Informatics in Medicine Unlocked最新文献_第3页

A systematic review on computer vision-based methods for cervical cancer detection 基于计算机视觉的宫颈癌检测方法综述

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-12-16 DOI: 10.1016/j.imu.2025.101726

Hope Mbelwa , Judith Leo , Crispin Kahesa , Elizabeth Mkoba

Cervical cancer is a leading cause of mortality among women globally, especially in regions where access to timely screening remains a challenge. With such concerns, accurate detection of cervical lesions is essential for effective diagnosis and treatment. This review aimed to explore the application of computer vision-based methods for detecting cervical cancer, identifying their potential, setbacks and areas for future development. A comprehensive literature search across Scopus, IEEE Xplore, PubMed, and Google Scholar identified 96 relevant studies published between 2014 and August 2025. These studies applied computer vision methods including CNNs, Vision Transformers, and multimodal models to cervical cancer detection using Pap smear, colposcopy, and histopathology images. They were analyzed based on the techniques employed, datasets used, evaluation metrics adopted, and reported results. This review highlights significant advancements in the field, particularly in lesion classification, precise segmentation of affected regions, and accurate detection of cancerous regions. However, some challenges were identified, including limited image datasets with insufficiently distributed normal and abnormal cases, aggravated by privacy issues and accurate labeling of medical images, which is critical and rigorous, often leading to annotation inconsistencies. Lastly, this study revealed that integrating Natural Language Processing and Computer Vision can enhance cervical cancer diagnosis through multi-modal models that combine both clinical text and imaging data. Additionally, this study proposes the use of techniques like annotation-efficient learning to manage limited labeled datasets using methods such as semi-supervised and transfer learning as well as the use of federated learning to ensure privacy in computer-aided diagnostic systems.

宫颈癌是全球妇女死亡的主要原因，特别是在获得及时筛查仍然是一项挑战的区域。考虑到这些问题，准确检测宫颈病变对于有效的诊断和治疗至关重要。本文旨在探讨基于计算机视觉的宫颈癌检测方法的应用，确定其潜力，挫折和未来发展的领域。在Scopus、IEEE explore、PubMed和b谷歌Scholar上进行了全面的文献检索，确定了2014年至2025年8月期间发表的96项相关研究。这些研究将计算机视觉方法包括cnn、视觉变形器和多模态模型应用于宫颈涂片、阴道镜检查和组织病理学图像的宫颈癌检测。根据采用的技术、使用的数据集、采用的评估指标和报告的结果对它们进行分析。这篇综述强调了该领域的重大进展，特别是在病变分类、受影响区域的精确分割和癌变区域的准确检测方面。然而，也发现了一些挑战，包括图像数据集有限，正常和异常病例分布不足，隐私问题和医学图像的准确标记（这是至关重要和严格的）加剧了这些问题，往往导致注释不一致。最后，本研究表明，将自然语言处理和计算机视觉相结合，可以通过结合临床文本和图像数据的多模态模型来提高宫颈癌的诊断。此外，本研究提出使用注释高效学习等技术来管理有限的标记数据集，使用半监督和迁移学习等方法，以及使用联邦学习来确保计算机辅助诊断系统中的隐私。

{"title":"A systematic review on computer vision-based methods for cervical cancer detection","authors":"Hope Mbelwa , Judith Leo , Crispin Kahesa , Elizabeth Mkoba","doi":"10.1016/j.imu.2025.101726","DOIUrl":"10.1016/j.imu.2025.101726","url":null,"abstract":"<div><div>Cervical cancer is a leading cause of mortality among women globally, especially in regions where access to timely screening remains a challenge. With such concerns, accurate detection of cervical lesions is essential for effective diagnosis and treatment. This review aimed to explore the application of computer vision-based methods for detecting cervical cancer, identifying their potential, setbacks and areas for future development. A comprehensive literature search across Scopus, IEEE Xplore, PubMed, and Google Scholar identified 96 relevant studies published between 2014 and August 2025. These studies applied computer vision methods including CNNs, Vision Transformers, and multimodal models to cervical cancer detection using Pap smear, colposcopy, and histopathology images. They were analyzed based on the techniques employed, datasets used, evaluation metrics adopted, and reported results. This review highlights significant advancements in the field, particularly in lesion classification, precise segmentation of affected regions, and accurate detection of cancerous regions. However, some challenges were identified, including limited image datasets with insufficiently distributed normal and abnormal cases, aggravated by privacy issues and accurate labeling of medical images, which is critical and rigorous, often leading to annotation inconsistencies. Lastly, this study revealed that integrating Natural Language Processing and Computer Vision can enhance cervical cancer diagnosis through multi-modal models that combine both clinical text and imaging data. Additionally, this study proposes the use of techniques like annotation-efficient learning to manage limited labeled datasets using methods such as semi-supervised and transfer learning as well as the use of federated learning to ensure privacy in computer-aided diagnostic systems.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"60 ","pages":"Article 101726"},"PeriodicalIF":0.0,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788409","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}

引用次数: 0

Epic overhaul at a Canadian hospital: Pre-Post evaluation insights from physicians and medical residents 加拿大一家医院史诗般的大修：来自医生和住院医生的前后评估见解

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-12-16 DOI: 10.1016/j.imu.2025.101725

Mirou Jaana , Edward Riachy , Erika MacPhee , Heather Sherrard

Background

The implementation of Electronic Medical Records (EMRs) in hospitals offers potential benefits but often disrupts clinician workflows, affecting care delivery and outcomes.

Objective

This study evaluates physicians' and medical residents’ perspectives on the impacts of introducing a new Epic system at a Canadian academic hospital.

Methods

A pre-post evaluation design was conducted using physician and resident surveys before (T0) and 4- and 9-months post-implementation (T1, T2) that assessed technology use, satisfaction with training and system use, and EMR's perceived impact on care delivery, work practices and quality.

Results

Satisfaction with training and system use declined for both groups in the first four months (more sharply for residents) but several measures improved at T2 as users readjusted to the system. There was a significant increase in physicians’ daily computer use (4 h at T0 to 6 h at T1; P < .001). Limited early benefits of the Epic system were observed and a decline in perceived improvement in clinical documentation (P = .006 and .0012), order entry (P = .018 and .002) and patient safety (P = .044 and .024) were reported at T1 for physicians and residents, respectively. Although some medical practice/work indicators improved by 9 months for physicians, the changes were not statistically significant; these benefits were not observed for residents at T2. Medical training was not significantly affected by the new Epic system either immediately or later post implementation. At T1, 83% of physicians reported that the new system sometimes or often improved the quality of care, as opposed to only 33% of residents; no significant improvements were noted at 9 months post implementation by both groups.

