In this paper, a low-dose CT denoising method based on L1/L2regularization method of Markov chain Monte Carlo is studied. Firstly, the mathematical model and regularization method of low-dose CT denoising are summariz...In this paper, a low-dose CT denoising method based on L1/L2regularization method of Markov chain Monte Carlo is studied. Firstly, the mathematical model and regularization method of low-dose CT denoising are summarized, and then the theoretical basis of MCMC method and its application in image denoising are introduced. We evaluated the performance of various regularization strategies by comparing the denoising effects of L1, L2, and L1/L2regularization terms in MCMC sampling at Gaussian noise levels. The experimental results show that L1/L2regularization has the best performance in balancing noise removal and image detail retention, significantly superior to single L1and L2regularization, which proves its effectiveness for low-dose CT denoising.展开更多
In order to solve the problems of artifacts and noise in low-dose computed tomography(CT)images in clinical medical diagnosis,an improved image denoising algorithm under the architecture of generative adversarial netw...In order to solve the problems of artifacts and noise in low-dose computed tomography(CT)images in clinical medical diagnosis,an improved image denoising algorithm under the architecture of generative adversarial network(GAN)was proposed.First,a noise model based on style GAN2 was constructed to estimate the real noise distribution,and the noise information similar to the real noise distribution was generated as the experimental noise data set.Then,a network model with encoder-decoder architecture as the core based on GAN idea was constructed,and the network model was trained with the generated noise data set until it reached the optimal value.Finally,the noise and artifacts in low-dose CT images could be removed by inputting low-dose CT images into the denoising network.The experimental results showed that the constructed network model based on GAN architecture improved the utilization rate of noise feature information and the stability of network training,removed image noise and artifacts,and reconstructed image with rich texture and realistic visual effect.展开更多
In this paper,we propose Hformer,a novel supervised learning model for low-dose computer tomography(LDCT)denoising.Hformer combines the strengths of convolutional neural networks for local feature extraction and trans...In this paper,we propose Hformer,a novel supervised learning model for low-dose computer tomography(LDCT)denoising.Hformer combines the strengths of convolutional neural networks for local feature extraction and transformer models for global feature capture.The performance of Hformer was verified and evaluated based on the AAPM-Mayo Clinic LDCT Grand Challenge Dataset.Compared with the former representative state-of-the-art(SOTA)model designs under different architectures,Hformer achieved optimal metrics without requiring a large number of learning parameters,with metrics of33.4405 PSNR,8.6956 RMSE,and 0.9163 SSIM.The experiments demonstrated designed Hformer is a SOTA model for noise suppression,structure preservation,and lesion detection.展开更多
In order to improve the quality of low-dose computational tomography (CT)images, the paper proposes an improved image denoising approach based on WGAN-gpwith Wasserstein distance. For improving the training and the co...In order to improve the quality of low-dose computational tomography (CT)images, the paper proposes an improved image denoising approach based on WGAN-gpwith Wasserstein distance. For improving the training and the convergence efficiency, thegiven method introduces the gradient penalty term to WGAN network. The novelperceptual loss is introduced to make the texture information of the low-dose imagessensitive to the diagnostician eye. The experimental results show that compared with thestate-of-art methods, the time complexity is reduced, and the visual quality of low-doseCT images is significantly improved.展开更多
AIM:To explore whether computer tomography coronary angiography(CTCA) using iterative reconstruction(IR) leads to significant radiation dose reduction without a significant loss in image interpretability compared to c...AIM:To explore whether computer tomography coronary angiography(CTCA) using iterative reconstruction(IR) leads to significant radiation dose reduction without a significant loss in image interpretability compared to conventional filtered back projection(FBP).METHODS:A consecutive series of 200 patients referred to our institution to undergo CTCA constituted the study population.Patients were sequentially assigned to FBP or IR.All studies were acquired with a 256-slice CT scanner.A coronary segment was considered interpretable if image quality was adequate for evaluation of