Pub Date : 2024-08-20DOI: 10.1007/s10278-024-01178-8
Weiguo Cao, Marc J Pomeroy, Zhengrong Liang, Yongfeng Gao, Yongyi Shi, Jiaxing Tan, Fangfang Han, Jing Wang, Jianhua Ma, Hongbin Lu, Almas F Abbasi, Perry J Pickhardt
The elasticity of soft tissues has been widely considered a characteristic property for differentiation of healthy and lesions and, therefore, motivated the development of several elasticity imaging modalities, for example, ultrasound elastography, magnetic resonance elastography, and optical coherence elastography to directly measure the tissue elasticity. This paper proposes an alternative approach of modeling the elasticity for prior knowledge-based extraction of tissue elastic characteristic features for machine learning (ML) lesion classification using computed tomography (CT) imaging modality. The model describes a dynamic non-rigid (or elastic) soft tissue deformation in differential manifold to mimic the tissues' elasticity under wave fluctuation in vivo. Based on the model, a local deformation invariant is formulated using the 1st and 2nd order derivatives of the lesion volumetric CT image and used to generate elastic feature map of the lesion volume. From the feature map, tissue elastic features are extracted and fed to ML to perform lesion classification. Two pathologically proven image datasets of colon polyps and lung nodules were used to test the modeling strategy. The outcomes reached the score of area under the curve of receiver operating characteristics of 94.2% for the polyps and 87.4% for the nodules, resulting in an average gain of 5 to 20% over several existing state-of-the-art image feature-based lesion classification methods. The gain demonstrates the importance of extracting tissue characteristic features for lesion classification, instead of extracting image features, which can include various image artifacts and may vary for different protocols in image acquisition and different imaging modalities.
软组织的弹性被广泛认为是区分健康组织和病变组织的一个特征特性,因此,人们开发了多种弹性成像模式,如超声弹性成像、磁共振弹性成像和光学相干弹性成像,以直接测量组织弹性。本文提出了一种弹性建模的替代方法,以基于先验知识提取组织弹性特征,用于使用计算机断层扫描(CT)成像模式进行机器学习(ML)病变分类。该模型在差分流形中描述了非刚性(或弹性)软组织的动态形变,以模拟组织在活体波波动下的弹性。根据该模型,利用病变容积 CT 图像的一阶和二阶导数制定了局部变形不变量,并用于生成病变容积的弹性特征图。从特征图中提取组织弹性特征,并将其输入 ML 以进行病变分类。结肠息肉和肺结节这两个病理证实的图像数据集被用来测试建模策略。结果显示,息肉的接收者操作特征曲线下面积得分率为 94.2%,结节的接收者操作特征曲线下面积得分率为 87.4%,与现有的几种最先进的基于图像特征的病变分类方法相比,平均增益 5% 至 20%。这种增益表明了提取组织特征对病变分类的重要性,而不是提取图像特征,因为图像特征可能包括各种图像伪影,而且在不同的图像采集协议和不同的成像模式下可能会有所不同。
{"title":"Lesion Classification by Model-Based Feature Extraction: A Differential Affine Invariant Model of Soft Tissue Elasticity in CT Images.","authors":"Weiguo Cao, Marc J Pomeroy, Zhengrong Liang, Yongfeng Gao, Yongyi Shi, Jiaxing Tan, Fangfang Han, Jing Wang, Jianhua Ma, Hongbin Lu, Almas F Abbasi, Perry J Pickhardt","doi":"10.1007/s10278-024-01178-8","DOIUrl":"https://doi.org/10.1007/s10278-024-01178-8","url":null,"abstract":"<p><p>The elasticity of soft tissues has been widely considered a characteristic property for differentiation of healthy and lesions and, therefore, motivated the development of several elasticity imaging modalities, for example, ultrasound elastography, magnetic resonance elastography, and optical coherence elastography to directly measure the tissue elasticity. This paper proposes an alternative approach of modeling the elasticity for prior knowledge-based extraction of tissue elastic characteristic features for machine learning (ML) lesion classification using computed tomography (CT) imaging modality. The model describes a dynamic non-rigid (or elastic) soft tissue deformation in differential manifold to mimic the tissues' elasticity under wave fluctuation in vivo. Based on the model, a local deformation invariant is formulated using the 1<sup>st</sup> and 2<sup>nd</sup> order derivatives of the lesion volumetric CT image and used to generate elastic feature map of the lesion volume. From the feature map, tissue elastic features are extracted and fed to ML to perform lesion classification. Two pathologically proven image datasets of colon polyps and lung nodules were used to test the modeling strategy. The outcomes reached the score of area under the curve of receiver operating characteristics of 94.2% for the polyps and 87.4% for the nodules, resulting in an average gain of 5 to 20% over several existing state-of-the-art image feature-based lesion classification methods. The gain demonstrates the importance of extracting tissue characteristic features for lesion classification, instead of extracting image features, which can include various image artifacts and may vary for different protocols in image acquisition and different imaging modalities.