drop-down menu with the categories of available AI models.
Figure 6
Figure 6
Overview of the pipeline created for the ovarian cancer segmentation and response prediction use case, showing its multiple components that seamlessly integrate AI models into visualisation tools.
Figure 7
Figure 7
Upper portion of API web-page created by the NVIDIA AIAA Docker container and accessed in the URL of the static (public) IP address of the server.
Figure 8
Figure 8
Basic administration commands (e.g., list, access, upload or delete models) accessible through the API interface.
Figure 9
Figure 9
Waiting screen seen by the end-user while the session is created at the AI server and the AI inference process is running.
Figure 10
Figure 10
API web page for uploading new models.
Figure 11
Figure 11
API command for new model uploading to the AIAA server.
Figure 12
Figure 12
Example 2D CT image with overlaid masks showing the outcome of the custom inference for abdominal lesions added to the pipeline: omentum (red), right upper quadrant (RUQ, green), left upper quadrant (LUQ, purple), mesentery (brown), left paracolic gutter (LPG, blue) and right paracolic gutter (RPG, pink).
Figure 13
Figure 13
Example outcome of the segmentation and radiomics-based prediction of omental lesions of ovarian cancer, showing the lesions in green as the prediction for this particular patient was to respond to the chemotherapy treatment.
See this image and copyright information in PMC

References

    1. Gillies R., Kinahan P., Hricak H. Radiomics: Images are more than pictures, they are data. Radiology. 2015;278:563–577. doi: 10.1148/radiol.2015151169. - DOI - PMC - PubMed
    1. Aerts H., Rios Velazquez E., Leijenaar R., Parmar C., Grossmann P., Cavalho S., Bussink J., Monshouwer R., Haibe-Kains B., Rietveld D., et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat. Commun. 2014;5:4006. doi: 10.1038/ncomms5006. - DOI - PMC - PubMed
    1. Lambin P., Leijenaar R.T., Deist T.M., Peerlings J., De Jong E.E., Van Timmeren J., Sanduleanu S., Larue R.T., Even A.J., Jochems A., et al. Radiomics: The bridge between medical imaging and personalized medicine. Nat. Rev. Clin. Oncol. 2017;14:749–762. doi: 10.1038/nrclinonc.2017.141. - DOI - PubMed
    1. Rundo L., Beer L., Escudero Sanchez L., Crispin-Ortuzar M., Reinius M., McCague C., Sahin H., Bura V., Pintican R., Zerunian M., et al. linically Interpretable Radiomics-Based Prediction of Histopathologic Response to Neoadjuvant Chemotherapy in High-Grade Serous Ovarian Carcinoma. Front Oncol. 2022 doi: 10.3389/fonc.2022.868265. - DOI - PMC - PubMed
    1. McKinney S.M., Sieniek M., Godbole V., Godwin J., Antropova N., Ashrafian H., Back T., Chesus M., Corrado G.S., Darzi A., et al. International evaluation of an AI system for breast cancer screening. Nature. 2020;577:89–94. doi: 10.1038/s41586-019-1799-6. - DOI - PubMed

LinkOut - more resources