coronary lesions in all segments ≥ 1.5 mm.RESULTS:The mean age was 56.3±9.6 years and165(83%) were male,with no significant differences between groups.Most scans were acquired using prospective ECG triggering,without differences between groups(FBP 84%vs IR 82%;P=0.71).A total of 3198(94%) coronary segments were deemed of diagnostic quality.The percent assessable coronary segments was similar between groups(FBP 91.7%±4.0% vs IR92.5% ± 2.8%; P=0.12).Radiation dose was significantly lower in the IR group(2.8±1.4 mSvvs 4.6±3.0mSv;P<0.0001).Image noise(37.8±1.4 HUvs 38.2±2.4 HU; P=0.20) and signal density(461.7±51.9HU vs 462.2±51.2 HU; P=0.54) levels did not differ between FBP and IR groups,respectively.The IR group was associated to significant effective dose reductions,irrespective of the acquisition mode.CONCLUSION:Application of IR in CTCA preserves image interpretability despite a significant reduction in radiation dose.展开更多
Background:The effectiveness of lung cancer screening with low-dose computed tomography(LDCT)has been established.The current study evaluates the cost-effectiveness of lung cancer screening with LDCT in a general popu...Background:The effectiveness of lung cancer screening with low-dose computed tomography(LDCT)has been established.The current study evaluates the cost-effectiveness of lung cancer screening with LDCT in a general population in China.Methods:A previously validated micro-simulation model was used to simulate a cohort of men and women on a lifetime horizon in the presence and absence of LDCT screening.The modeling data were collected from numerous national and international sources.Simulated screening scenarios included different combinations of screening intervals and start and stop ages.Additional costs(valued in Chinese Yuan,CNY;1 USD=6.8976 CNY,1 EUR=7.8755 CNY in 2020),life-years gained(LYG)and mortality reduction due to screening were also determined.The costs and life-years were discounted by 3%.All results were scaled to 1,000 individuals.The average cost-effectiveness ratio(ACER)was calculated.A willingness-to-pay threshold of CNY 217.3k/LYG was considered.A healthcare system perspective was adopted.Results:Compared to no screening,lung cancer screening by LDCT in a general Chinese population yielded 21.0-36.7 LYG in men and 9.2-16.6 LYG in women across the scenarios.For men,biennial LDCT screening yielded an ACER of CNY 171.4k-306.3k/LYG relative to no screening.Biennial screening performed between 55 and 75 years of age was optimal at the defined threshold;it resulted in CNY 174.6k/LYG and a lung cancer mortality reduction of 9.1%,and this scenario had a 75%probability of being cost-effective.For women,the ACER ranged from CNY 364.2k to 1193.3k/LYG.Conclusions:In China,lung cancer screening with LDCT in the general population including never smokers could be cost-effective for men with 75%probability,but not for women.The optimal strategy for men would be performing biennial screening between 55 and 75 years of age.展开更多
Purpose: The purpose of this study has been to evaluate the diagnostic information contained in the CT scout view in the detection of body packing. Materials and methods: Retrospect analysis of 43 CT scans between Jul...Purpose: The purpose of this study has been to evaluate the diagnostic information contained in the CT scout view in the detection of body packing. Materials and methods: Retrospect analysis of 43 CT scans between July 2011 and June 2013 in asymptomatic suspects of body packing (29 men, 14 females, mean age 38 ± 9 years). Results: A total of 11 positive cases of body packing were identified. In 10 (91%) of the cases packets were relatively large and spares in number (3 or less);in 7 (64%) a single packet has been identified. 6 (55%) of the packets were located rectally, 4 (36%) vaginally and in 1 (9%) case multiple small packets of approximately 1 cm in size were found to have been ingested orally. Maximum and minimum diameters were 5.9 ± 3 cm and 2.9 ± 1.4 cm, respectively. The mean weight of packets was 7.5 ± 4.2 g (range 2 - 54 g). In 73% (n = 8) heroin had been detected;other drugs such as cocaine (n = 1) and cannabis (n = 1) were encountered once, respectively. One packet was identified retrospectively and its content could therefore not be identified. The average effective dose was 3.8 ± 2.1 mSv for CT, of that 0.12 ± 0.01 mSv was required for the CT scout view. Conclusion: If CT scout view were treated as a diagnostic image, some CT scans may be omitted, thereby maintaining streamlined operations and achieving further dose reduction jointly in the workup of body packing.展开更多