Conclusions

Physicians and residents adapt differently to Epic and full system assimilation does not happen in one year. Early perceptions of Epic do not reflect its long-term potential, and meaningful benefits require prolonged stabilization periods for user satisfaction and efficiency gains. We caution hospital leaders not to rely heavily on a vendor-driven implementation, and recommend tailored training, rapid-cycle improvements, transparent communication, and monitoring of agreed-upon performance indicators to strengthen clinician engagement and support long-term success.

在医院实施电子病历（emr）提供了潜在的好处，但往往会扰乱临床医生的工作流程，影响医疗服务的提供和结果。目的本研究评估医生和住院医师对加拿大一家学术医院引入新的Epic系统的影响的看法。方法采用实施前（T0）和实施后4个月、9个月（T1、T2）的医师和住院医师调查进行前后评估设计，评估技术使用情况、培训和系统使用满意度，以及EMR对护理交付、工作实践和质量的感知影响。结果在前四个月，两组患者对培训和系统使用的满意度都有所下降（住院患者的满意度下降得更明显），但随着用户重新适应系统，在T2阶段有几项指标有所改善。医生每天使用电脑的时间显著增加（T0时为4小时，T1时为6小时；P < 0.001）。观察到Epic系统的早期益处有限，临床文献中感知到的改善有所下降（P = 0.006和P = 0.006）。0012)，订单输入(P = .018和。002)和患者安全(P = 0.044；在T1时，医生和住院医生分别报告了024例。虽然医生的一些医疗实践/工作指标提高了9个月，但变化没有统计学意义；T2的居民没有观察到这些益处。新的Epic系统对医疗训练没有明显的影响，无论是立即还是之后。在T1, 83%的医生报告说，新系统有时或经常提高护理质量，而只有33%的住院医生这样认为；两组在实施后9个月均未见明显改善。结论医师和住院医师对Epic的适应程度存在差异，并不是在一年内完全同化。早期对Epic的看法并没有反映出它的长期潜力，真正有意义的收益需要长期的稳定期来满足用户满意度和提高效率。我们提醒医院领导不要严重依赖供应商驱动的实施，并建议定制培训、快速周期改进、透明沟通和监测商定的绩效指标，以加强临床医生的参与并支持长期成功。

{"title":"Epic overhaul at a Canadian hospital: Pre-Post evaluation insights from physicians and medical residents","authors":"Mirou Jaana , Edward Riachy , Erika MacPhee , Heather Sherrard","doi":"10.1016/j.imu.2025.101725","DOIUrl":"10.1016/j.imu.2025.101725","url":null,"abstract":"<div><h3>Background</h3><div>The implementation of Electronic Medical Records (EMRs) in hospitals offers potential benefits but often disrupts clinician workflows, affecting care delivery and outcomes.</div></div><div><h3>Objective</h3><div>This study evaluates physicians' and medical residents’ perspectives on the impacts of introducing a new Epic system at a Canadian academic hospital.</div></div><div><h3>Methods</h3><div>A pre-post evaluation design was conducted using physician and resident surveys before (T0) and 4- and 9-months post-implementation (T1, T2) that assessed technology use, satisfaction with training and system use, and EMR's perceived impact on care delivery, work practices and quality.</div></div><div><h3>Results</h3><div>Satisfaction with training and system use declined for both groups in the first four months (more sharply for residents) but several measures improved at T2 as users readjusted to the system. There was a significant increase in physicians’ daily computer use (4 h at T0 to 6 h at T1; P < .<em>001</em>). Limited early benefits of the Epic system were observed and a decline in perceived improvement in clinical documentation (P = .<em>006</em> and <em>.0012)</em>, order entry (P = .<em>018</em> and <em>.002)</em> and patient safety (P = .<em>044</em> and <em>.024)</em> were reported at T1 for physicians and residents, respectively. Although some medical practice/work indicators improved by 9 months for physicians, the changes were not statistically significant; these benefits were not observed for residents at T2. Medical training was not significantly affected by the new Epic system either immediately or later post implementation. At T1, 83% of physicians reported that the new system sometimes or often improved the quality of care, as opposed to only 33% of residents; no significant improvements were noted at 9 months post implementation by both groups.</div></div><div><h3>Conclusions</h3><div>Physicians and residents adapt differently to Epic and full system assimilation does not happen in one year. Early perceptions of Epic do not reflect its long-term potential, and meaningful benefits require prolonged stabilization periods for user satisfaction and efficiency gains. We caution hospital leaders not to rely heavily on a vendor-driven implementation, and recommend tailored training, rapid-cycle improvements, transparent communication, and monitoring of agreed-upon performance indicators to strengthen clinician engagement and support long-term success.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"61 ","pages":"Article 101725"},"PeriodicalIF":0.0,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075004","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}

引用次数: 0

Comparative development and evaluation of multilayer fuzzy and adaptive neuro-fuzzy inference systems for early diagnosis of chronic kidney disease 多层模糊与自适应神经模糊推理系统在慢性肾脏疾病早期诊断中的比较开发与评价

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-12-13 DOI: 10.1016/j.imu.2025.101724

Arvind Sharma , Dalwinder Singh , Arun Singh , MVV Prasad Kantipudi , Saiprasd Potharaju , B. Suresh Babu

Chronic Kidney Disease (CKD) is a gradual yet life-threatening condition that often remains undetected until significant renal impairment has occurred. Early recognition is essential to slow its progression and reduce the risk of associated complications. Conventional diagnostic procedures have limitations; they are subjective and may be ineffective when symptoms take longer to manifest. In this paper, we propose and compare two intelligent diagnostic models—a Multilayer Fuzzy Inference System (MLFIS) and an Adaptive Neuro-Fuzzy Inference System (ANFIS)—that aim to diagnose early-stage CKD using clinical symptoms and laboratory indicators.

The model of MLFIS is constructed in dual form i.e. the first layer determines the probable presence of CKD, which is based on symptoms like hypertension, diabetes, and family history, and the second layer specifies the disease stage, which is done based on biochemical parameters attached with or without or increasing glomerular filtration rate, blood urea nitrogen, serum creatinine, albumin, and urinary pus cells. The ANFIS model, however, combines the strengths of both fuzzy logic and neural networks, providing automatic learning of diagnosis rules based on data, thereby increasing accuracy and flexibility.

The models had been constructed and benchmarked on a retrospective dataset consisting of 500 anonymised patient records retrieved from a publicly available resource, and then tested using standard performance measures. Although the ANFIS had a classification accuracy of 99.5 % which is a higher percentage than the 99.33 per cent accuracy of the MLFIS, the difference is not that great. Findings demonstrate the benchmarking performance of fuzzy and neuro-fuzzy systems for CKD staging, providing a foundation for future validation on hospital datasets. However, it is important to note that the high accuracies reported are based on a relatively small UCI dataset and should be interpreted as internal validation only. Broader clinical validation is necessary before deployment in real-world settings. The paper also compares these two systems and determines their performance based on current machine learning models. This study seeks to fill the gap between the laboratory-based diagnostic and real clinical practice by presenting a scalable and understandable solution to screening of CKD.