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142010199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In clinical practice, the anatomical classification of pulmonary veins plays a crucial role in the preoperative assessment of atrial fibrillation radiofrequency ablation surgery. Accurate classification of pulmonary vein anatomy assists physicians in selecting appropriate mapping electrodes and avoids causing pulmonary arterial hypertension. Due to the diverse and subtly different anatomical classifications of pulmonary veins, as well as the imbalance in data distribution, deep learning models often exhibit poor expression capability in extracting deep features, leading to misjudgments and affecting classification accuracy. Therefore, in order to solve the problem of unbalanced classification of left atrial pulmonary veins, this paper proposes a network integrating multi-scale feature-enhanced attention and dual-feature extraction classifiers, called DECNet. The multi-scale feature-enhanced attention utilizes multi-scale information to guide the reinforcement of deep features, generating channel weights and spatial weights to enhance the expression capability of deep features. The dual-feature extraction classifier assigns a fixed number of channels to each category, equally evaluating all categories, thus alleviating the learning bias and overfitting caused by data imbalance. By combining the two, the expression capability of deep features is strengthened, achieving accurate classification of left atrial pulmonary vein morphology and providing support for subsequent clinical treatment. The proposed method is evaluated on datasets provided by the People's Hospital of Liaoning Province and the publicly available DermaMNIST dataset, achieving average accuracies of 78.81% and 83.44%, respectively, demonstrating the effectiveness of the proposed approach.
{"title":"DECNet: Left Atrial Pulmonary Vein Class Imbalance Classification Network.","authors":"GuoDong Zhang, WenWen Gu, TingYu Liang, YanLin Li, Wei Guo, ZhaoXuan Gong, RongHui Ju","doi":"10.1007/s10278-024-01221-8","DOIUrl":"https://doi.org/10.1007/s10278-024-01221-8","url":null,"abstract":"<p><p>In clinical practice, the anatomical classification of pulmonary veins plays a crucial role in the preoperative assessment of atrial fibrillation radiofrequency ablation surgery. Accurate classification of pulmonary vein anatomy assists physicians in selecting appropriate mapping electrodes and avoids causing pulmonary arterial hypertension. Due to the diverse and subtly different anatomical classifications of pulmonary veins, as well as the imbalance in data distribution, deep learning models often exhibit poor expression capability in extracting deep features, leading to misjudgments and affecting classification accuracy. Therefore, in order to solve the problem of unbalanced classification of left atrial pulmonary veins, this paper proposes a network integrating multi-scale feature-enhanced attention and dual-feature extraction classifiers, called DECNet. The multi-scale feature-enhanced attention utilizes multi-scale information to guide the reinforcement of deep features, generating channel weights and spatial weights to enhance the expression capability of deep features. The dual-feature extraction classifier assigns a fixed number of channels to each category, equally evaluating all categories, thus alleviating the learning bias and overfitting caused by data imbalance. By combining the two, the expression capability of deep features is strengthened, achieving accurate classification of left atrial pulmonary vein morphology and providing support for subsequent clinical treatment. The proposed method is evaluated on datasets provided by the People's Hospital of Liaoning Province and the publicly available DermaMNIST dataset, achieving average accuracies of 78.81% and 83.44%, respectively, demonstrating the effectiveness of the proposed approach.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142010198","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-20DOI: 10.1007/s10278-024-01147-1
Mitchell Goldburgh, Michael LaChance, Julia Komissarchik, Julia Patriarche, Joe Chapa, Oliver Chen, Priya Deshpande, Matthew Geeslin, Julia Komissarchik, Nina Kottler, Julia Patriarche, Jennifer Sommer, Marcus Ayers, Vedrana Vujic
This SIIM-sponsored 2023 report highlights an industry view on artificial intelligence adoption barriers and success related to diagnostic imaging, life sciences, and contrasts. In general, our 2023 survey indicates that there has been progress in adopting AI across multiple uses, and there continues to be an optimistic forecast for the impact on workflow and clinical outcomes. This report, as in prior years, should be seen as a snapshot of the use of AI in imaging. Compared to our 2021 survey, the 2023 respondents expressed wider AI adoption but felt this was behind the potential. Specifically, the adoption has increased as sources of return on investment with AI in radiology are better understood as documented by vendor/client use case studies. Generally, the discussions of AI solutions centered on workflow triage, visualization, detection, and characterization. Generative AI was also mentioned for improving productivity in reporting. As payor reimbursement remains elusive, the ROI discussions expanded to look at other factors, including increased hospital procedures and admissions, enhanced radiologist productivity for practices, and improved patient outcomes for integrated health networks. When looking at the longer-term horizon for AI adoption, respondents frequently mentioned that the opportunity for AI to achieve greater adoption with more complex AI and a more manageable/visible ROI is outside the USA. Respondents focused on the barriers to trust in AI and the FDA processes.