AIM:To report a one-year clinical outcomes of low-dose laser cycloplasty(LCP)among malignant glaucoma patients.METHODS:In this prospective,multicenter,noncomparative clinical study,participants with malignant glaucoma...AIM:To report a one-year clinical outcomes of low-dose laser cycloplasty(LCP)among malignant glaucoma patients.METHODS:In this prospective,multicenter,noncomparative clinical study,participants with malignant glaucoma were recruited and underwent LCP at eight ophthalmic centers in China.Patients were followed up at 1wk,1,3,6,and 12mo.Intraocular pressure(IOP),number of glaucoma medications,anterior chamber depth(ACD),and complications were recorded.Anatomical success was defined as the reformation of the anterior chamber based on slit-lamp biomicroscopy.Recurrence was defined by the presence of a shallow orflat anterior chamber after initial recovery from treatment.RESULTS:A total of 34 eyes received LCP.Mean IOP and medications decreased from 36.1±11.5 mm Hg with 3.3±1.5 glaucoma medications pre-treatment to 20.9±9.8 mm Hg(P<0.001)with 2.9±1.6 medications(P=0.046)at 1d,and 17.4±6.7 mm Hg(P<0.001)with 1.3±1.7 medications(P<0.001)at 12mo.The ACD increased from 1.1±0.8 mm at baseline to 1.7±1.0 mm and to 2.0±0.5 mm at 1d and 12mo,respectively.A total of 32(94.1%)eyes achieved initial anatomical success.During follow-up,2(5.9%)eyes failed and 8(23.5%)eyes relapsed,yielding a 12-month anatomical success rate of 64.3%.Complications including anterior synechia(8.82%),choroidal/ciliary detachment(5.88%)and hypopyon(2.94%)were observed within 1wk.CONCLUSION:LCP is simple,safe,and effective in reforming the anterior chamber in malignant glaucoma.展开更多
Objective: Unresectable hepatocellular carcinoma(uHCC) continues to pose effective treatment options. The objective of this study was to assess the efficacy and safety of combining low-dose cyclophosphamide with lenva...Objective: Unresectable hepatocellular carcinoma(uHCC) continues to pose effective treatment options. The objective of this study was to assess the efficacy and safety of combining low-dose cyclophosphamide with lenvatinib, pembrolizumab and transarterial chemoembolization(TACE) for the treatment of uHCC.Methods: From February 2022 to November 2023, a total of 40 patients diagnosed with uHCC were enrolled in this small-dose, single-center, single-arm, prospective study. They received a combined treatment of low-dose cyclophosphamide with lenvatinib, pembrolizumab, and TACE. Study endpoints included progression-free survival(PFS), objective response rate(ORR), and safety assessment. Tumor response was assessed using the modified Response Evaluation Criteria in Solid Tumors(mRECIST), while survival analysis was conducted through KaplanMeier curve analysis for overall survival(OS) and PFS. Adverse events(AEs) were evaluated according to the National Cancer Institute Common Terminology Criteria for Adverse Events(version 5.0).Results: A total of 34 patients were included in the study. The median follow-up duration was 11.2 [95% confidence interval(95% CI), 5.3-14.6] months, and the median PFS(mPFS) was 15.5(95% CI, 5.4-NA) months.Median OS(mOS) was not attained during the study period. The ORR was 55.9%, and the disease control rate(DCR) was 70.6%. AEs were reported in 27(79.4%) patients. The most frequently reported AEs(with an incidence rate >10%) included abnormal liver function(52.9%), abdominal pain(44.1%), abdominal distension and constipation(29.4%), hypertension(20.6%), leukopenia(17.6%), constipation(17.6%), ascites(14.7%), and insomnia(14.7%). Abnormal liver function(14.7%) had the most common grade 3 or higher AEs.Conclusions: A combination of low-dose cyclophosphamide with lenvatinib, pembrolizumab, and TACE is safe and effective for u HCC, showcasing a promising therapeutic strategy for managing uHCC.展开更多
Low-dose computed tomography(LDCT)contains the mixed noise of Poisson and Gaus-sian,which makes the image reconstruction a challenging task.In order to describe the statistical characteristics of the mixed noise,we ad...Low-dose computed tomography(LDCT)contains the mixed noise of Poisson and Gaus-sian,which makes the image reconstruction a challenging task.In order to describe the statistical characteristics of the mixed noise,we adopt the sinogram preprocessing as a stan-dard maximum a posteriori(MAP).Based on the fact that the sinogram of LDCT has non-local self-similarity property,it exhibits low-rank characteristics.The conventional way of solving the low-rank problem is implemented in matrix forms,and ignores the correlations among similar patch groups.To avoid this issue,we make use of a nonlocal Kronecker-Basis-Representation(KBR)method to depict the low-rank problem.A new denoising model,which consists of the sinogram preprocessing for data fidelity and the nonlocal KBR term,is developed in this work.The proposed denoising model can better illustrate the generative mechanism of the mixed noise and the prior knowledge of the LDCT.Nu-merical results show that the proposed denoising model outperforms the state-of-the-art algorithms in terms of peak-signal-to-noise ratio(PSNR),feature similarity(FSIM),and normalized mean square error(NMSE).展开更多