慢性肾脏疾病（CKD）是一种渐进但危及生命的疾病，通常在发生重大肾脏损害之前未被发现。早期识别对于减缓其进展和减少相关并发症的风险至关重要。传统的诊断程序有局限性；它们是主观的，当症状需要较长时间才能显现时，它们可能无效。在本文中，我们提出并比较了两种智能诊断模型-多层模糊推理系统（MLFIS）和自适应神经模糊推理系统(ANFIS) -旨在通过临床症状和实验室指标诊断早期CKD。MLFIS模型采用双重形式构建，第一层根据高血压、糖尿病、家族史等症状判断是否存在CKD，第二层根据肾小球滤过率、血尿素氮、血清肌酐、白蛋白、尿脓细胞等生化指标有无升高来确定疾病分期。然而，ANFIS模型结合了模糊逻辑和神经网络的优势，提供了基于数据的诊断规则的自动学习，从而提高了准确性和灵活性。这些模型是在一个回顾性数据集上构建的，该数据集由500个从公共资源中检索的匿名患者记录组成，然后使用标准性能指标进行测试。虽然ANFIS的分类准确率为99.5%，高于MLFIS的99.33%，但差异并不大。研究结果证明了模糊和神经模糊系统对CKD分期的基准性能，为未来在医院数据集上的验证提供了基础。然而，值得注意的是，报告的高准确性是基于相对较小的UCI数据集，应该只解释为内部验证。在实际应用之前，需要进行更广泛的临床验证。本文还比较了这两种系统，并基于当前的机器学习模型确定了它们的性能。本研究旨在通过提出一种可扩展且可理解的CKD筛查解决方案来填补实验室诊断与实际临床实践之间的空白。

{"title":"Comparative development and evaluation of multilayer fuzzy and adaptive neuro-fuzzy inference systems for early diagnosis of chronic kidney disease","authors":"Arvind Sharma , Dalwinder Singh , Arun Singh , MVV Prasad Kantipudi , Saiprasd Potharaju , B. Suresh Babu","doi":"10.1016/j.imu.2025.101724","DOIUrl":"10.1016/j.imu.2025.101724","url":null,"abstract":"<div><div>Chronic Kidney Disease (CKD) is a gradual yet life-threatening condition that often remains undetected until significant renal impairment has occurred. Early recognition is essential to slow its progression and reduce the risk of associated complications. Conventional diagnostic procedures have limitations; they are subjective and may be ineffective when symptoms take longer to manifest. In this paper, we propose and compare two intelligent diagnostic models—a Multilayer Fuzzy Inference System (MLFIS) and an Adaptive Neuro-Fuzzy Inference System (ANFIS)—that aim to diagnose early-stage CKD using clinical symptoms and laboratory indicators.</div><div>The model of MLFIS is constructed in dual form i.e. the first layer determines the probable presence of CKD, which is based on symptoms like hypertension, diabetes, and family history, and the second layer specifies the disease stage, which is done based on biochemical parameters attached with or without or increasing glomerular filtration rate, blood urea nitrogen, serum creatinine, albumin, and urinary pus cells. The ANFIS model, however, combines the strengths of both fuzzy logic and neural networks, providing automatic learning of diagnosis rules based on data, thereby increasing accuracy and flexibility.</div><div>The models had been constructed and benchmarked on a retrospective dataset consisting of 500 anonymised patient records retrieved from a publicly available resource, and then tested using standard performance measures. Although the ANFIS had a classification accuracy of 99.5 % which is a higher percentage than the 99.33 per cent accuracy of the MLFIS, the difference is not that great. Findings demonstrate the benchmarking performance of fuzzy and neuro-fuzzy systems for CKD staging, providing a foundation for future validation on hospital datasets. However, it is important to note that the high accuracies reported are based on a relatively small UCI dataset and should be interpreted as internal validation only. Broader clinical validation is necessary before deployment in real-world settings. The paper also compares these two systems and determines their performance based on current machine learning models. This study seeks to fill the gap between the laboratory-based diagnostic and real clinical practice by presenting a scalable and understandable solution to screening of CKD.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"60 ","pages":"Article 101724"},"PeriodicalIF":0.0,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788410","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}

引用次数: 0

Review of machine learning models in short- and long-term glucose forecasting and hypoglycemia classification 短期和长期血糖预测和低血糖分类中的机器学习模型综述

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-12-08 DOI: 10.1016/j.imu.2025.101723

Beyza Cinar , Louisa van den Boom , Maria Maleshkova

Type 1 Diabetes (T1D) is a chronic autoimmune disorder that requires lifelong insulin therapy. A common side effect is hypoglycemia, characterized by decreased blood glucose levels (BGL) below 70 mg/dL. Diabetes care can be optimized using machine learning (ML) models that can predict and alert patients to potential glycemic abnormalities. The ML models can be classified into regression-based, in which glucose levels are forecasted, and classification-based, in which adverse events are classified. This review analyzes the performance of ML models applied to T1D and compares these in terms of short- and long-term prediction horizons (PHs), defined as 15–120 min and 3 to more than 24 h, respectively. This review investigates: 1) How much in advance can glucose values or a hypoglycemic event be accurately predicted? 2) Which ML methods have the best performance? 3) Which factors impact the performance? And 4) Does personalization increase performance? The results indicate that 1) a PH of up to 1 h provides the best results. 2) Conventional ML methods yield the best results for classification and deep learning (DL) for regression. A single model cannot adequately classify across multiple PHs. 3) The model performance is influenced by multivariate datasets and the input sequence length (ISL). 4) Finally, personal data enhances performance, but due to limited data quality, population-based models are preferred.

1型糖尿病（T1D）是一种需要终身胰岛素治疗的慢性自身免疫性疾病。常见的副作用是低血糖，其特征是血糖水平（BGL）降至70 mg/dL以下。糖尿病护理可以使用机器学习（ML）模型进行优化，该模型可以预测并提醒患者潜在的血糖异常。ML模型可以分为基于回归的模型（预测葡萄糖水平）和基于分类的模型（对不良事件进行分类）。本文分析了应用于T1D的ML模型的性能，并在短期和长期预测范围（ph）方面进行了比较，分别定义为15-120分钟和3至24小时以上。本文综述了以下问题：1)血糖值或低血糖事件可以提前多少准确预测？2)哪种ML方法性能最好？3)哪些因素影响绩效？4)个性化是否能提高性能？结果表明：1)PH为1 h时效果最好。2)传统的机器学习方法在分类和深度学习（DL）回归方面的效果最好。单个模型不能对多个ph值进行充分的分类。3)模型性能受多元数据集和输入序列长度（ISL）的影响。4)最后，个人数据增强了性能，但由于数据质量有限，基于人口的模型更受青睐。

