Department of Molecular and Clinical Medicine, Clinical Physiology, Sahlgrenska University Hospital, Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden.
School of Biomedical Engineering and Imaging Sciences Kings College, King's College London and Guy's and St Thomas' PET Centre, London, UK.
Clin Physiol Funct Imaging. 2024 May;44(3):220-227. doi: 10.1111/cpf.12868. Epub 2023 Dec 2.
To compare total metabolic tumour volume (tMTV), calculated using two artificial intelligence (AI)-based tools, with manual segmentation by specialists as the reference.
Forty-eight consecutive Hodgkin lymphoma (HL) patients staged with [18F] fluorodeoxyglucose positron emission tomography/computed tomography were included. The median age was 35 years (range: 7-75), 46% female. The tMTV was automatically measured using the AI-based tools positron emission tomography assisted reporting system (PARS) (from Siemens) and RECOMIA (recomia.org) without any manual adjustments. A group of eight nuclear medicine specialists manually segmented lesions for tMTV calculations; each patient was independently segmented by two specialists.
The median of the manual tMTV was 146 cm (interquartile range [IQR]: 79-568 cm) and the median difference between two tMTV values segmented by different specialists for the same patient was 26 cm (IQR: 10-86 cm). In 22 of the 48 patients, the manual tMTV value was closer to the RECOMIA tMTV value than to the manual tMTV value segmented by the second specialist. In 11 of the remaining 26 patients, the difference between the RECOMIA tMTV and the manual tMTV was small (<26 cm, which was the median difference between two manual tMTV values from the same patient). The corresponding numbers for PARS were 18 and 10 patients, respectively.
The results of this study indicate that RECOMIA and Siemens PARS AI tools could be used without any major manual adjustments in 69% (33/48) and 58% (28/48) of HL patients, respectively. This demonstrates the feasibility of using AI tools to support physicians measuring tMTV for assessment of prognosis in clinical practice.
比较两种基于人工智能（AI）的工具计算的总代谢肿瘤体积（tMTV）与专家手动分割的参考值。
纳入 48 例连续进行[18F]氟脱氧葡萄糖正电子发射断层扫描/计算机断层扫描（PET/CT）分期的霍奇金淋巴瘤（HL）患者。中位年龄为 35 岁（范围：7-75 岁），46%为女性。使用基于 AI 的工具 positron emission tomography assisted reporting system（PARS）（来自西门子）和 RECOMIA（recomia.org）自动测量 tMTV，无需任何手动调整。一组 8 名核医学专家手动分割病变以计算 tMTV；每位患者均由两位专家独立分割。
手动 tMTV 的中位数为 146cm（四分位距[IQR]：79-568cm），同一患者由两位专家分割的两种 tMTV 值的中位数差异为 26cm（IQR：10-86cm）。在 48 例患者中的 22 例中，手动 tMTV 值更接近 RECOMIA tMTV 值，而非第二位专家手动分割的 tMTV 值。在其余 26 例患者中的 11 例中，RECOMIA tMTV 值与手动 tMTV 值的差异较小（<26cm，这是同一患者的两种手动 tMTV 值的中位数差异）。对于 PARS，相应的数字分别为 18 例和 10 例患者。
这项研究的结果表明，在 69%（33/48）和 58%（28/48）的 HL 患者中，分别可以在无需进行重大手动调整的情况下使用 RECOMIA 和西门子 PARS AI 工具。这证明了在临床实践中使用 AI 工具支持医生测量 tMTV 以评估预后的可行性。