这份由 SIIM 赞助的 2023 年报告重点介绍了与诊断成像、生命科学和对比度有关的人工智能应用障碍和成功的行业观点。总体而言,我们的 2023 年调查表明,在多种用途中采用人工智能方面取得了进展,而且对人工智能对工作流程和临床结果的影响仍持乐观预测。与往年一样,本报告应被视为人工智能在成像领域应用的一个缩影。与 2021 年的调查相比,2023 年的受访者表示更广泛地采用了人工智能,但他们认为这还远远不够。具体来说,随着人们对人工智能在放射学中的投资回报来源有了更好的了解,供应商/客户的使用案例研究也证明了这一点。一般来说,人工智能解决方案的讨论集中在工作流程分流、可视化、检测和特征描述方面。此外,还提到了用于提高报告效率的生成性人工智能。由于支付方的报销仍然遥遥无期,投资回报率的讨论扩展到了其他因素,包括医院手术和入院人数的增加、放射医师工作效率的提高以及综合医疗网络患者治疗效果的改善。在展望人工智能应用的长远前景时,受访者经常提到,人工智能在美国以外的地区有机会通过更复杂的人工智能和更易于管理/可见的投资回报率获得更广泛的应用。受访者关注的重点是人工智能和食品与药物管理局流程的信任障碍。
{"title":"2023 Industry Perceptions Survey on AI Adoption and Return on Investment.","authors":"Mitchell Goldburgh, Michael LaChance, Julia Komissarchik, Julia Patriarche, Joe Chapa, Oliver Chen, Priya Deshpande, Matthew Geeslin, Julia Komissarchik, Nina Kottler, Julia Patriarche, Jennifer Sommer, Marcus Ayers, Vedrana Vujic","doi":"10.1007/s10278-024-01147-1","DOIUrl":"https://doi.org/10.1007/s10278-024-01147-1","url":null,"abstract":"<p><p>This SIIM-sponsored 2023 report highlights an industry view on artificial intelligence adoption barriers and success related to diagnostic imaging, life sciences, and contrasts. In general, our 2023 survey indicates that there has been progress in adopting AI across multiple uses, and there continues to be an optimistic forecast for the impact on workflow and clinical outcomes. This report, as in prior years, should be seen as a snapshot of the use of AI in imaging. Compared to our 2021 survey, the 2023 respondents expressed wider AI adoption but felt this was behind the potential. Specifically, the adoption has increased as sources of return on investment with AI in radiology are better understood as documented by vendor/client use case studies. Generally, the discussions of AI solutions centered on workflow triage, visualization, detection, and characterization. Generative AI was also mentioned for improving productivity in reporting. As payor reimbursement remains elusive, the ROI discussions expanded to look at other factors, including increased hospital procedures and admissions, enhanced radiologist productivity for practices, and improved patient outcomes for integrated health networks. When looking at the longer-term horizon for AI adoption, respondents frequently mentioned that the opportunity for AI to achieve greater adoption with more complex AI and a more manageable/visible ROI is outside the USA. Respondents focused on the barriers to trust in AI and the FDA processes.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142010196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-20DOI: 10.1007/s10278-024-01151-5
Mohammadreza Zandehshahvar, Marly van Assen, Eun Kim, Yashar Kiarashi, Vikranth Keerthipati, Giovanni Tessarin, Emanuele Muscogiuri, Arthur E Stillman, Peter Filev, Amir H Davarpanah, Eugene A Berkowitz, Stefan Tigges, Scott J Lee, Brianna L Vey, Carlo De Cecco, Ali Adibi
In this study, we present a method based on Monte Carlo Dropout (MCD) as Bayesian neural network (BNN) approximation for confidence-aware severity classification of lung diseases in COVID-19 patients using chest X-rays (CXRs). Trained and tested on 1208 CXRs from Hospital 1 in the USA, the model categorizes severity into four levels (i.e., normal, mild, moderate, and severe) based on lung consolidation and opacity. Severity labels, determined by the median consensus of five radiologists, serve as the reference standard. The model's performance is internally validated against evaluations from an additional radiologist and two residents that were excluded from the median. The performance of the model is further evaluated on additional internal and external datasets comprising 2200 CXRs from the same hospital and 1300 CXRs from Hospital 2 in South Korea. The model achieves an average area under the curve (AUC) of 0.94 ± 0.01 across all classes in the primary dataset, surpassing human readers in each severity class and achieves a higher Kendall correlation coefficient (KCC) of 0.80 ± 0.03. The performance of the model is consistent across varied datasets, highlighting its generalization. A key aspect of the model is its predictive uncertainty (PU), which is inversely related to the level of agreement among radiologists, particularly in mild and moderate cases. The study concludes that the model outperforms human readers in severity assessment and maintains consistent accuracy across diverse datasets. Its ability to provide confidence measures in predictions is pivotal for potential clinical use, underscoring the BNN's role in enhancing diagnostic precision in lung disease analysis through CXR.