{"title":"Review of machine learning models in short- and long-term glucose forecasting and hypoglycemia classification","authors":"Beyza Cinar , Louisa van den Boom , Maria Maleshkova","doi":"10.1016/j.imu.2025.101723","DOIUrl":"10.1016/j.imu.2025.101723","url":null,"abstract":"<div><div>Type 1 Diabetes (T1D) is a chronic autoimmune disorder that requires lifelong insulin therapy. A common side effect is hypoglycemia, characterized by decreased blood glucose levels (BGL) below 70 mg/dL. Diabetes care can be optimized using machine learning (ML) models that can predict and alert patients to potential glycemic abnormalities. The ML models can be classified into regression-based, in which glucose levels are forecasted, and classification-based, in which adverse events are classified. This review analyzes the performance of ML models applied to T1D and compares these in terms of short- and long-term prediction horizons (PHs), defined as 15–120 min and 3 to more than 24 h, respectively. This review investigates: 1) How much in advance can glucose values or a hypoglycemic event be accurately predicted? 2) Which ML methods have the best performance? 3) Which factors impact the performance? And 4) Does personalization increase performance? The results indicate that 1) a PH of up to 1 h provides the best results. 2) Conventional ML methods yield the best results for classification and deep learning (DL) for regression. A single model cannot adequately classify across multiple PHs. 3) The model performance is influenced by multivariate datasets and the input sequence length (ISL). 4) Finally, personal data enhances performance, but due to limited data quality, population-based models are preferred.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"60 ","pages":"Article 101723"},"PeriodicalIF":0.0,"publicationDate":"2025-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145735871","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}

引用次数: 0

3D SPECT-based machine learning approach to early Parkinson’s diagnosis 基于3D spect的机器学习方法在帕金森早期诊断中的应用

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-12-08 DOI: 10.1016/j.imu.2025.101722

Jihad Boucherouite, Abdelilah Jilbab, Atman Jbari

Purpose:

Dopamine transporter binding measurements in the striatum play a critical role in identifying dopaminergic deficits in Parkinson’s disease (PD), particularly during early stages. Using 3D SPECT imaging data from the PPMI database (928 subjects), we developed an automated feature extraction pipeline that captures subtle morphological changes characteristic of early neurodegeneration.

Approach:

Our methodology involved a sub-volume from each 3D SPECT scan and applied several image processing techniques to each slice. Unlike approaches relying on external datasets or derived features, we focused exclusively on direct geometric and shape features. We extracted 30 novel discriminative features, including volumetric measures, global thresholds, circularity, compactness, and curvature, with rigorous statistical validation using parametric tests (t-tests, ANOVA). The top 19 features, prioritized through MRMR to eliminate redundancy, were used to train and evaluate 34 machine learning classifiers.

Results:

For binary classification (Healthy Control (HC) vs. early PD), our Medium Gaussian SVM achieved excellent testing performance (98.56% accuracy, 98.81% precision), significantly higher than recent state-of-the-art approaches. Meanwhile, for three-class classification (HC, PD stage 1, and PD stage 2), the Subspace Discriminant Ensemble achieved the highest accuracy of 76.26%, solving a critical real-world scenario.

Conclusions:

This stage-aware approach advances the field by enabling automatic bilateral analysis of the striatum and accurate classification of early-stage PD, potentially supporting earlier clinical intervention. It also addresses a critical gap in the literature where most methods focus solely on binary classification (HC vs. PD) without capturing disease progression.

目的：纹状体多巴胺转运体结合测量在识别帕金森病（PD）的多巴胺能缺陷中起关键作用，特别是在早期阶段。利用来自PPMI数据库（928名受试者）的3D SPECT成像数据，我们开发了一种自动特征提取管道，可以捕获早期神经变性的细微形态学变化特征。方法：我们的方法涉及每个3D SPECT扫描的子体积，并对每个切片应用几种图像处理技术。与依赖外部数据集或衍生特征的方法不同，我们专注于直接的几何和形状特征。我们提取了30个新的判别特征，包括体积度量、全局阈值、圆度、紧度和曲率，并使用参数检验（t检验、方差分析）进行了严格的统计验证。通过MRMR对前19个特征进行优先排序以消除冗余，用于训练和评估34个机器学习分类器。结果：对于二分类（健康对照与早期PD），我们的中高斯支持向量机获得了出色的测试性能（98.56%的准确率，98.81%的精密度），显著高于目前最先进的方法。同时，对于三类分类（HC、PD阶段1和PD阶段2），子空间判别集成（Subspace Discriminant Ensemble）的准确率最高，达到76.26%，解决了一个关键的现实场景。结论：这种阶段感知方法通过实现纹状体的自动双侧分析和早期PD的准确分类，推动了该领域的发展，可能支持早期临床干预。它还解决了文献中大多数方法仅关注二元分类（HC与PD）而不捕获疾病进展的关键空白。

{"title":"3D SPECT-based machine learning approach to early Parkinson’s diagnosis","authors":"Jihad Boucherouite, Abdelilah Jilbab, Atman Jbari","doi":"10.1016/j.imu.2025.101722","DOIUrl":"10.1016/j.imu.2025.101722","url":null,"abstract":"<div><h3>Purpose:</h3><div>Dopamine transporter binding measurements in the striatum play a critical role in identifying dopaminergic deficits in Parkinson’s disease (PD), particularly during early stages. Using 3D SPECT imaging data from the PPMI database (928 subjects), we developed an automated feature extraction pipeline that captures subtle morphological changes characteristic of early neurodegeneration.</div></div><div><h3>Approach:</h3><div>Our methodology involved a sub-volume from each 3D SPECT scan and applied several image processing techniques to each slice. Unlike approaches relying on external datasets or derived features, we focused exclusively on direct geometric and shape features. We extracted 30 novel discriminative features, including volumetric measures, global thresholds, circularity, compactness, and curvature, with rigorous statistical validation using parametric tests (t-tests, ANOVA). The top 19 features, prioritized through MRMR to eliminate redundancy, were used to train and evaluate 34 machine learning classifiers.</div></div><div><h3>Results:</h3><div>For binary classification (Healthy Control (HC) vs. early PD), our Medium Gaussian SVM achieved excellent testing performance (98.56% accuracy, 98.81% precision), significantly higher than recent state-of-the-art approaches. Meanwhile, for three-class classification (HC, PD stage 1, and PD stage 2), the Subspace Discriminant Ensemble achieved the highest accuracy of 76.26%, solving a critical real-world scenario.</div></div><div><h3>Conclusions:</h3><div>This stage-aware approach advances the field by enabling automatic bilateral analysis of the striatum and accurate classification of early-stage PD, potentially supporting earlier clinical intervention. It also addresses a critical gap in the literature where most methods focus solely on binary classification (HC vs. PD) without capturing disease progression.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"60 ","pages":"Article 101722"},"PeriodicalIF":0.0,"publicationDate":"2025-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145697959","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}

