Objective:To investigate the denoising performance of different deep learning (DL) strategies on low-dose brain 18F-FDG PET images. Methods:This retrospective methodological study was conducted on brain PET/CT images of 50 patients (35 males, 15 females, age 20-87 years) who received 3.7MBq/kg 18F-FDG at the Fifth Affiliated Hospital of Sun Yat-sen University between May 2023 and January 2024. Full-dose PET data were acquired with 2min scan. CT scans were acquired before PET scanning. Low-dose PET sinograms were generated by down-sampling the full-dose list mode data to 1/2, 1/4, and 1/20 of full-dose count level. Both full-dose and low-dose sinograms were reconstructed with random, CT-based attenuation and scatter corrections using the three-dimensional (3D) ordered-subsets expectation maximization (OSEM) algorithm (2 iterations, 20 subsets). A total of 4 DL denoising methods were established: (1) 3D conditional generative adversarial networks (GAN) using only low-dose PET as input (GAN-1); (2) 3D attention-based GAN (AttGAN) with low-dose PET input (AttGAN-1); (3) 3D AttGAN with low-dose PET and CT inputs (AttGAN-2); (4) 3D AttGAN with frequency-separation using low-dose PET and CT inputs (AttGAN-FS-2). For AttGAN-FS-2, during the frequency division process, high- and low-frequency components were extracted from the PET reconstructed images via Fourier transform, then inversed Fourier transform, denoised separately, and finally combined to produce the final denoised images. The dataset was separated into training (70%), validation (10%) and testing (20%) sets using simple random sampling without replacement with a fixed random seed. A 5-fold cross-validation scheme was then applied to test all 50 patients. Performance was evaluated against full-dose PET using normalized mean square error (NMSE), structural similarity (SSIM), peak signal-to-noise ratio (PSNR), contrast-to-noise ratio (CNR), SUV mean and SUV max bias of selected brain ROIs. Wilcoxon signed rank test was used to analyze the differences between the denoising methods. Results:AttGAN-FS-2 showed the best performance among all dose levels, with statistical difference as compared by low-dose PET and GAN-1 denoised images for NMSE, SSIM, PSNR, and CNR ( Z values: 2.92-6.15, all P<0.005). NMSE, SSIM quantitative evaluation results (median) of each model at 1/20 dose were: GAN-1: 0.08, 0.87, AttGAN-1: 0.08, 0.88, AttGAN-2: 0.07, 0.89, AttGAN-FS-2: 0.06, 0.91, respectively ( Z values: 3.24-5.77, all P<0.005). Conclusion:The DL-based method combined with multiple strategies AttGAN-FS-2 shows improved denoising performance for low-dose brain PET images.

New Medical Sciences, as a crucial initiative in transforming and upgrading medical education and healthcare services in China, has promoted deep integration of public health with multiple disciplines. However, the public health discipline still faces numerous challenges such as disease prevention system reconstruction, insufficient technological innovation, and the shortage of professional talents. To address these issues, this paper discusses the current status, challenges, and development of the public health discipline in the context of New Medical Sciences, and proposes strategies including reforming talent cultivation models, enhancing practical capabilities, strengthening faculty development, promoting the innovation of scientific research and social service, actively utilizing artificial intelligence technology, and optimizing international cooperation. These measures aim to achieve innovative development in the public health discipline, and better serve the Healthy China strategy and global public health initiatives.

New Medical Sciences, as a crucial initiative in transforming and upgrading medical education and healthcare services in China, has promoted deep integration of public health with multiple disciplines. However, the public health discipline still faces numerous challenges such as disease prevention system reconstruction, insufficient technological innovation, and the shortage of professional talents. To address these issues, this paper discusses the current status, challenges, and development of the public health discipline in the context of New Medical Sciences, and proposes strategies including reforming talent cultivation models, enhancing practical capabilities, strengthening faculty development, promoting the innovation of scientific research and social service, actively utilizing artificial intelligence technology, and optimizing international cooperation. These measures aim to achieve innovative development in the public health discipline, and better serve the Healthy China strategy and global public health initiatives.