{"title":"Confidence-Aware Severity Assessment of Lung Disease from Chest X-Rays Using Deep Neural Network on a Multi-Reader Dataset.","authors":"Mohammadreza Zandehshahvar, Marly van Assen, Eun Kim, Yashar Kiarashi, Vikranth Keerthipati, Giovanni Tessarin, Emanuele Muscogiuri, Arthur E Stillman, Peter Filev, Amir H Davarpanah, Eugene A Berkowitz, Stefan Tigges, Scott J Lee, Brianna L Vey, Carlo De Cecco, Ali Adibi","doi":"10.1007/s10278-024-01151-5","DOIUrl":"https://doi.org/10.1007/s10278-024-01151-5","url":null,"abstract":"<p><p>In this study, we present a method based on Monte Carlo Dropout (MCD) as Bayesian neural network (BNN) approximation for confidence-aware severity classification of lung diseases in COVID-19 patients using chest X-rays (CXRs). Trained and tested on 1208 CXRs from Hospital 1 in the USA, the model categorizes severity into four levels (i.e., normal, mild, moderate, and severe) based on lung consolidation and opacity. Severity labels, determined by the median consensus of five radiologists, serve as the reference standard. The model's performance is internally validated against evaluations from an additional radiologist and two residents that were excluded from the median. The performance of the model is further evaluated on additional internal and external datasets comprising 2200 CXRs from the same hospital and 1300 CXRs from Hospital 2 in South Korea. The model achieves an average area under the curve (AUC) of 0.94 ± 0.01 across all classes in the primary dataset, surpassing human readers in each severity class and achieves a higher Kendall correlation coefficient (KCC) of 0.80 ± 0.03. The performance of the model is consistent across varied datasets, highlighting its generalization. A key aspect of the model is its predictive uncertainty (PU), which is inversely related to the level of agreement among radiologists, particularly in mild and moderate cases. The study concludes that the model outperforms human readers in severity assessment and maintains consistent accuracy across diverse datasets. Its ability to provide confidence measures in predictions is pivotal for potential clinical use, underscoring the BNN's role in enhancing diagnostic precision in lung disease analysis through CXR.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142010197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Anemia is a significant global health issue, affecting over a billion people worldwide, according to the World Health Organization. Generally, the gold standard for diagnosing anemia relies on laboratory measurements of hemoglobin. To meet the need in clinical practice, physicians often rely on visual examination of specific areas, such as conjunctiva, to assess pallor. However, this method is subjective and relies on the physician's experience. Therefore, we proposed a deep learning prediction model based on three input images from different body parts, namely, conjunctiva, palm, and fingernail. By incorporating additional body part labels and employing a fusion attention mechanism, the model learns and enhances the salient features of each body part during training, enabling it to produce reliable results. Additionally, we employ a dual loss function that allows the regression model to benefit from well-established classification methods, thereby achieving stable handling of minority samples. We used a retrospective data set (EYES-DEFY-ANEMIA) to develop this model called Body-Part-Anemia Network (BPANet). The BPANet showed excellent performance in detecting anemia, with accuracy of 0.849 and an F1-score of 0.828. Our multi-body-part model has been validated on a prospectively collected data set of 101 patients in National Taiwan University Hospital. The prediction accuracy as well as F1-score can achieve as high as 0.716 and 0.788, respectively. To sum up, we have developed and validated a novel non-invasive hemoglobin prediction model based on image input from multiple body parts, with the potential of real-time use at home and in clinical settings.