引用次数: 0

SuSastho.AI: A multimodal medical copilot for adolescents using evidence-based medicine and large language models SuSastho。人工智能：使用循证医学和大型语言模型为青少年提供多模式医疗副驾驶

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-12-03 DOI: 10.1016/j.imu.2025.101720

Shamim Ahamed , Moinul H. Chowdhury , Marzia Zaman , Abhishek Agarwala , Md Abdullah Al Mashud , Tariqul , Tareq Al Muntasir , Rifat Rahman , Rifat Shahriyar , Farhana Sarker , Khondaker A. Mamun

Adolescents in Bangladesh face serious sexual, reproductive, and mental health challenges due to cultural stigma, poverty, and limited healthcare infrastructure. Within the country, 63% of adolescents are deprived of essential sexual and reproductive health services, and only 13% receive mental health support. Adolescents living in urban slums and with disabilities face additional challenges in receiving reliable health information. This limited access exposes them to a high risk of sexually transmitted infections (STIs), unintended pregnancies, and serious mental health issues. Addressing these challenges, our study introduces SuSastho.AI, a healthcare copilot providing access to reliable health information to adolescents. We utilized large language models, along with Evidence-Based Medicine, retrieval-augmented generation, and a clinically validated dataset to provide evidence-based responses, supporting both voice and text-based interactions. Clinical evaluation of a pilot study shows our method reduces incorrect responses by 26.9% and increases response correctness by 16.1% compared to other methods. It achieved an accuracy rate of 86.7% when specifically evaluated based on available knowledge. While the responses are mostly consistent with up-to-date medical practices, occasional, less precise responses highlight the need for further refinement. Participants reported overall positive feedback, where 87% found answers to their questions, and 90.7% found responses relevant. Our results show that SuSastho.AI can provide reliable and evidence-based information while being an affordable way to support traditional healthcare systems with a high potential to transform digital health. The study sets an example as an evidence-based healthcare copilot to support adolescents and lays the foundation for future research where evidence-based tools overcome social barriers.

由于文化耻辱、贫困和有限的医疗基础设施，孟加拉国的青少年面临着严重的性、生殖和精神健康挑战。在国内，63%的青少年被剥夺了基本的性健康和生殖健康服务，只有13%的青少年获得心理健康支持。生活在城市贫民窟的残疾青少年在获得可靠的保健信息方面面临更多挑战。这种有限的获取途径使她们面临性传播感染、意外怀孕和严重精神健康问题的高风险。为了解决这些挑战，我们的研究引入了susasto。人工智能，一个向青少年提供可靠健康信息的医疗保健副驾驶员。我们利用大型语言模型、循证医学、检索增强生成和临床验证数据集来提供基于证据的响应，支持基于语音和基于文本的交互。一项初步研究的临床评估表明，与其他方法相比，我们的方法减少了26.9%的错误反应，提高了16.1%的反应正确性。当基于现有知识进行具体评估时，准确率达到86.7%。虽然这些答复大多符合最新的医疗实践，但偶尔出现的不太精确的答复突出表明需要进一步改进。参与者报告了总体的积极反馈，其中87%的人找到了问题的答案，90.7%的人认为答案相关。我们的结果表明，susasto。人工智能可以提供可靠和基于证据的信息，同时也是支持传统医疗保健系统的一种经济实惠的方式，具有很大的潜力来转变数字卫生。该研究为支持青少年的循证医疗辅助试点树立了榜样，并为未来的循证工具克服社会障碍的研究奠定了基础。

{"title":"SuSastho.AI: A multimodal medical copilot for adolescents using evidence-based medicine and large language models","authors":"Shamim Ahamed , Moinul H. Chowdhury , Marzia Zaman , Abhishek Agarwala , Md Abdullah Al Mashud , Tariqul , Tareq Al Muntasir , Rifat Rahman , Rifat Shahriyar , Farhana Sarker , Khondaker A. Mamun","doi":"10.1016/j.imu.2025.101720","DOIUrl":"10.1016/j.imu.2025.101720","url":null,"abstract":"<div><div>Adolescents in Bangladesh face serious sexual, reproductive, and mental health challenges due to cultural stigma, poverty, and limited healthcare infrastructure. Within the country, 63% of adolescents are deprived of essential sexual and reproductive health services, and only 13% receive mental health support. Adolescents living in urban slums and with disabilities face additional challenges in receiving reliable health information. This limited access exposes them to a high risk of sexually transmitted infections (STIs), unintended pregnancies, and serious mental health issues. Addressing these challenges, our study introduces SuSastho.AI, a healthcare copilot providing access to reliable health information to adolescents. We utilized large language models, along with Evidence-Based Medicine, retrieval-augmented generation, and a clinically validated dataset to provide evidence-based responses, supporting both voice and text-based interactions. Clinical evaluation of a pilot study shows our method reduces incorrect responses by 26.9% and increases response correctness by 16.1% compared to other methods. It achieved an accuracy rate of 86.7% when specifically evaluated based on available knowledge. While the responses are mostly consistent with up-to-date medical practices, occasional, less precise responses highlight the need for further refinement. Participants reported overall positive feedback, where 87% found answers to their questions, and 90.7% found responses relevant. Our results show that SuSastho.AI can provide reliable and evidence-based information while being an affordable way to support traditional healthcare systems with a high potential to transform digital health. The study sets an example as an evidence-based healthcare copilot to support adolescents and lays the foundation for future research where evidence-based tools overcome social barriers.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"60 ","pages":"Article 101720"},"PeriodicalIF":0.0,"publicationDate":"2025-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788038","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}

引用次数: 0

A stochastic model for evaluating the progression of ductal carcinoma in situ breast cancer using Norwegian breast cancer screening program data 使用挪威乳腺癌筛查项目数据评估导管原位癌进展的随机模型

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-01-01 DOI: 10.1016/j.imu.2025.101647

