Determining Body Surface Area Using Long‐Axial Field‐of‐View Total‐Body PET: A Novel Method Without Additional Radiation Exposure

Background Body surface area (BSA) is used to normalize physiological measures and guide chemotherapy dosing and is traditionally estimated using height‐ and weight‐based formulas. Although whole‐body computed tomography has been investigated for direct BSA estimation, BSA determination from total‐body positron emission tomography (PET) has not been reported. We developed a method to determine BSA and morphological parameters, including total body volume, from routine long‐axial field‐of‐view (LAFOV) total‐body PET (TBPET) data without additional imaging or radiation exposure. Methods Software was developed in C# using the open‐source Insight Segmentation and Registration Toolkit and Visualization Toolkit, with SimpleITK and ActiViz wrappers ( https://simpleitk.org/ , https://www.kitware.eu/activiz/ ). SimpleITK was used for DICOM import and body‐mask segmentation, vtkMarchingCubes for surface reconstruction, and vtkMassProperties for BSA calculation. Technical validation was performed using a routine quality‐control phantom. Clinical data from 20 patients were acquired using a uEXPLORER LAFOV TBPET scanner. Results PET‐derived BSA differed from the physically measured phantom surface area by 0.19%. In patients, PET‐derived BSA was strongly correlated with BSA estimated using the DuBois, Mosteller, and Haycock formulas ( r = 0.962, 0.973, and 0.974, respectively). Maximum percentage differences were 6.58%, −5.84%, and −8.34%, respectively. Conclusions BSA can be determined from routine LAFOV TBPET data without additional radiation exposure, with potential applicability to multibed whole‐body PET. PET‐derived BSA correlated strongly with conventional estimates, although individual differences exceeded 8%. Larger studies using appropriate reference methods are needed to establish accuracy and clinical utility.