{"title":"Deep Learning-Based Model for Non-invasive Hemoglobin Estimation via Body Parts Images: A Retrospective Analysis and a Prospective Emergency Department Study.","authors":"En-Ting Lin, Shao-Chi Lu, An-Sheng Liu, Chia-Hsin Ko, Chien-Hua Huang, Chu-Lin Tsai, Li-Chen Fu","doi":"10.1007/s10278-024-01209-4","DOIUrl":"https://doi.org/10.1007/s10278-024-01209-4","url":null,"abstract":"<p><p>Anemia is a significant global health issue, affecting over a billion people worldwide, according to the World Health Organization. Generally, the gold standard for diagnosing anemia relies on laboratory measurements of hemoglobin. To meet the need in clinical practice, physicians often rely on visual examination of specific areas, such as conjunctiva, to assess pallor. However, this method is subjective and relies on the physician's experience. Therefore, we proposed a deep learning prediction model based on three input images from different body parts, namely, conjunctiva, palm, and fingernail. By incorporating additional body part labels and employing a fusion attention mechanism, the model learns and enhances the salient features of each body part during training, enabling it to produce reliable results. Additionally, we employ a dual loss function that allows the regression model to benefit from well-established classification methods, thereby achieving stable handling of minority samples. We used a retrospective data set (EYES-DEFY-ANEMIA) to develop this model called Body-Part-Anemia Network (BPANet). The BPANet showed excellent performance in detecting anemia, with accuracy of 0.849 and an F1-score of 0.828. Our multi-body-part model has been validated on a prospectively collected data set of 101 patients in National Taiwan University Hospital. The prediction accuracy as well as F1-score can achieve as high as 0.716 and 0.788, respectively. To sum up, we have developed and validated a novel non-invasive hemoglobin prediction model based on image input from multiple body parts, with the potential of real-time use at home and in clinical settings.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142006261","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Expert feedback on trainees' preliminary reports is crucial for radiologic training, but real-time feedback can be challenging due to non-contemporaneous, remote reading and increasing imaging volumes. Trainee report revisions contain valuable educational feedback, but synthesizing data from raw revisions is challenging. Generative AI models can potentially analyze these revisions and provide structured, actionable feedback. This study used the OpenAI GPT-4 Turbo API to analyze paired synthesized and open-source analogs of preliminary and finalized reports, identify discrepancies, categorize their severity and type, and suggest review topics. Expert radiologists reviewed the output by grading discrepancies, evaluating the severity and category accuracy, and suggested review topic relevance. The reproducibility of discrepancy detection and maximal discrepancy severity was also examined. The model exhibited high sensitivity, detecting significantly more discrepancies than radiologists (W = 19.0, p < 0.001) with a strong positive correlation (r = 0.778, p < 0.001). Interrater reliability for severity and type were fair (Fleiss' kappa = 0.346 and 0.340, respectively; weighted kappa = 0.622 for severity). The LLM achieved a weighted F1 score of 0.66 for severity and 0.64 for type. Generated teaching points were considered relevant in ~ 85% of cases, and relevance correlated with the maximal discrepancy severity (Spearman ρ = 0.76, p < 0.001). The reproducibility was moderate to good (ICC (2,1) = 0.690) for the number of discrepancies and substantial for maximal discrepancy severity (Fleiss' kappa = 0.718; weighted kappa = 0.94). Generative AI models can effectively identify discrepancies in report revisions and generate relevant educational feedback, offering promise for enhancing radiology training.
专家对学员初步报告的反馈意见对放射学培训至关重要,但由于非同步、远程阅片和成像量不断增加,实时反馈可能具有挑战性。学员报告的修改包含宝贵的教育反馈,但从原始修改中综合数据是一项挑战。生成式人工智能模型有可能分析这些修订,并提供结构化、可操作的反馈。本研究使用 OpenAI GPT-4 Turbo API 分析初步报告和最终报告的配对合成和开源模拟,识别差异,对其严重程度和类型进行分类,并提出审查主题。放射科专家通过对差异进行分级、评估严重程度和类别的准确性以及建议审查主题的相关性来审查输出结果。此外,还对差异检测的再现性和最大差异严重程度进行了检查。该模型表现出很高的灵敏度,检测到的差异明显多于放射科医生(W = 19.0,p
{"title":"From Revisions to Insights: Converting Radiology Report Revisions into Actionable Educational Feedback Using Generative AI Models.","authors":"Shawn Lyo, Suyash Mohan, Alvand Hassankhani, Abass Noor, Farouk Dako, Tessa Cook","doi":"10.1007/s10278-024-01233-4","DOIUrl":"https://doi.org/10.1007/s10278-024-01233-4","url":null,"abstract":"<p><p>Expert feedback on trainees' preliminary reports is crucial for radiologic training, but real-time feedback can be challenging due to non-contemporaneous, remote reading and increasing imaging volumes. Trainee report revisions contain valuable educational feedback, but synthesizing data from raw revisions is challenging. Generative AI models can potentially analyze these revisions and provide structured, actionable feedback. This study used the OpenAI GPT-4 Turbo API to analyze paired synthesized and open-source analogs of preliminary and finalized reports, identify discrepancies, categorize their severity and type, and suggest review topics. Expert radiologists reviewed the output by grading discrepancies, evaluating the severity and category accuracy, and suggested review topic relevance. The reproducibility of discrepancy detection and maximal discrepancy severity was also examined. The model