Lung nodule classification based on low-dose computed tomography(LDCT)images has attracted major attention thanks to the reduced radiation dose and its potential for early diagnosis of lung cancer from LDCT-based lung...Lung nodule classification based on low-dose computed tomography(LDCT)images has attracted major attention thanks to the reduced radiation dose and its potential for early diagnosis of lung cancer from LDCT-based lung cancer screening.However,LDCT images suffer from severe noise,largely influencing the performance of lung nodule classification.Current methods combining denoising and classification tasks typically require the corresponding normal-dose CT(NDCT)images as the supervision for the denoising task,which is impractical in the context of clinical diagnosis using LDCT.To jointly train these two tasks in a unified framework without the NDCT images,this paper introduces a novel self-supervised method,termed strided Noise2Neighbors or SN2N,for blind medical image denoising and lung nodule classification,where the supervision is generated from noisy input images.More specifically,the proposed SN2N can construct the supervision infor-mation from its neighbors for LDCT denoising,which does not need NDCT images anymore.The proposed SN2N method enables joint training of LDCT denoising and lung nodule classification tasks by using self-supervised loss for denoising and cross-entropy loss for classification.Extensively experimental results on the Mayo LDCT dataset demonstrate that our SN2N achieves competitive performance compared with the supervised learning methods that have paired NDCT images as supervision.Moreover,our results on the LIDC-IDRI dataset show that the joint training of LDCT denoising and lung nodule classification significantly improves the performance of LDCT-based lung nodule classification.展开更多
文摘In this paper, a low-dose CT denoising method based on L1/L2regularization method of Markov chain Monte Carlo is studied. Firstly, the mathematical model and regularization method of low-dose CT denoising are summarized, and then the theoretical basis of MCMC method and its application in image denoising are introduced. We evaluated the performance of various regularization strategies by comparing the denoising effects of L1, L2, and L1/L2regularization terms in MCMC sampling at Gaussian noise levels. The experimental results show that L1/L2regularization has the best performance in balancing noise removal and image detail retention, significantly superior to single L1and L2regularization, which proves its effectiveness for low-dose CT denoising.
基金supported by National Natural Science Foundation of China(No.11802272)China Postdoctoral Science Foundation(No.2019M651085)。
文摘In order to solve the problems of artifacts and noise in low-dose computed tomography(CT)images in clinical medical diagnosis,an improved image denoising algorithm under the architecture of generative adversarial network(GAN)was proposed.First,a noise model based on style GAN2 was constructed to estimate the real noise distribution,and the noise information similar to the real noise distribution was generated as the experimental noise data set.Then,a network model with encoder-decoder architecture as the core based on GAN idea was constructed,and the network model was trained with the generated noise data set until it reached the optimal value.Finally,the noise and artifacts in low-dose CT images could be removed by inputting low-dose CT images into the denoising network.The experimental results showed that the constructed network model based on GAN architecture improved the utilization rate of noise feature information and the stability of network training,removed image noise and artifacts,and reconstructed image with rich texture and realistic visual effect.
基金supported by the National Natural Science Foundation of China(Nos.11975292,12222512)the CAS"Light of West Chin"Program+1 种基金the CAS Pioneer Hundred Talent Programthe Guangdong Major Project of Basic and Applied Basic Research(No.2020B0301030008)。
文摘In this paper,we propose Hformer,a novel supervised learning model for low-dose computer tomography(LDCT)denoising.Hformer combines the strengths of convolutional neural networks for local feature extraction and transformer models for global feature capture.The performance of Hformer was verified and evaluated based on the AAPM-Mayo Clinic LDCT Grand Challenge Dataset.Compared with the former representative state-of-the-art(SOTA)model designs under different architectures,Hformer achieved optimal metrics without requiring a large number of learning parameters,with metrics of33.4405 PSNR,8.6956 RMSE,and 0.9163 SSIM.The experiments demonstrated designed Hformer is a SOTA model for noise suppression,structure preservation,and lesion detection.