Objective:To investigate the denoising performance of different deep learning (DL) strategies on low-dose brain 18F-FDG PET images. Methods:This retrospective methodological study was conducted on brain PET/CT images of 50 patients (35 males, 15 females, age 20-87 years) who received 3.7MBq/kg 18F-FDG at the Fifth Affiliated Hospital of Sun Yat-sen University between May 2023 and January 2024. Full-dose PET data were acquired with 2min scan. CT scans were acquired before PET scanning. Low-dose PET sinograms were generated by down-sampling the full-dose list mode data to 1/2, 1/4, and 1/20 of full-dose count level. Both full-dose and low-dose sinograms were reconstructed with random, CT-based attenuation and scatter corrections using the three-dimensional (3D) ordered-subsets expectation maximization (OSEM) algorithm (2 iterations, 20 subsets). A total of 4 DL denoising methods were established: (1) 3D conditional generative adversarial networks (GAN) using only low-dose PET as input (GAN-1); (2) 3D attention-based GAN (AttGAN) with low-dose PET input (AttGAN-1); (3) 3D AttGAN with low-dose PET and CT inputs (AttGAN-2); (4) 3D AttGAN with frequency-separation using low-dose PET and CT inputs (AttGAN-FS-2). For AttGAN-FS-2, during the frequency division process, high- and low-frequency components were extracted from the PET reconstructed images via Fourier transform, then inversed Fourier transform, denoised separately, and finally combined to produce the final denoised images. The dataset was separated into training (70%), validation (10%) and testing (20%) sets using simple random sampling without replacement with a fixed random seed. A 5-fold cross-validation scheme was then applied to test all 50 patients. Performance was evaluated against full-dose PET using normalized mean square error (NMSE), structural similarity (SSIM), peak signal-to-noise ratio (PSNR), contrast-to-noise ratio (CNR), SUV mean and SUV max bias of selected brain ROIs. Wilcoxon signed rank test was used to analyze the differences between the denoising methods. Results:AttGAN-FS-2 showed the best performance among all dose levels, with statistical difference as compared by low-dose PET and GAN-1 denoised images for NMSE, SSIM, PSNR, and CNR ( Z values: 2.92-6.15, all P<0.005). NMSE, SSIM quantitative evaluation results (median) of each model at 1/20 dose were: GAN-1: 0.08, 0.87, AttGAN-1: 0.08, 0.88, AttGAN-2: 0.07, 0.89, AttGAN-FS-2: 0.06, 0.91, respectively ( Z values: 3.24-5.77, all P<0.005). Conclusion:The DL-based method combined with multiple strategies AttGAN-FS-2 shows improved denoising performance for low-dose brain PET images.

Objective To investigate the epidemiology of chronic kidney disease (CKD) in patients with hypertension and diabetes mellitus in Kunming urban area.Methods A multistage cluster randomized sampling method was used to collect 400 randomly selected patients (community managed hypertension and diabetes mellitus) in community service centers in the 4 main urban districts of Kunming,Yunnan province.The subjects were screened for CKD by questionnaires,physical examinations,and microalbuminuria tests.Results A total of 343 people were surveyed.The prevalence of albuminuria,proteinuria by routine urinalysis,decreased glomerular filtration rate,and CKD prevalence were respectively 37.3％,12.2％,5.0％ and 39.1％.A total of 134 patients with CKD (134/343) were screened.Logistic regression analysis showed male (OR=2.312,95％CI 1.325-4.037,P=0.003),hyperuricemia (OR=1.751,95％ CI 1.109-2.765,P=0.016) and obesity (OR=2.150,95％ CI 1.115-4.146,P=0.022) were related to CKD.Conclusions The prevalences of CKD and albuminuria are 39.1％ and 37.3％ in patients with chronic diseases (hypertension and diabetes) in the main urban community of Kunming,Yunnan.Hyperuricemia,male and obesity are independent risk factors for CKD.

