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
Review
. 2023 Sep 1;15(17):4389.
doi: 10.3390/cancers15174389.

Automatic Segmentation with Deep Learning in Radiotherapy

Lars Johannes Isaksson  1   2 , Paul Summers  3 , Federico Mastroleo  1   4 , Giulia Marvaso  1 , Giulia Corrao  1 , Maria Giulia Vincini  1 , Mattia Zaffaroni  1 , Francesco Ceci  2   5 , Giuseppe Petralia  2   6 , Roberto Orecchia  7 , Barbara Alicja Jereczek-Fossa  1   2
Affiliations
Review

Automatic Segmentation with Deep Learning in Radiotherapy

Lars Johannes Isaksson et al. Cancers (Basel). .

Abstract

This review provides a formal overview of current automatic segmentation studies that use deep learning in radiotherapy. It covers 807 published papers and includes multiple cancer sites, image types (CT/MRI/PET), and segmentation methods. We collect key statistics about the papers to uncover commonalities, trends, and methods, and identify areas where more research might be needed. Moreover, we analyzed the corpus by posing explicit questions aimed at providing high-quality and actionable insights, including: "What should researchers think about when starting a segmentation study?", "How can research practices in medical image segmentation be improved?", "What is missing from the current corpus?", and more. This allowed us to provide practical guidelines on how to conduct a good segmentation study in today's competitive environment that will be useful for future research within the field, regardless of the specific radiotherapeutic subfield. To aid in our analysis, we used the large language model ChatGPT to condense information.

Keywords: artificial intelligence; artificial neural networks; automatic; deep learning; radiotherapy; segmentation.

PubMed Disclaimer

Conflict of interest statement

B.A.J.-F. received speaker fees from Roche, Bayer, Janssen, Ipsen, Accuray, Astellas, Elekta, IBA and Astra Zeneca (all outside the current project). The remaining authors declare no conflict of interest. M.G.V. and G.C. received a research fellowship from the Associazione Italiana per la Ricerca sul Cancro (AIRC) entitled “Radioablation ± hormonotherapy for prostate cancer oligorecurrences (RADIOSA trial): potential of imaging and biology” registered at Clinical Trials.gov NCT03940235, approved by the Ethics Committee of IEO and Centro Cardiologico Monzino (IEO-997). The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Figures

Figure 1
Figure 1
Flowchart of the paper selection and analysis process.
Figure 2
Figure 2
Number of deep learning medical image segmentation studies published per year. Not pictured: five papers published between 1999 and 2012.
Figure 3
Figure 3
The total number of published studies per cancer site.
Figure 4
Figure 4
The total number of published studies organized by (a) image type and (b) segmentation type.
Figure 5
Figure 5
Distribution of the number of patients included in the studies (if a study used multiple datasets, their sizes were added together).
Figure 6
Figure 6
Distribution of dataset sizes per body region. The numbers to the right indicate the median sample size. Note the log scale on the x-axis.
Figure 7
Figure 7
The number of yes and no responses when asked: “Did the paper propose a novel segmentation method or deep learning architecture? (Yes/No)”. The number of studies proposing novel models/architectures shows that the vast majority of papers focus on developing new methods.
See this image and copyright information in PMC

References

    1. Brouwer C.L., Steenbakkers R.J., van den Heuvel E., Duppen J.C., Navran A., Bijl H.P., Chouvalova O., Burlage F.R., Meertens H., Langendijk J.A., et al. 3D Variation in Delineation of Head and Neck Organs at Risk. Radiat. Oncol. 2012;7:32. doi: 10.1186/1748-717X-7-32. - DOI - PMC - PubMed
    1. Veiga-Canuto D., Cerdà-Alberich L., Sangüesa Nebot C., Martínez de las Heras B., Pötschger U., Gabelloni M., Carot Sierra J.M., Taschner-Mandl S., Düster V., Cañete A., et al. Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images. Cancers. 2022;14:3648. doi: 10.3390/cancers14153648. - DOI - PMC - PubMed
    1. Van de Steene J., Linthout N., de Mey J., Vinh-Hung V., Claassens C., Noppen M., Bel A., Storme G. Definition of Gross Tumor Volume in Lung Cancer: Inter-Observer Variability. Radiother. Oncol. 2002;62:37–49. doi: 10.1016/S0167-8140(01)00453-4. - DOI - PubMed
    1. Senan S., van Sörnsen de Koste J., Samson M., Tankink H., Jansen P., Nowak P.J.C.M., Krol A.D.G., Schmitz P., Lagerwaard F.J. Evaluation of a Target Contouring Protocol for 3D Conformal Radiotherapy in Non-Small Cell Lung Cancer. Radiother. Oncol. 1999;53:247–255. doi: 10.1016/S0167-8140(99)00143-7. - DOI - PubMed
    1. Covert E.C., Fitzpatrick K., Mikell J., Kaza R.K., Millet J.D., Barkmeier D., Gemmete J., Christensen J., Schipper M.J., Dewaraja Y.K. Intra- and Inter-Operator Variability in MRI-Based Manual Segmentation of HCC Lesions and Its Impact on Dosimetry. EJNMMI Phys. 2022;9:90. doi: 10.1186/s40658-022-00515-6. - DOI - PMC - PubMed