基金supported by National Natural Science Foundation ofChina (61672279)Project of “Six Talents Peak” in Jiangsu (2012-WLW-023)OpenFoundation of State Key Laboratory of Hydrology-Water Resources and HydraulicEngineering, Nanjing Hydraulic Research Institute, China (2016491411).
文摘In order to improve the quality of low-dose computational tomography (CT)images, the paper proposes an improved image denoising approach based on WGAN-gpwith Wasserstein distance. For improving the training and the convergence efficiency, thegiven method introduces the gradient penalty term to WGAN network. The novelperceptual loss is introduced to make the texture information of the low-dose imagessensitive to the diagnostician eye. The experimental results show that compared with thestate-of-art methods, the time complexity is reduced, and the visual quality of low-doseCT images is significantly improved.
文摘AIM:To explore whether computer tomography coronary angiography(CTCA) using iterative reconstruction(IR) leads to significant radiation dose reduction without a significant loss in image interpretability compared to conventional filtered back projection(FBP).METHODS:A consecutive series of 200 patients referred to our institution to undergo CTCA constituted the study population.Patients were sequentially assigned to FBP or IR.All studies were acquired with a 256-slice CT scanner.A coronary segment was considered interpretable if image quality was adequate for evaluation of coronary lesions in all segments ≥ 1.5 mm.RESULTS:The mean age was 56.3±9.6 years and165(83%) were male,with no significant differences between groups.Most scans were acquired using prospective ECG triggering,without differences between groups(FBP 84%vs IR 82%;P=0.71).A total of 3198(94%) coronary segments were deemed of diagnostic quality.The percent assessable coronary segments was similar between groups(FBP 91.7%±4.0% vs IR92.5% ± 2.8%; P=0.12).Radiation dose was significantly lower in the IR group(2.8±1.4 mSvvs 4.6±3.0mSv;P<0.0001).Image noise(37.8±1.4 HUvs 38.2±2.4 HU; P=0.20) and signal density(461.7±51.9HU vs 462.2±51.2 HU; P=0.54) levels did not differ between FBP and IR groups,respectively.The IR group was associated to significant effective dose reductions,irrespective of the acquisition mode.CONCLUSION:Application of IR in CTCA preserves image interpretability despite a significant reduction in radiation dose.
基金support from China Schol-arship Council(CSC file No.201708340072).
文摘Background:The effectiveness of lung cancer screening with low-dose computed tomography(LDCT)has been established.The current study evaluates the cost-effectiveness of lung cancer screening with LDCT in a general population in China.Methods:A previously validated micro-simulation model was used to simulate a cohort of men and women on a lifetime horizon in the presence and absence of LDCT screening.The modeling data were collected from numerous national and international sources.Simulated screening scenarios included different combinations of screening intervals and start and stop ages.Additional costs(valued in Chinese Yuan,CNY;1 USD=6.8976 CNY,1 EUR=7.8755 CNY in 2020),life-years gained(LYG)and mortality reduction due to screening were also determined.The costs and life-years were discounted by 3%.All results were scaled to 1,000 individuals.The average cost-effectiveness ratio(ACER)was calculated.A willingness-to-pay threshold of CNY 217.3k/LYG was considered.A healthcare system perspective was adopted.Results:Compared to no screening,lung cancer screening by LDCT in a general Chinese population yielded 21.0-36.7 LYG in men and 9.2-16.6 LYG in women across the scenarios.For men,biennial LDCT screening yielded an ACER of CNY 171.4k-306.3k/LYG relative to no screening.Biennial screening performed between 55 and 75 years of age was optimal at the defined threshold;it resulted in CNY 174.6k/LYG and a lung cancer mortality reduction of 9.1%,and this scenario had a 75%probability of being cost-effective.For women,the ACER ranged from CNY 364.2k to 1193.3k/LYG.Conclusions:In China,lung cancer screening with LDCT in the general population including never smokers could be cost-effective for men with 75%probability,but not for women.The optimal strategy for men would be performing biennial screening between 55 and 75 years of age.
