An AI model based on SPECT images has shown promise in helping nuclear medicine physicians identify kidney defects in children, according to a research group in Taiwan. The deep-learning model was able to differentiate between normal and scarred kidneys using Tc-99m DMSA SPECT scans, achieving almost perfect agreement with reads by experienced experts. The researchers believe that this AI model can be used as a computer-aided diagnosis system to assist less-experienced physicians in diagnosing renal cortical defects. The study, although limited by a small number of patient images, is the first to use DL with Tc-99m DMSA SPECT images in evaluating the differentiation between normal and abnormal kidneys. Further research is needed, and the application of these DL models in adults, including renal transplant donors, should be explored.
Thu, 17 Aug 2023 07:03:07 GMT | AuntMinnie