目的建立人工智能辅助结直肠癌病理切片分子分型诊断系统。方法在癌症基因组图谱(the cancer genome Atlas,TCGA)数据库中筛选出422例结直肠癌患者的812张病理切片,分为训练集(75%)和测试集(25%);存入www.paiwsit.com数据库中,根据资深...目的建立人工智能辅助结直肠癌病理切片分子分型诊断系统。方法在癌症基因组图谱(the cancer genome Atlas,TCGA)数据库中筛选出422例结直肠癌患者的812张病理切片,分为训练集(75%)和测试集(25%);存入www.paiwsit.com数据库中,根据资深的病理医生标注的数据进行处理及分割,得到超过400万张带有标签的训练集,最后利用深度学习模型进行训练。结果在经过多种卷积神经网络模型训练后,在110例203张切片的测试集上测试,子图级别达到53.04%的准确率,切片级别准确率达到51.72%,其中结直肠癌共识分子亚型之一的经典型(CMS2)切片级准确率达到75.00%。结论本研究对促进结直肠癌筛查和精准治疗具有重要意义。展开更多
Objective: The Immunoscore method has proved fruitful for predicting prognosis in patients with colon cancer.However, there is still room for improvement in this scoring method to achieve further advances in its clini...Objective: The Immunoscore method has proved fruitful for predicting prognosis in patients with colon cancer.However, there is still room for improvement in this scoring method to achieve further advances in its clinical translation. This study aimed to develop and validate a modified Immunoscore(IS-mod) system for predicting overall survival(OS) in patients with stage Ⅰ-Ⅲ colon cancer.Methods: The IS-mod was proposed by counting CD3+ and CD8+ immune cells in regions of the tumor core and its invasive margin by drawing two lines of interest. A discovery cohort(N=212) and validation cohort(N=103)from two centers were used to evaluate the prognostic value of the IS-mod.Results: In the discovery cohort, 5-year survival rates were 88.6% in the high IS-mod group and 60.7% in the low IS-mod group. Multivariate analysis confirmed that the IS-mod was an independent prognostic factor for OS[adjusted hazard ratio(HR)=0.36, 95% confidence interval(95% CI): 0.20-0.63]. With less annotation and computation cost, the IS-mod achieved performance comparable to that of the Immunoscore-like(IS-like) system(C-index, 0.676 vs. 0.661, P=0.231). The 2-category IS-mod using 47.5% as the threshold had a better prognostic value than that using a fixed threshold of 25%(C-index, 0.653 vs. 0.573, P=0.004). Similar results were confirmed in the validation cohort.Conclusions: Our method simplifies the annotation and accelerates the calculation of Immunoscore method,thus making it easier for clinical implementation. The IS-mod achieved comparable prognostic performance when compared to the IS-like system in both cohorts. Besides, we further found that even with a small reference set(N≥120), the IS-mod still demonstrated a stable prognostic value. This finding may inspire other institutions to develop a local reference set of an IS-mod system for more accurate risk stratification of colon cancer.展开更多
Background The prevalence of thyroid cancer is growing rapidly.Early and precise diagnosis is critical in thy-roid cancer caring.An automatic thyroid cancer diagnostic tool can be valuable to achieve early detection a...Background The prevalence of thyroid cancer is growing rapidly.Early and precise diagnosis is critical in thy-roid cancer caring.An automatic thyroid cancer diagnostic tool can be valuable to achieve early detection and diagnostic consistency.Only the follicular areas in the sample contain useful information to the thyroid cancer diagnosis based on fine needle aspiration(FNA).This study aimed to develop a highly efficient accurate method for follicular cell areas segmentation(FCAS)of thyroid cytopathological whole slide images(WSIs).Methods A total of 96 cell samples from July 2017 to July 2018 were collected in one hospital in Beijing,China.Forty-three WSIs were selected and manually labeled,including 17 cases of papillary thyroid carci-noma sample and 26 cases of benign sample.Six thousand and nine hundred cropped typical image patches(available on https://github.com/bupt-ai-cz/Hybrid-Model-Enabling-Highly-Efficient-Follicular-Segmentation)of 1024×1024 pixels from 13 large WSIs were used for patch-level model training and testing and all of the 13 large WSIs were papillary thyroid carcinoma samples.Thirty testing WSIs with an average size 36,217×29,400(from 10,240×10,240 to 81,920×61,440)were used to test the effectiveness of the hybrid model.Based on the traditional semantic segmentation model deeplabv3,we constructed a hybrid segmentation architecture by adding a classification branch into the segmentation scheme to improve efficiency.Accuracy was used to measure the performance of the classification model;pixel accuracy(pAcc),mean accuracy(mAcc),mean intersection over union(mIoU),and frequency weighted intersection over union(fwIoU)were used to measure the performance of the segmentation model,respectively.Results Using this method,up to 93%WSI segmentation time was reduced by skipping the colloidal areas and the blank background areas.The average processing time of 30 WSI was 49.49 s.On the patch dataset,this hybrid model might reach pAcc=98.65%,mAcc=85.60%,mIoU=79.61%,and fwIoU=97.54%.On the WSI datase展开更多