exhibited high sensitivity, detecting significantly more discrepancies than radiologists (W = 19.0, p < 0.001) with a strong positive correlation (r = 0.778, p < 0.001). Interrater reliability for severity and type were fair (Fleiss' kappa = 0.346 and 0.340, respectively; weighted kappa = 0.622 for severity). The LLM achieved a weighted F1 score of 0.66 for severity and 0.64 for type. Generated teaching points were considered relevant in ~ 85% of cases, and relevance correlated with the maximal discrepancy severity (Spearman ρ = 0.76, p < 0.001). The reproducibility was moderate to good (ICC (2,1) = 0.690) for the number of discrepancies and substantial for maximal discrepancy severity (Fleiss' kappa = 0.718; weighted kappa = 0.94). Generative AI models can effectively identify discrepancies in report revisions and generate relevant educational feedback, offering promise for enhancing radiology training.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142006262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-16DOI: 10.1007/s10278-024-01107-9
Chongxuan Tian, Yue Xi, Yuting Ma, Cai Chen, Cong Wu, Kun Ru, Wei Li, Miaoqing Zhao
Primary diffuse central nervous system large B-cell lymphoma (CNS-pDLBCL) and high-grade glioma (HGG) often present similarly, clinically and on imaging, making differentiation challenging. This similarity can complicate pathologists' diagnostic efforts, yet accurately distinguishing between these conditions is crucial for guiding treatment decisions. This study leverages a deep learning model to classify brain tumor pathology images, addressing the common issue of limited medical imaging data. Instead of training a convolutional neural network (CNN) from scratch, we employ a pre-trained network for extracting deep features, which are then used by a support vector machine (SVM) for classification. Our evaluation shows that the Resnet50 (TL + SVM) model achieves a 97.4% accuracy, based on tenfold cross-validation on the test set. These results highlight the synergy between deep learning and traditional diagnostics, potentially setting a new standard for accuracy and efficiency in the pathological diagnosis of brain tumors.
原发性弥漫性中枢神经系统大 B 细胞淋巴瘤(CNS-pDLBCL)和高级别胶质瘤(HGG)在临床上和影像学上的表现往往很相似,因此很难区分。这种相似性会使病理学家的诊断工作复杂化,但准确区分这些病症对于指导治疗决策至关重要。本研究利用深度学习模型对脑肿瘤病理图像进行分类,解决了医学影像数据有限这一常见问题。我们没有从头开始训练卷积神经网络(CNN),而是采用了一个预先训练好的网络来提取深度特征,然后由支持向量机(SVM)进行分类。我们的评估结果表明,基于测试集上的十倍交叉验证,Resnet50(TL + SVM)模型达到了 97.4% 的准确率。这些结果凸显了深度学习与传统诊断之间的协同作用,有可能为脑肿瘤病理诊断的准确性和效率设定一个新标准。
{"title":"Harnessing Deep Learning for Accurate Pathological Assessment of Brain Tumor Cell Types.","authors":"Chongxuan Tian, Yue Xi, Yuting Ma, Cai Chen, Cong Wu, Kun Ru, Wei Li, Miaoqing Zhao","doi":"10.1007/s10278-024-01107-9","DOIUrl":"https://doi.org/10.1007/s10278-024-01107-9","url":null,"abstract":"<p><p>Primary diffuse central nervous system large B-cell lymphoma (CNS-pDLBCL) and high-grade glioma (HGG) often present similarly, clinically and on imaging, making differentiation challenging. This similarity can complicate pathologists' diagnostic efforts, yet accurately distinguishing between these conditions is crucial for guiding treatment decisions. This study leverages a deep learning model to classify brain tumor pathology images, addressing the common issue of limited medical imaging data. Instead of training a convolutional neural network (CNN) from scratch, we employ a pre-trained network for extracting deep features, which are then used by a support vector machine (SVM) for classification. Our evaluation shows that the Resnet50 (TL + SVM) model achieves a 97.4% accuracy, based on tenfold cross-validation on the test set. These results highlight the synergy between deep learning and traditional diagnostics, potentially setting a new standard for accuracy and efficiency in the pathological diagnosis of brain tumors.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141989827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate segmentation of skin lesions in dermoscopic images is of key importance for quantitative analysis of melanoma. Although existing medical image segmentation methods significantly improve skin lesion segmentation, they still have limitations in extracting local features with global information, do not handle challenging lesions well, and usually have a large number of parameters and high computational complexity. To address these issues, this paper proposes an efficient adaptive attention and convolutional fusion network for skin lesion segmentation (EAAC-Net). We designed two parallel encoders, where the efficient adaptive attention feature extraction module (EAAM) adaptively establishes global spatial dependence and global channel dependence by constructing the adjacency matrix of the directed graph and can adaptively filter out the least relevant tokens at the coarse-grained region level, thus reducing the computational complexity of the self-attention mechanism. The efficient multiscale attention-based convolution module (EMA⋅C) utilizes multiscale attention for cross-space learning of local features extracted from the convolutional layer to enhance the representation of richly detailed local features. In addition, we designed a reverse attention feature fusion module (RAFM) to enhance the effective boundary information gradually. To validate the performance of our proposed network, we compared it with other methods on ISIC 2016, ISIC 2018, and PH2 public datasets, and the experimental results show that EAAC-Net has superior segmentation performance under commonly used evaluation metrics.