文摘Purpose: The purpose of this study has been to evaluate the diagnostic information contained in the CT scout view in the detection of body packing. Materials and methods: Retrospect analysis of 43 CT scans between July 2011 and June 2013 in asymptomatic suspects of body packing (29 men, 14 females, mean age 38 ± 9 years). Results: A total of 11 positive cases of body packing were identified. In 10 (91%) of the cases packets were relatively large and spares in number (3 or less);in 7 (64%) a single packet has been identified. 6 (55%) of the packets were located rectally, 4 (36%) vaginally and in 1 (9%) case multiple small packets of approximately 1 cm in size were found to have been ingested orally. Maximum and minimum diameters were 5.9 ± 3 cm and 2.9 ± 1.4 cm, respectively. The mean weight of packets was 7.5 ± 4.2 g (range 2 - 54 g). In 73% (n = 8) heroin had been detected;other drugs such as cocaine (n = 1) and cannabis (n = 1) were encountered once, respectively. One packet was identified retrospectively and its content could therefore not be identified. The average effective dose was 3.8 ± 2.1 mSv for CT, of that 0.12 ± 0.01 mSv was required for the CT scout view. Conclusion: If CT scout view were treated as a diagnostic image, some CT scans may be omitted, thereby maintaining streamlined operations and achieving further dose reduction jointly in the workup of body packing.
基金Supported by the Program for Zhejiang Leading Talent of S&T Innovation(No.2021R52012)Development Projects of Zhejiang Province(No.2022C03112)Innovation Team Program of Wenzhou.
文摘AIM:To report a one-year clinical outcomes of low-dose laser cycloplasty(LCP)among malignant glaucoma patients.METHODS:In this prospective,multicenter,noncomparative clinical study,participants with malignant glaucoma were recruited and underwent LCP at eight ophthalmic centers in China.Patients were followed up at 1wk,1,3,6,and 12mo.Intraocular pressure(IOP),number of glaucoma medications,anterior chamber depth(ACD),and complications were recorded.Anatomical success was defined as the reformation of the anterior chamber based on slit-lamp biomicroscopy.Recurrence was defined by the presence of a shallow orflat anterior chamber after initial recovery from treatment.RESULTS:A total of 34 eyes received LCP.Mean IOP and medications decreased from 36.1±11.5 mm Hg with 3.3±1.5 glaucoma medications pre-treatment to 20.9±9.8 mm Hg(P<0.001)with 2.9±1.6 medications(P=0.046)at 1d,and 17.4±6.7 mm Hg(P<0.001)with 1.3±1.7 medications(P<0.001)at 12mo.The ACD increased from 1.1±0.8 mm at baseline to 1.7±1.0 mm and to 2.0±0.5 mm at 1d and 12mo,respectively.A total of 32(94.1%)eyes achieved initial anatomical success.During follow-up,2(5.9%)eyes failed and 8(23.5%)eyes relapsed,yielding a 12-month anatomical success rate of 64.3%.Complications including anterior synechia(8.82%),choroidal/ciliary detachment(5.88%)and hypopyon(2.94%)were observed within 1wk.CONCLUSION:LCP is simple,safe,and effective in reforming the anterior chamber in malignant glaucoma.
基金financially supported by the Science and Technology Plan Project of Guangzhou (No. 202102010171)National Natural Science Foundation Cultivation Project of the Third Affiliated Hospital of Sun Yat-sen University (No. 2020GZRPYMS11)+2 种基金Natural Science Foundation of Guangdong Province (No. 2018A030313641)Natural Science Foundation of Guangdong Province (No. 2016A030313848)Science and Technology Plan Project of Guangzhou (No. 201704020175)。
文摘Objective: Unresectable hepatocellular carcinoma(uHCC) continues to pose effective treatment options. The objective of this study was to assess the efficacy and safety of combining low-dose cyclophosphamide with lenvatinib, pembrolizumab and transarterial chemoembolization(TACE) for the treatment of uHCC.Methods: From February 2022 to November 2023, a total of 40 patients diagnosed with uHCC were enrolled in this small-dose, single-center, single-arm, prospective study. They received a combined treatment of low-dose cyclophosphamide with lenvatinib, pembrolizumab, and TACE. Study endpoints included progression-free survival(PFS), objective response rate(ORR), and safety assessment. Tumor response was assessed using the modified Response Evaluation Criteria in Solid Tumors(mRECIST), while survival analysis was conducted through KaplanMeier curve analysis for overall survival(OS) and PFS. Adverse events(AEs) were evaluated according to the National Cancer Institute Common Terminology Criteria for Adverse Events(version 5.0).Results: A total of 34 patients were included in the study. The median follow-up duration was 11.2 [95% confidence interval(95% CI), 5.3-14.6] months, and the median PFS(mPFS) was 15.5(95% CI, 5.4-NA) months.Median OS(mOS) was not attained during the study period. The ORR was 55.9%, and the disease control rate(DCR) was 70.6%. AEs were reported in 27(79.4%) patients. The most frequently reported AEs(with an incidence rate >10%) included abnormal liver function(52.9%), abdominal pain(44.1%), abdominal distension and constipation(29.4%), hypertension(20.6%), leukopenia(17.6%), constipation(17.6%), ascites(14.7%), and insomnia(14.7%). Abnormal liver function(14.7%) had the most common grade 3 or higher AEs.Conclusions: A combination of low-dose cyclophosphamide with lenvatinib, pembrolizumab, and TACE is safe and effective for u HCC, showcasing a promising therapeutic strategy for managing uHCC.