Epithelial and stromal tissues are components of the tumor microenvironment and play a major role in tumor initiation and progression.Distinguishing stroma from epithelial tissues is critically important for spatial c...Epithelial and stromal tissues are components of the tumor microenvironment and play a major role in tumor initiation and progression.Distinguishing stroma from epithelial tissues is critically important for spatial characterization of the tumor microenvironment.Here,we propose BrcaSeg,an image analysis pipeline based on a convolutional neural network(CNN)model to classify epithelial and stromal regions in whole-slide hematoxylin and eosin(H&E)stained histopathological images.The CNN model is trained using well-annotated breast cancer tissue microarrays and validated with images from The Cancer Genome Atlas(TCGA)Program.BrcaSeg achieves a classification accuracy of 91.02%,which outperforms other state-of-the-art methods.Using this model,we generate pixel-level epithelial/stromal tissue maps for 1000 TCGA breast cancer slide images that are paired with gene expression data.We subsequently estimate the epithelial and stromal ratios and perform correlation analysis to model the relationship between gene expression and tissue ratios.Gene Ontology(GO)enrichment analyses of genes that are highly correlated with tissue ratios suggest that the same tissue is associated with similar biological processes in different breast cancer subtypes,whereas each subtype also has its own idiosyncratic biological processes governing the development of these tissues.Taken all together,our approach can lead to new insights in exploring relationships between image-based phenotypes and their underlying genomic events and biological processes for all types of solid tumors.BrcaSeg can be accessed at https://github.com/Serian1992/ImgBio.展开更多
文摘目的建立人工智能辅助结直肠癌病理切片分子分型诊断系统。方法在癌症基因组图谱(the cancer genome Atlas,TCGA)数据库中筛选出422例结直肠癌患者的812张病理切片,分为训练集(75%)和测试集(25%);存入www.paiwsit.com数据库中,根据资深的病理医生标注的数据进行处理及分割,得到超过400万张带有标签的训练集,最后利用深度学习模型进行训练。结果在经过多种卷积神经网络模型训练后,在110例203张切片的测试集上测试,子图级别达到53.04%的准确率,切片级别准确率达到51.72%,其中结直肠癌共识分子亚型之一的经典型(CMS2)切片级准确率达到75.00%。结论本研究对促进结直肠癌筛查和精准治疗具有重要意义。
基金supported by the National Key Research and Development Program of China(No.2017YFC1309102)National Natural Science Foundation of China(No.81771912,No.82001986,No.82071892)+1 种基金National Science Fund for Distinguished Young Scholars(No.81925023)High-level Hospital Construction Project(No.DFJH201805 and No.DFJH201914)。
文摘Objective: The Immunoscore method has proved fruitful for predicting prognosis in patients with colon cancer.However, there is still room for improvement in this scoring method to achieve further advances in its clinical translation. This study aimed to develop and validate a modified Immunoscore(IS-mod) system for predicting overall survival(OS) in patients with stage Ⅰ-Ⅲ colon cancer.Methods: The IS-mod was proposed by counting CD3+ and CD8+ immune cells in regions of the tumor core and its invasive margin by drawing two lines of interest. A discovery cohort(N=212) and validation cohort(N=103)from two centers were used to evaluate the prognostic value of the IS-mod.Results: In the discovery cohort, 5-year survival rates were 88.6% in the high IS-mod group and 60.7% in the low IS-mod group. Multivariate analysis confirmed that the IS-mod was an independent prognostic factor for OS[adjusted hazard ratio(HR)=0.36, 95% confidence interval(95% CI): 0.20-0.63]. With less annotation and computation cost, the IS-mod achieved performance comparable to that of the Immunoscore-like(IS-like) system(C-index, 0.676 vs. 0.661, P=0.231). The 2-category IS-mod using 47.5% as the threshold had a better prognostic value than that using a fixed threshold of 25%(C-index, 0.653 vs. 0.573, P=0.004). Similar results were confirmed in the validation cohort.Conclusions: Our method simplifies the annotation and accelerates the calculation of Immunoscore method,thus making it easier for clinical implementation. The IS-mod achieved comparable prognostic performance when compared to the IS-like system in both cohorts. Besides, we further found that even with a small reference set(N≥120), the IS-mod still demonstrated a stable prognostic value. This finding may inspire other institutions to develop a local reference set of an IS-mod system for more accurate risk stratification of colon cancer.