Xiaoxue Li , Harald Weedon-Fekjær , Bo Zhang , Sandra J. Lee

<div><h3>Background</h3><div>Following widespread mammography screening for breast cancer, the incidence of ductal carcinoma in situ (DCIS) has increased sharply. However, the value of detecting DCIS by screening is uncertain as not all DCIS progresses to invasive breast cancer. Knowledge about the sojourn time in the screen-detectable DCIS state and the progression or regression of DCIS to other stages (i.e., the natural history of DCIS) is essential to treat screen-detected DCIS lesions.</div></div><div><h3>Methods</h3><div>We developed a stochastic model for DCIS natural history, characterized by DCIS states, invasive breast cancer states, and transition probabilities between the states. The model included DCIS lesions in the screen-detectable preclinical state and their progression to clinical DCIS, invasive breast cancer, or regression to a state undetectable by screening. Unlike currently available DCIS Markov models, the proposed model assumed no relationship between the sojourn time and transition probabilities in DCIS states and used age-specific transition probabilities. In the absence of ideal data for DCIS modeling, the Norwegian Breast Cancer Screening Program data, specifically arranged by screening round and mode of detection, was applied to obtain maximum likelihood estimates of DCIS natural history parameters, including transition probabilities and the mean sojourn time in the preclinical screen-detectable DCIS state.</div></div><div><h3>Results</h3><div>By indirectly specifying a range of the proportion of breast lesions in the preclinical undetectable DCIS state (S<sub>du</sub>) that progress through the preclinical screen-detectable DCIS state (S<sub>dp</sub>), <em>P</em><sub><em>d</em></sub><em>(t)</em>, not going directly to preclinical invasive breast cancer (S<sub>p</sub>), plausible sets of DCIS natural history parameters were systematically evaluated. All estimates indicated that the mean sojourn time in S<sub>dp</sub> was relatively short (≤3.5 years). For the age group 50–54 years, the best fitting mean sojourn time in S<sub>dp</sub> was 3.4–3.5 years, with mammography sensitivity 0.60–0.61 when <em>P</em><sub><em>d</em></sub><em>(t)</em> was 0.31–0.34. When <em>P</em><sub><em>d</em></sub><em>(t)</em> was larger, mean sojourn times in S<sub>dp</sub> likely varied by the pathway. In general, assuming higher <em>P</em><sub><em>d</em></sub><em>(t)</em>—that is, a higher proportion of DCIS lesions that progress to from S<sub>dp</sub> to S<sub>p</sub>—the mean sojourn time became shorter. Regression to no cancer or undetectable state might be possible, but the quantified level of regression was associated with great uncertainties.</div></div><div><h3>Conclusion</h3><div>While difficult to point to a unique set of DCIS natural history estimates, identifying broader sets of plausible estimates is possible. Estimates reported here provide a comprehensive view of potential progression paths of DCIS while acknowledging the limit

背景：随着乳房x线摄影对乳腺癌的广泛筛查，导管原位癌（DCIS）的发病率急剧上升。然而，通过筛查发现DCIS的价值尚不确定，因为并非所有DCIS都进展为浸润性乳腺癌。了解筛查到的DCIS状态的停留时间以及DCIS向其他阶段的进展或倒退（即DCIS的自然史）对于治疗筛查到的DCIS病变至关重要。方法我们建立了一个DCIS自然史的随机模型，以DCIS状态、浸润性乳腺癌状态和状态之间的转移概率为特征。该模型包括筛查可检测到的临床前DCIS病变及其进展为临床DCIS、浸润性乳腺癌或退至筛查无法检测到的状态。与现有的DCIS马尔可夫模型不同，该模型没有假设DCIS状态的停留时间和转移概率之间的关系，而是使用特定年龄的转移概率。在缺乏理想的DCIS建模数据的情况下，我们应用挪威乳腺癌筛查项目的数据，根据筛查轮和检测方式进行了特别安排，以获得DCIS自然史参数的最大似然估计，包括转移概率和临床前筛查可检测到的DCIS状态的平均停留时间。结果通过间接指定临床前无法检测到的DCIS状态（Sdu）的乳腺病变的比例范围，通过临床前可筛查的DCIS状态（Sdp）， Pd(t)，而不是直接进入临床前浸润性乳腺癌（Sp），系统地评估了DCIS自然史参数的合理集合。所有的估计表明，在Sdp的平均逗留时间相对较短（≤3.5年）。对于50-54岁年龄组，最佳拟合的平均Sdp停留时间为3.4-3.5年，当Pd(t)为0.31-0.34时，乳房x线摄影灵敏度为0.60-0.61。当Pd(t)较大时，Sdp的平均停留时间可能因途径而异。一般来说，假设Pd(t)较高，即从Sdp发展到sp的DCIS病变比例较高，则平均停留时间变短。回归到无癌或检测不到的状态是可能的，但回归的量化水平与很大的不确定性有关。结论：虽然很难得出一组独特的DCIS自然历史估计，但确定更广泛的合理估计是可能的。在承认现有数据的局限性的同时，本文报告的估计提供了DCIS潜在进展路径的全面视图。

{"title":"A stochastic model for evaluating the progression of ductal carcinoma in situ breast cancer using Norwegian breast cancer screening program data","authors":"Xiaoxue Li , Harald Weedon-Fekjær , Bo Zhang , Sandra J. Lee","doi":"10.1016/j.imu.2025.101647","DOIUrl":"10.1016/j.imu.2025.101647","url":null,"abstract":"<div><h3>Background</h3><div>Following widespread mammography screening for breast cancer, the incidence of ductal carcinoma in situ (DCIS) has increased sharply. However, the value of detecting DCIS by screening is uncertain as not all DCIS progresses to invasive breast cancer. Knowledge about the sojourn time in the screen-detectable DCIS state and the progression or regression of DCIS to other stages (i.e., the natural history of DCIS) is essential to treat screen-detected DCIS lesions.</div></div><div><h3>Methods</h3><div>We developed a stochastic model for DCIS natural history, characterized by DCIS states, invasive breast cancer states, and transition probabilities between the states. The model included DCIS lesions in the screen-detectable preclinical state and their progression to clinical DCIS, invasive breast cancer, or regression to a state undetectable by screening. Unlike currently available DCIS Markov models, the proposed model assumed no relationship between the sojourn time and transition probabilities in DCIS states and used age-specific transition probabilities. In the absence of ideal data for DCIS modeling, the Norwegian Breast Cancer Screening Program data, specifically arranged by screening round and mode of detection, was applied to obtain maximum likelihood estimates of DCIS natural history parameters, including transition probabilities and the mean sojourn time in the preclinical screen-detectable DCIS state.</div></div><div><h3>Results</h3><div>By indirectly specifying a range of the proportion of breast lesions in the preclinical undetectable DCIS state (S<sub>du</sub>) that progress through the preclinical screen-detectable DCIS state (S<sub>dp</sub>), <em>P</em><sub><em>d</em></sub><em>(t)</em>, not going directly to preclinical invasive breast cancer (S<sub>p</sub>), plausible sets of DCIS natural history parameters were systematically evaluated. All estimates indicated that the mean sojourn time in S<sub>dp</sub> was relatively short (≤3.5 years). For the age group 50–54 years, the best fitting mean sojourn time in S<sub>dp</sub> was 3.4–3.5 years, with mammography sensitivity 0.60–0.61 when <em>P</em><sub><em>d</em></sub><em>(t)</em> was 0.31–0.34. When <em>P</em><sub><em>d</em></sub><em>(t)</em> was larger, mean sojourn times in S<sub>dp</sub> likely varied by the pathway. In general, assuming higher <em>P</em><sub><em>d</em></sub><em>(t)</em>—that is, a higher proportion of DCIS lesions that progress to from S<sub>dp</sub> to S<sub>p</sub>—the mean sojourn time became shorter. Regression to no cancer or undetectable state might be possible, but the quantified level of regression was associated with great uncertainties.</div></div><div><h3>Conclusion</h3><div>While difficult to point to a unique set of DCIS natural history estimates, identifying broader sets of plausible estimates is possible. Estimates reported here provide a comprehensive view of potential progression paths of DCIS while acknowledging the limit","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"55 ","pages":"Article 101647"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143899764","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}