{"title":"EAAC-Net: An Efficient Adaptive Attention and Convolution Fusion Network for Skin Lesion Segmentation.","authors":"Chao Fan, Zhentong Zhu, Bincheng Peng, Zhihui Xuan, Xinru Zhu","doi":"10.1007/s10278-024-01223-6","DOIUrl":"https://doi.org/10.1007/s10278-024-01223-6","url":null,"abstract":"<p><p>Accurate segmentation of skin lesions in dermoscopic images is of key importance for quantitative analysis of melanoma. Although existing medical image segmentation methods significantly improve skin lesion segmentation, they still have limitations in extracting local features with global information, do not handle challenging lesions well, and usually have a large number of parameters and high computational complexity. To address these issues, this paper proposes an efficient adaptive attention and convolutional fusion network for skin lesion segmentation (EAAC-Net). We designed two parallel encoders, where the efficient adaptive attention feature extraction module (EAAM) adaptively establishes global spatial dependence and global channel dependence by constructing the adjacency matrix of the directed graph and can adaptively filter out the least relevant tokens at the coarse-grained region level, thus reducing the computational complexity of the self-attention mechanism. The efficient multiscale attention-based convolution module (EMA⋅C) utilizes multiscale attention for cross-space learning of local features extracted from the convolutional layer to enhance the representation of richly detailed local features. In addition, we designed a reverse attention feature fusion module (RAFM) to enhance the effective boundary information gradually. To validate the performance of our proposed network, we compared it with other methods on ISIC 2016, ISIC 2018, and PH<sup>2</sup> public datasets, and the experimental results show that EAAC-Net has superior segmentation performance under commonly used evaluation metrics.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141989826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-15DOI: 10.1007/s10278-024-01231-6
Yueyan Wang, Aiqi Chen, Kai Wang, Yihui Zhao, Xiaomeng Du, Yan Chen, Lei Lv, Yimin Huang, Yichuan Ma
This study aimed to establish and validate the efficacy of a nomogram model, synthesized through the integration of multi-parametric magnetic resonance radiomics and clinical risk factors, for forecasting perineural invasion in rectal cancer. We retrospectively collected data from 108 patients with pathologically confirmed rectal adenocarcinoma who underwent preoperative multiparametric MRI at the First Affiliated Hospital of Bengbu Medical College between April 2019 and August 2023. This dataset was subsequently divided into training and validation sets following a ratio of 7:3. Both univariate and multivariate logistic regression analyses were implemented to identify independent clinical risk factors associated with perineural invasion (PNI) in rectal cancer. We manually delineated the region of interest (ROI) layer-by-layer on T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) sequences and extracted the image features. Five machine learning algorithms were used to construct radiomics model with the features selected by least absolute shrinkage and selection operator (LASSO) method. The optimal radiomics model was then selected and combined with clinical features to formulate a nomogram model. The model performance was evaluated using receiver operating characteristic (ROC) curve analysis, and its clinical value was assessed via decision curve analysis (DCA). Our final selection comprised 10 optimal radiological features and the SVM model showcased superior predictive efficiency and robustness among the five classifiers. The area under the curve (AUC) values of the nomogram model were 0.945 (0.899, 0.991) and 0.846 (0.703, 0.99) for the training and validation sets, respectively. The nomogram model developed in this study exhibited excellent predictive performance in foretelling PNI of rectal cancer, thereby offering valuable guidance for clinical decision-making. The nomogram could predict the perineural invasion status of rectal cancer in early stage.