基金supported in part by the Overseas Expertise Introduc-tion Project for Discipline Innovation(Grant No.B17007)the National Natural Science Foundation of China(Grant No.81972248)+1 种基金the Natural Science Foundation of Beijing Municipality(Grant No.7202056)by the Beijing Municipal Administration of Hospitals Incubating Program(Grant No.PX2021013).
文摘Background The prevalence of thyroid cancer is growing rapidly.Early and precise diagnosis is critical in thy-roid cancer caring.An automatic thyroid cancer diagnostic tool can be valuable to achieve early detection and diagnostic consistency.Only the follicular areas in the sample contain useful information to the thyroid cancer diagnosis based on fine needle aspiration(FNA).This study aimed to develop a highly efficient accurate method for follicular cell areas segmentation(FCAS)of thyroid cytopathological whole slide images(WSIs).Methods A total of 96 cell samples from July 2017 to July 2018 were collected in one hospital in Beijing,China.Forty-three WSIs were selected and manually labeled,including 17 cases of papillary thyroid carci-noma sample and 26 cases of benign sample.Six thousand and nine hundred cropped typical image patches(available on https://github.com/bupt-ai-cz/Hybrid-Model-Enabling-Highly-Efficient-Follicular-Segmentation)of 1024×1024 pixels from 13 large WSIs were used for patch-level model training and testing and all of the 13 large WSIs were papillary thyroid carcinoma samples.Thirty testing WSIs with an average size 36,217×29,400(from 10,240×10,240 to 81,920×61,440)were used to test the effectiveness of the hybrid model.Based on the traditional semantic segmentation model deeplabv3,we constructed a hybrid segmentation architecture by adding a classification branch into the segmentation scheme to improve efficiency.Accuracy was used to measure the performance of the classification model;pixel accuracy(pAcc),mean accuracy(mAcc),mean intersection over union(mIoU),and frequency weighted intersection over union(fwIoU)were used to measure the performance of the segmentation model,respectively.Results Using this method,up to 93%WSI segmentation time was reduced by skipping the colloidal areas and the blank background areas.The average processing time of 30 WSI was 49.49 s.On the patch dataset,this hybrid model might reach pAcc=98.65%,mAcc=85.60%,mIoU=79.61%,and fwIoU=97.54%.On the WSI datase
基金supported by Indiana University Precision Health Initiative to KH and JZthe NSFC-Guangdong Joint Fund of China (Grant No. U1501256) to QFShenzhen Peacock Plan (Grant No. KQTD2016053112051497) to XZ and ND.
文摘Epithelial and stromal tissues are components of the tumor microenvironment and play a major role in tumor initiation and progression.Distinguishing stroma from epithelial tissues is critically important for spatial characterization of the tumor microenvironment.Here,we propose BrcaSeg,an image analysis pipeline based on a convolutional neural network(CNN)model to classify epithelial and stromal regions in whole-slide hematoxylin and eosin(H&E)stained histopathological images.The CNN model is trained using well-annotated breast cancer tissue microarrays and validated with images from The Cancer Genome Atlas(TCGA)Program.BrcaSeg achieves a classification accuracy of 91.02%,which outperforms other state-of-the-art methods.Using this model,we generate pixel-level epithelial/stromal tissue maps for 1000 TCGA breast cancer slide images that are paired with gene expression data.We subsequently estimate the epithelial and stromal ratios and perform correlation analysis to model the relationship between gene expression and tissue ratios.Gene Ontology(GO)enrichment analyses of genes that are highly correlated with tissue ratios suggest that the same tissue is associated with similar biological processes in different breast cancer subtypes,whereas each subtype also has its own idiosyncratic biological processes governing the development of these tissues.Taken all together,our approach can lead to new insights in exploring relationships between image-based phenotypes and their underlying genomic events and biological processes for all types of solid tumors.BrcaSeg can be accessed at https://github.com/Serian1992/ImgBio.