Objective To study the relationship between status of methylation of tumor suppressor candidate 1 gene(TUSC1) promoter and expression of its protein in adolescent papillary thyroid carcinoma(PTC). Methods Forty cases of adolescent PTC were chosen and the corresponding para carcinoma tissues were taken from July 2010 to Decem-ber 2013 in the First Affiliated Hospital of Zhengzhou University surgical specimens of the thyroid gland and were con-firmed by pathology. Male 12 cases,female 28 cases,median age 14 (10-18) years old. Tumor node metastasis (TNM) stageⅠ-Ⅱ13 cases,Ⅲ-Ⅳstage 27 cases;gradeⅠin 15 cases,gradeⅡin 25 cases;lymph node metastasis in 22 cases,18 cases were negative. Methylation-specific polymerase chain reaction (MSP) and Western blot were applied respectively to examine the methyaltion of TUSC1 gene promoter and its protein expression of 40 samples of adolescent PTC and their matched adjacent non-cancerous epithelium. Results The results of MSP revealed that there was no methylation of TUSC1 gene promoter in adjacent non-cancerous epithelium,while in the adolescent PTC,the hyper-methylation rate was 60%(24/60 cases,χ2=34. 28,P<0. 05). In additional,it was related to the TNM stage,pathological grade and lymph node metastasis (χ2=4. 862,7. 111,5. 625,all P<0. 05). The result of Western blot revealed that the positive expression rate of TUSC1 protein was 100% in adjacent non-cancerous epithelium and 30%(12/40 cases) in adolescent PTC (χ2=14. 118,P<0. 05),which was related to the TNM stage,pathological grade and lymph node metastasis (χ2=5. 215,6. 222, 5. 079,all P<0. 05). There was distinct correlation between methylation of TUSC1 gene promoter and the protein expres-sion (r=-0. 84,P<0. 05). Conclusions Methylation of promoter might be one of the important mechanisms of inactiva-tion of TUSC1 gene,and might play an important role in carcinogenesis and progression of adolescent PTC.

BACKGROUNDThe prognostic values of interim and post-therapy fluorine-18-fluorodeoxyglucose ((18)F-FDG) positron emission tomography (PET) and PET/computed tomography (CT) scanning have been confirmed in several subtypes of lymphoma. However, its prognostic value in Burkitt's lymphoma has not been clearly defined. The aim of the present study was to assess the prognostic value of PET/CT scanning during different treatment processes of Burkitt's lymphoma.

METHODSA total of 29 adult patients with newly diagnosed Burkitt's lymphoma were retrospectively involved in this study; of them, 23 patients underwent baseline PET/CT, 15 patients underwent mid-therapy PET/CT after 1-4 cycles of chemotherapy, and 17 patients underwent post-therapy PET/CT after all planned first-line chemotherapy cycles. Mid-therapy and post-therapy PET/CT results (positive vs. negative) were visually interpreted according to the criteria of the International Harmonization Project. The reduction in the maximum standardizes uptake values (∆SUVmax) of 25%, 50%, and 75% were regarded as cutoff points. Overall survival (OS) and progression-free survival (PFS) were regarded as the major endpoints.

RESULTSThe median OS and PFS were 27.6 months (range 6.5-78.3 months) and 27.2 months (range 3.0-78.3 months), respectively. The median SUVmax of the baseline PET/CT was 18.3 (range 1.6-35.9), whereas the median SUVmax of the mid-therapy and post-therapy PET/CT decreased to 4.0 (range 0-17.6) and 3.0 (range 0-14.5), respectively. The patients' Eastern Cooperative Oncology Group (ECOG) scores (<2 vs. ≥2) were significantly associated with the baseline PET/CT SUVmax. The mid-therapy and post-therapy PET/CT results (positive vs. negative) showed no significant association with OS or PFS. The optimal cutoff ∆SUVmax from the baseline to the post-therapy PET/CT that could predict a change in OS in patients with Burkitt's lymphoma was 50% (P = 0.019).

CONCLUSIONS(18)F-FDG uptake was intense in Burkitt's lymphoma, and there was a significant reduction in SUVmax during the interim and post-therapy PET/CT procedures. A ∆SUVmax of greater than 50% was a favorable cutoff point to predict the OS of Burkitt's lymphoma patients.

Aged ; Antineoplastic Combined Chemotherapy Protocols ; therapeutic use ; Burkitt Lymphoma ; diagnostic imaging ; drug therapy ; Female ; Fluorodeoxyglucose F18 ; Humans ; Kaplan-Meier Estimate ; Male ; Middle Aged ; Positron Emission Tomography Computed Tomography ; methods ; Prognosis ; Radiopharmaceuticals ; Retrospective Studies ; Treatment Outcome