<dd id="rw0xn"></dd>

  • <label id="rw0xn"></label>

  • <sup id="rw0xn"><strike id="rw0xn"></strike></sup><label id="rw0xn"></label>
      <th id="rw0xn"></th>
    1. <var id="rw0xn"></var>
        1. <table id="rw0xn"></table>

          <sub id="rw0xn"><meter id="rw0xn"></meter></sub>

          IAEA Research on AI-Assisted Contouring Suggests Benefits for Cancer Patients

          A radiation oncologist contouring a head and neck cancer case. (Photo: G. Ferraris)

          Research involving 23 countries has highlighted the safety and benefit of using artificial intelligence for a key and often time-consuming step of the cancer treatment process: contouring organs at risk. By adding unique data from low- and middle-income countries to a growing body of scientific evidence, an IAEA coordinated research project (the ELAISA Study) shows how this technology can enhance radiotherapy access around the world.

          This contouring of tumours and nearby healthy tissues (organs-at-risk) is essential for the safe, effective and optimal use of radiotherapy to treat cancer. However, variations in how different observers may contour (inter-observer variability) can impact both the accuracy and consistency of radiotherapy planning. Previous studies have demonstrated that instructor-led guidance workshops can reduce this inter-observer variation.

          Despite nearly half of all cancer patients requiring radiotherapy at some point, this treatment type is underused across the globe – in part because there are not enough clinically trained professionals. The IAEA-led Lancet Oncology Commission on Radiotherapy and Theranostics shows that over 84 000 radiation oncologists are needed by 2050 just to meet the global cancer demand of 35.2 million new cases. “This figure reflects a more than 60 per cent increase in the number of radiation oncologists in 2022,” said May Abdel-Wahab, Director of the IAEA Division of Human Health and the commission’s co-lead. “As cancer cases and treatment complexity increase, radiation oncologists will have to spend even more of their already limited capacity on contouring cancerous tissues and the surrounding healthy ones.” 

          AI Investigated to Help Treat Head and Neck Cancers

          In addressing these challenges within radiation oncology, the IAEA investigated how artificial intelligence (AI) could assist with contouring head and neck cancers in low- and middle-income countries (LMICs).

          While AI-based algorithms have shown promising potential in automatically outlining structures (auto-segmentation), this was mostly observed in retrospective studies. Its actual clinical benefit within LMIC-contexts and in terms of interobserver variability had largely been unresearched until recently.

          “The use of AI to assist with contouring can be an important tool for supporting the efficiency of radiation oncologists,” noted Abdel-Wahab.

          Radiation Oncologists from 22 Countries Participated

          Almost 100 radiation oncologists from 22 different radiotherapy centres — in Albania, Argentina, Azerbaijan, Bangladesh, Belarus, Costa Rica, Georgia, India, Indonesia, Jordan, Kazakhstan, Kenya, Kyrgyzstan, Malaysia, Moldova, Mongolia, Nepal, North Macedonia, Pakistan, Sudan, Tunisia and Uganda — participated in the IAEA’s study while Aarhus University Hospital in Denmark provided a total of 16 head and neck cancer cases. 

          For the study, radiation oncologists were randomly split into two groups: one that delineated organs at risk with AI and another that used manual methods. After an online IAEA workshop on AI-assisted contouring, both groups continued outlining cases, first with their original approach and then all using AI. A final round using AI was done six months later as a follow up.

          AI-assisted Contouring Increased Quality

          Results of the IAEA’s coordinated research project showed that AI assistance not only increased contouring quality by considerably minimizing inter-observer variability but also reduced contouring times — even without prior instruction. While instruction only increased the contouring quality of two organs-at-risk, it did magnify the time-saving effect of AI-assisted contouring. This phenomenon could also be seen over time in the study’s short- and long-term follow-up to the instructor-led workshop.

          “The ELAISA Study demonstrates that teaching combined with AI-assisted contouring was the most effective strategy to reduce contouring time,” explained Jesper Grau Eriksen, clinical professor at Aarhus University and one of the study’s lead investigators. “If applied appropriately, the safe implementation of AI-assisted contouring tools can result in resource sparing, enabling more radiation oncologists, especially those working in LMIC-contexts, to treat even more patients.” 

          The study’s results have been published in the Journal of Global Oncology and have been presented at the annual meetings of the European Society for Radiotherapy and Oncology.

          <dd id="rw0xn"></dd>

        2. <label id="rw0xn"></label>

        3. <sup id="rw0xn"><strike id="rw0xn"></strike></sup><label id="rw0xn"></label>
            <th id="rw0xn"></th>
          1. <var id="rw0xn"></var>
              1. <table id="rw0xn"></table>

                <sub id="rw0xn"><meter id="rw0xn"></meter></sub>
                97碰成人国产免费公开视频