基金supported by the National Natural Science Foundation of China(Grant Nos.U21A20455,61972265,11871348)by the Natural Science Foundation of Guangdong Province of China(Grant No.2020B1515310008)+3 种基金by the Department of Education of Guangdong Province of China(Grant No.2019KZDZX1007)by the PolyU internal Grant No.P0040271by the Pazhou Laboratory,Guangzhou,China(Grant No.PZL2021KF0017)by the Guangdong Key Laboratory of Intelligent Information Processing,China.
文摘Low-dose computed tomography(LDCT)contains the mixed noise of Poisson and Gaus-sian,which makes the image reconstruction a challenging task.In order to describe the statistical characteristics of the mixed noise,we adopt the sinogram preprocessing as a stan-dard maximum a posteriori(MAP).Based on the fact that the sinogram of LDCT has non-local self-similarity property,it exhibits low-rank characteristics.The conventional way of solving the low-rank problem is implemented in matrix forms,and ignores the correlations among similar patch groups.To avoid this issue,we make use of a nonlocal Kronecker-Basis-Representation(KBR)method to depict the low-rank problem.A new denoising model,which consists of the sinogram preprocessing for data fidelity and the nonlocal KBR term,is developed in this work.The proposed denoising model can better illustrate the generative mechanism of the mixed noise and the prior knowledge of the LDCT.Nu-merical results show that the proposed denoising model outperforms the state-of-the-art algorithms in terms of peak-signal-to-noise ratio(PSNR),feature similarity(FSIM),and normalized mean square error(NMSE).
基金supported in part by National Natural Science Foundation of China(No.62101136)Shanghai Municipal of Science and Technology Project(No.20JC1419500)+3 种基金Shanghai Sailing Program(No.21YF1402800)the Shanghai Municipal Science and Technology Major Project(No.2018SHZDZX01)ZJLab,Shanghai Center for Brain Science and Brain-Inspired Technology,the National Key R&D Program of China(No.2018YFB1305104)the Natural Science Foundation of Shanghai(No.21ZR1403600).
文摘Lung nodule classification based on low-dose computed tomography(LDCT)images has attracted major attention thanks to the reduced radiation dose and its potential for early diagnosis of lung cancer from LDCT-based lung cancer screening.However,LDCT images suffer from severe noise,largely influencing the performance of lung nodule classification.Current methods combining denoising and classification tasks typically require the corresponding normal-dose CT(NDCT)images as the supervision for the denoising task,which is impractical in the context of clinical diagnosis using LDCT.To jointly train these two tasks in a unified framework without the NDCT images,this paper introduces a novel self-supervised method,termed strided Noise2Neighbors or SN2N,for blind medical image denoising and lung nodule classification,where the supervision is generated from noisy input images.More specifically,the proposed SN2N can construct the supervision infor-mation from its neighbors for LDCT denoising,which does not need NDCT images anymore.The proposed SN2N method enables joint training of LDCT denoising and lung nodule classification tasks by using self-supervised loss for denoising and cross-entropy loss for classification.Extensively experimental results on the Mayo LDCT dataset demonstrate that our SN2N achieves competitive performance compared with the supervised learning methods that have paired NDCT images as supervision.Moreover,our results on the LIDC-IDRI dataset show that the joint training of LDCT denoising and lung nodule classification significantly improves the performance of LDCT-based lung nodule classification.
