In recent years there has been significant developments in photosensitizers(PSs), light sources and light delivery systems that have allowed decreasing the treatment time and skin phototoxicity resulting in more frequ...In recent years there has been significant developments in photosensitizers(PSs), light sources and light delivery systems that have allowed decreasing the treatment time and skin phototoxicity resulting in more frequent use of photodynamic therapy(PDT) in the clinical settings. Compared to standard treatment approaches such as chemo-radiation and surgery, PDT has much reduced morbidity for head and neck malignancies and is becoming an alternative treatment option. It can be used as an adjunct therapy to other treatment modalities without any additive cumulative side effects. Surface illumination can be an option for pre-malignant and early-stage malignancies while interstitial treatment is for debulking of thick tumors in the head and neck region. PDT can achieve equivalent or greater efficacy in treating head and neck malignancies, suggesting that it may be considered as a first line therapy in the future. Despite progressive development, clinical PDT needs improvement in several topics for wider acceptance including standardization of protocols that involve the same administrated light and PS doses and establishing quantitative tools for PDT dosimetry planning and response monitoring. Quantitative measures such as optical parameters, PS concentration, tissueoxygenation and blood flow are essential for accurate PDT dosimetry as well as PDT response monitoring and assessing therapy outcome. Unlike conventional imaging modalities like magnetic resonance imaging, novel optical imaging techniques can quantify PDT-related parameters without any contrast agent administration and enable real-time assessment during PDT for providing fast feedback to clinicians. Ongoing developments in optical imaging offer the promise of optimization of PDT protocols with improved outcomes.展开更多
交通流量因受周期性特征、突发状况等多重因素影响,现有模型的预测精度无法满足实际要求.对此,本文提出了基于误差补偿的多模态协同交通流预测模型(Multimodal Collaborative traffic flow prediction model based on Error Compensatio...交通流量因受周期性特征、突发状况等多重因素影响,现有模型的预测精度无法满足实际要求.对此,本文提出了基于误差补偿的多模态协同交通流预测模型(Multimodal Collaborative traffic flow prediction model based on Error Compensation,MCEC).针对传统预测模型不能兼顾时间序列和协变量的问题,提出基于小波分析的特征拓展方法,该方法引入聚类算法得到节假日标签特征,将拥堵指数、交通事故图、天气信息作为拓展特征,对特征进行多尺度分解.在训练阶段,为达到充分学习各部分数据、最优匹配模型的效果,采用差分整合移动平均自回归模型(Autoreg Ressive Integrated Moving Average Model,ARIMA)、长短期记忆神经网络(Long Short-Term Memory network,LSTM)、限制动态时间规整技术(Dynamic Time Warping,DTW)以及自注意力机制(Self-Attention),设计了多模态协同模型训练.在误差补偿阶段,将得到的相应过程值输入基于支持向量机回归(Support Vector Regression,SVR)的误差补偿模块,对各分量的误差进行学习、补偿,并重构得到预测结果.使用公开的高速公路数据集对MCEC进行验证,在多个时间间隔下对比实验结果表明,MCEC在交通流量预测中的平均绝对百分比误差(Mean Absolute Percentage Error,MAPE)达到17.02%,比LSTM-SVR、ConvLSTM(Convolutional Long Short-Term Memory network)、ST-GCN(Spatial Temporal Graph Convolutional Networks)、MFFB(Multi-stream Feature Fusion Block)、Transformer等预测模型具有更高的预测精度,MCEC模型具有较好的有效性与合理性.展开更多
基金Supported by RPCI Startup Grant(P30CA16056)NCI CA55791
文摘In recent years there has been significant developments in photosensitizers(PSs), light sources and light delivery systems that have allowed decreasing the treatment time and skin phototoxicity resulting in more frequent use of photodynamic therapy(PDT) in the clinical settings. Compared to standard treatment approaches such as chemo-radiation and surgery, PDT has much reduced morbidity for head and neck malignancies and is becoming an alternative treatment option. It can be used as an adjunct therapy to other treatment modalities without any additive cumulative side effects. Surface illumination can be an option for pre-malignant and early-stage malignancies while interstitial treatment is for debulking of thick tumors in the head and neck region. PDT can achieve equivalent or greater efficacy in treating head and neck malignancies, suggesting that it may be considered as a first line therapy in the future. Despite progressive development, clinical PDT needs improvement in several topics for wider acceptance including standardization of protocols that involve the same administrated light and PS doses and establishing quantitative tools for PDT dosimetry planning and response monitoring. Quantitative measures such as optical parameters, PS concentration, tissueoxygenation and blood flow are essential for accurate PDT dosimetry as well as PDT response monitoring and assessing therapy outcome. Unlike conventional imaging modalities like magnetic resonance imaging, novel optical imaging techniques can quantify PDT-related parameters without any contrast agent administration and enable real-time assessment during PDT for providing fast feedback to clinicians. Ongoing developments in optical imaging offer the promise of optimization of PDT protocols with improved outcomes.