引用次数: 0

Regression and classification of Windkessel parameters from non-invasive cardiovascular quantities using a fully connected neural network 使用全连接神经网络从非侵入性心血管数量中回归和分类Windkessel参数

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-01-01 DOI: 10.1016/j.imu.2025.101614

Ahmed Gdoura , Stefan Bernhard

Despite their simplicity, three-element Windkessel models (WK-3) provide an effective and straightforward representation of the aortic input impedance. The WK-3 model not only captures valuable information about the mechanical and structural characteristics of the aortic arch but also generates reliable estimations of the central blood pressure (cBP) wave, a significant cardiovascular risk indicator. However, fitting the parameters of the WK-3 model typically requires invasively collected data, which carries substantial risk and high cost for patients.

This study aims to enable non-invasive impedance estimation of the WK-3 model using cardiovascular signals. As a proof of concept, we developed and trained a fully connected neural network (FCNN) on an in-silico dataset to predict the WK-3 parameters: characteristic impedance, peripheral arterial resistance, and arterial compliance. These predictions are based on non-invasive parameters, including zero-flow pressure intercept, heart rate, stroke volume, and left ventricular ejection time.

The proposed model achieved an overall accuracy of 80% with an average area under the curve (AUC) of

0.91 \pm 0.11

. The implementation and best-fitting model are available for download from this link.

尽管简单，三元素Windkessel模型（WK-3）提供了主动脉输入阻抗的有效和直接的表示。WK-3模型不仅捕获了有关主动脉弓的力学和结构特征的有价值的信息，而且还生成了对中心血压（cBP）波的可靠估计，这是一个重要的心血管风险指标。然而，拟合WK-3模型的参数通常需要侵入性地收集数据，这对患者来说风险很大，成本也很高。本研究旨在利用心血管信号实现WK-3模型的无创阻抗估计。作为概念验证，我们在一个计算机数据集上开发并训练了一个全连接神经网络（FCNN），以预测WK-3参数：特征阻抗、外周动脉阻力和动脉顺应性。这些预测是基于无创参数，包括零流压力截距、心率、卒中量和左心室射血时间。该模型总体精度为80%，平均曲线下面积（AUC）为0.91±0.11。实现和最佳拟合模型可从此链接下载。

{"title":"Regression and classification of Windkessel parameters from non-invasive cardiovascular quantities using a fully connected neural network","authors":"Ahmed Gdoura , Stefan Bernhard","doi":"10.1016/j.imu.2025.101614","DOIUrl":"10.1016/j.imu.2025.101614","url":null,"abstract":"<div><div>Despite their simplicity, three-element Windkessel models (WK-3) provide an effective and straightforward representation of the aortic input impedance. The WK-3 model not only captures valuable information about the mechanical and structural characteristics of the aortic arch but also generates reliable estimations of the central blood pressure (cBP) wave, a significant cardiovascular risk indicator. However, fitting the parameters of the WK-3 model typically requires invasively collected data, which carries substantial risk and high cost for patients.</div><div>This study aims to enable non-invasive impedance estimation of the WK-3 model using cardiovascular signals. As a proof of concept, we developed and trained a fully connected neural network (FCNN) on an in-silico dataset to predict the WK-3 parameters: characteristic impedance, peripheral arterial resistance, and arterial compliance. These predictions are based on non-invasive parameters, including zero-flow pressure intercept, heart rate, stroke volume, and left ventricular ejection time.</div><div>The proposed model achieved an overall accuracy of 80% with an average area under the curve (AUC) of <span><math><mrow><mn>0</mn><mo>.</mo><mn>91</mn><mo>±</mo><mn>0</mn><mo>.</mo><mn>11</mn></mrow></math></span>. The implementation and best-fitting model are available for download from <span><span>this link</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"53 ","pages":"Article 101614"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143103204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Development of deep learning-based classification models for opacity differentiation in pediatric chest radiography 基于深度学习的儿童胸片不透明分类模型的开发

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-01-01 DOI: 10.1016/j.imu.2024.101605

Germán Enrique Galvis Ruiz , Johana Benavides-Cruz , Daniela Muñoz Corredor , Esteban Morales-Mendoza , Héctor Daniel Alejandro Cotrino Palma , Andrés Cely-Jiménez

Opacities of non-interstitial origin in a pediatric patient's chest radiograph may indicate either consolidations and/or atelectasis, based on the appropriate clinical context. However, the overlapping and complex symptomatology of respiratory tract diseases in pediatric patients can make it difficult for physicians to interpret opacities. Artificial intelligence models are frequently employed by physicians for diagnostic support in healthcare, especially to evaluate aspects of radiographs that are not visible with the naked eye. In this study, a prediction model based on deep learning was used to differentiate between atelectasis and consolidations in pediatric chest radiographs from a clinical perspective. The radiologist can assist pediatricians in diagnosing respiratory pathologies based on the type of opacities using the machine learning model. We used 1297 chest X-ray images of pediatric patients with opacities including consolidations (

n = 500

), atelectasis (

n = 499

); and images without opacities (

n = 298

). The images were preprocessed, and various deep learning models were applied to determine the model with the best metrics. The InceptionV3 model demonstrated a significant improvement over its initial results.

儿科患者胸片上的非间质性混浊可能表明实变和/或肺不张，这取决于相应的临床情况。然而，儿科患者呼吸道疾病的重叠和复杂的症状使医生难以解释混浊。人工智能模型经常被医生用于医疗保健领域的诊断支持，特别是在评估裸眼无法看到的x光片方面。在本研究中，从临床角度使用基于深度学习的预测模型来区分小儿胸片中的肺不张和实变。放射科医生可以使用机器学习模型帮助儿科医生根据混浊的类型诊断呼吸系统疾病。我们使用1297张儿童患者的胸部x线图像，包括实变（n=500）、肺不张（n=499）；以及没有不透明的图像（n=298）。对图像进行预处理，并应用各种深度学习模型来确定具有最佳度量的模型。InceptionV3模型比它最初的结果有了显著的改进。