{"title":"Predictive Study of Machine Learning-Based Multiparametric MRI Radiomics Nomogram for Perineural Invasion in Rectal Cancer: A Pilot Study.","authors":"Yueyan Wang, Aiqi Chen, Kai Wang, Yihui Zhao, Xiaomeng Du, Yan Chen, Lei Lv, Yimin Huang, Yichuan Ma","doi":"10.1007/s10278-024-01231-6","DOIUrl":"https://doi.org/10.1007/s10278-024-01231-6","url":null,"abstract":"<p><p>This study aimed to establish and validate the efficacy of a nomogram model, synthesized through the integration of multi-parametric magnetic resonance radiomics and clinical risk factors, for forecasting perineural invasion in rectal cancer. We retrospectively collected data from 108 patients with pathologically confirmed rectal adenocarcinoma who underwent preoperative multiparametric MRI at the First Affiliated Hospital of Bengbu Medical College between April 2019 and August 2023. This dataset was subsequently divided into training and validation sets following a ratio of 7:3. Both univariate and multivariate logistic regression analyses were implemented to identify independent clinical risk factors associated with perineural invasion (PNI) in rectal cancer. We manually delineated the region of interest (ROI) layer-by-layer on T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) sequences and extracted the image features. Five machine learning algorithms were used to construct radiomics model with the features selected by least absolute shrinkage and selection operator (LASSO) method. The optimal radiomics model was then selected and combined with clinical features to formulate a nomogram model. The model performance was evaluated using receiver operating characteristic (ROC) curve analysis, and its clinical value was assessed via decision curve analysis (DCA). Our final selection comprised 10 optimal radiological features and the SVM model showcased superior predictive efficiency and robustness among the five classifiers. The area under the curve (AUC) values of the nomogram model were 0.945 (0.899, 0.991) and 0.846 (0.703, 0.99) for the training and validation sets, respectively. The nomogram model developed in this study exhibited excellent predictive performance in foretelling PNI of rectal cancer, thereby offering valuable guidance for clinical decision-making. The nomogram could predict the perineural invasion status of rectal cancer in early stage.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141989829","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-15DOI: 10.1007/s10278-024-01222-7
Vasanthi Durairaj, Palani Uthirapathy
Multi-modal medical image (MI) fusion assists in generating collaboration images collecting complement features through the distinct images of several conditions. The images help physicians to diagnose disease accurately. Hence, this research proposes a novel multi-modal MI fusion modal named guided filter-based interactive multi-scale and multi-modal transformer (Trans-IMSM) fusion approach to develop high-quality computed tomography-magnetic resonance imaging (CT-MRI) fused images for brain tumor detection. This research utilizes the CT and MRI brain scan dataset to gather the input CT and MRI images. At first, the data preprocessing is carried out to preprocess these input images to improve the image quality and generalization ability for further analysis. Then, these preprocessed CT and MRI are decomposed into detail and base components utilizing the guided filter-based MI decomposition approach. This approach involves two phases: such as acquiring the image guidance and decomposing the images utilizing the guided filter. A canny operator is employed to acquire the image guidance comprising robust edge for CT and MRI images, and the guided filter is applied to decompose the guidance and preprocessed images. Then, by applying the Trans-IMSM model, fuse the detail components, while a weighting approach is used for the base components. The fused detail and base components are subsequently processed through a gated fusion and reconstruction network, and the final fused images for brain tumor detection are generated. Extensive tests are carried out to compute the Trans-IMSM method's efficacy. The evaluation results demonstrated the robustness and effectiveness, achieving an accuracy of 98.64% and an SSIM of 0.94.
{"title":"Interactive Multi-scale Fusion: Advancing Brain Tumor Detection Through Trans-IMSM Model.","authors":"Vasanthi Durairaj, Palani Uthirapathy","doi":"10.1007/s10278-024-01222-7","DOIUrl":"https://doi.org/10.1007/s10278-024-01222-7","url":null,"abstract":"<p><p>Multi-modal medical image (MI) fusion assists in generating collaboration images collecting complement features through the distinct images of several conditions. The images help physicians to diagnose disease accurately. Hence, this research proposes a novel multi-modal MI fusion modal named guided filter-based interactive multi-scale and multi-modal transformer (Trans-IMSM) fusion approach to develop high-quality computed tomography-magnetic resonance imaging (CT-MRI) fused images for brain tumor detection. This research utilizes the CT and MRI brain scan dataset to gather the input CT and MRI images. At first, the data preprocessing is carried out to preprocess these input images to improve the image quality and generalization ability for further analysis. Then, these preprocessed CT and MRI are decomposed into detail and base components utilizing the guided filter-based MI decomposition approach. This approach involves two phases: such as acquiring the image guidance and decomposing the images utilizing the guided filter. A canny operator is employed to acquire the image guidance comprising robust edge for CT and MRI images, and the guided filter is applied to decompose the guidance and preprocessed images. Then, by applying the Trans-IMSM model, fuse the detail components, while a weighting approach is used for the base components. The fused detail and base components are subsequently processed through a gated fusion and reconstruction network, and the final fused images for brain tumor detection are generated. Extensive tests are carried out to compute the Trans-IMSM method's efficacy. The evaluation results demonstrated the robustness and effectiveness, achieving an accuracy of 98.64% and an SSIM of 0.94.</p>","PeriodicalId":516858,"journal":{"name":"Journal of imaging informatics in medicine","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141989828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}