{"title":"Development of deep learning-based classification models for opacity differentiation in pediatric chest radiography","authors":"Germán Enrique Galvis Ruiz , Johana Benavides-Cruz , Daniela Muñoz Corredor , Esteban Morales-Mendoza , Héctor Daniel Alejandro Cotrino Palma , Andrés Cely-Jiménez","doi":"10.1016/j.imu.2024.101605","DOIUrl":"10.1016/j.imu.2024.101605","url":null,"abstract":"<div><div>Opacities of non-interstitial origin in a pediatric patient's chest radiograph may indicate either consolidations and/or atelectasis, based on the appropriate clinical context. However, the overlapping and complex symptomatology of respiratory tract diseases in pediatric patients can make it difficult for physicians to interpret opacities. Artificial intelligence models are frequently employed by physicians for diagnostic support in healthcare, especially to evaluate aspects of radiographs that are not visible with the naked eye. In this study, a prediction model based on deep learning was used to differentiate between atelectasis and consolidations in pediatric chest radiographs from a clinical perspective. The radiologist can assist pediatricians in diagnosing respiratory pathologies based on the type of opacities using the machine learning model. We used 1297 chest X-ray images of pediatric patients with opacities including consolidations (<span><math><mrow><mi>n</mi><mo>=</mo><mn>500</mn></mrow></math></span>), atelectasis (<span><math><mrow><mi>n</mi><mo>=</mo><mn>499</mn></mrow></math></span>); and images without opacities (<span><math><mrow><mi>n</mi><mo>=</mo><mn>298</mn></mrow></math></span>). The images were preprocessed, and various deep learning models were applied to determine the model with the best metrics. The InceptionV3 model demonstrated a significant improvement over its initial results.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"52 ","pages":"Article 101605"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143178370","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust assessment of cervical precancerous lesions from pre- and post-acetic acid cervicography by combining deep learning and medical guidelines 结合深度学习和医学指南，对醋酸前后宫颈造影的宫颈癌前病变进行强有力的评估

Q1 Medicine

Informatics in Medicine Unlocked

Pub Date : 2025-01-01 DOI: 10.1016/j.imu.2024.101609

Siti Nurmaini , Patiyus Agustiyansyah , Muhammad Naufal Rachmatullah , Firdaus Firdaus , Annisa Darmawahyuni , Bambang Tutuko , Ade Iriani Sapitri , Anggun Islami , Akhiar Wista Arum , Rizal Sanif , Irawan Sastradinata , Legiran Legiran , Radiyati Umi Partan

Cervical cancer remains a major public health challenge, particularly in low-resource settings where access to regular screening and expert medical evaluation is limited. Traditional visual inspection with acetic acid (VIA) has been widely used for cervical cancer screening but is subjective and highly dependent on the expertise of the healthcare provider. This study presents a comprehensive methodology for decision-making regarding cervical precancerous lesions using cervicograms taken before and after the application of acetic acid. By leveraging the power of the deep learning (DL) model with You Only Look Once (Yolo) version 8, Slicing Aided Hyper Inference (SAHI), and oncology medical guidelines, the system aims to improve the accuracy and consistency of VIA assessments. The method involves training a Yolov8xl model on our cervicogram dataset, annotated by two oncologists using VIA screening results, to distinguish between the cervical area, columnar area, and lesions. The model is designed to process cervicography images taken both before and after the application of acetic acid, capturing the dynamic changes in tissue appearance indicative of precancerous conditions. The automated evaluation system demonstrated high sensitivity and specificity in detecting cervical lesions with 90.78 % accuracy, 91.67 % sensitivity, and 90.96 % specificity, outperforming other existing methods. This work represents a significant step towards deploying AI-driven solutions in cervical cancer screening, potentially reducing the global burden of the disease. It can be integrated into existing screening programs, providing a valuable tool for early detection and intervention, especially in regions with limited access to trained medical personnel.

子宫颈癌仍然是一项重大的公共卫生挑战，特别是在资源匮乏、定期筛查和专家医疗评估机会有限的地区。传统的醋酸目视检查（VIA）已广泛用于宫颈癌筛查，但它是主观的，高度依赖于医疗保健提供者的专业知识。本研究提出了一种综合的方法，用于决策关于宫颈癌前病变使用前后采取的宫颈造影醋酸。通过利用深度学习（DL）模型（You Only Look Once (Yolo) version 8）、切片辅助超推断（SAHI）和肿瘤医学指南的强大功能，该系统旨在提高VIA评估的准确性和一致性。该方法包括在我们的宫颈图数据集上训练一个Yolov8xl模型，由两名肿瘤学家使用VIA筛查结果进行注释，以区分宫颈区域、柱状区域和病变。该模型旨在处理应用醋酸前后拍摄的宫颈造影图像，捕捉指示癌前病变的组织外观的动态变化。该系统检测宫颈病变的准确率为90.78%，灵敏度为91.67%，特异性为90.96%，优于现有的其他方法。这项工作是朝着在宫颈癌筛查中部署人工智能驱动解决方案迈出的重要一步，有可能减轻该疾病的全球负担。它可以整合到现有的筛查方案中，为早期发现和干预提供宝贵的工具，特别是在缺乏训练有素的医务人员的地区。

{"title":"Robust assessment of cervical precancerous lesions from pre- and post-acetic acid cervicography by combining deep learning and medical guidelines","authors":"Siti Nurmaini , Patiyus Agustiyansyah , Muhammad Naufal Rachmatullah , Firdaus Firdaus , Annisa Darmawahyuni , Bambang Tutuko , Ade Iriani Sapitri , Anggun Islami , Akhiar Wista Arum , Rizal Sanif , Irawan Sastradinata , Legiran Legiran , Radiyati Umi Partan","doi":"10.1016/j.imu.2024.101609","DOIUrl":"10.1016/j.imu.2024.101609","url":null,"abstract":"<div><div>Cervical cancer remains a major public health challenge, particularly in low-resource settings where access to regular screening and expert medical evaluation is limited. Traditional visual inspection with acetic acid (VIA) has been widely used for cervical cancer screening but is subjective and highly dependent on the expertise of the healthcare provider. This study presents a comprehensive methodology for decision-making regarding cervical precancerous lesions using cervicograms taken before and after the application of acetic acid. By leveraging the power of the deep learning (DL) model with You Only Look Once (Yolo) version 8, Slicing Aided Hyper Inference (SAHI), and oncology medical guidelines, the system aims to improve the accuracy and consistency of VIA assessments. The method involves training a Yolov8xl model on our cervicogram dataset, annotated by two oncologists using VIA screening results, to distinguish between the cervical area, columnar area, and lesions. The model is designed to process cervicography images taken both before and after the application of acetic acid, capturing the dynamic changes in tissue appearance indicative of precancerous conditions. The automated evaluation system demonstrated high sensitivity and specificity in detecting cervical lesions with 90.78 % accuracy, 91.67 % sensitivity, and 90.96 % specificity, outperforming other existing methods. This work represents a significant step towards deploying AI-driven solutions in cervical cancer screening, potentially reducing the global burden of the disease. It can be integrated into existing screening programs, providing a valuable tool for early detection and intervention, especially in regions with limited access to trained medical personnel.</div></div>","PeriodicalId":13953,"journal":{"name":"Informatics in Medicine Unlocked","volume":"52 ","pages":"Article 101609"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143178786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0