Computer methods and programs in biomedicine最新文献_第2页

Artificial intelligence approaches for non-invasive diabetes prediction using ECG signals: A systematic review 利用心电信号进行无创糖尿病预测的人工智能方法：系统综述。

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-05-01 Epub Date: 2026-02-03 DOI: 10.1016/j.cmpb.2026.109264

Kiruthika Balakrishnan , Durgadevi Velusamy , Karthikeyan Ramasamy , Hana E. Hinkle , Holly J. Hudson , Ram Bilas Pachori , Hikmat Khan

Diabetes is a major global health challenge, with many individuals remaining undiagnosed due to the limitations of traditional screening methods. Artificial intelligence (AI)-based electrocardiogram (ECG) analysis offers a promising, non-invasive approach for the early detection of diabetes. This systematic review aims to critically evaluate machine learning (ML) and deep learning (DL) models developed for non-invasive prediction of diabetes and prediabetes using ECG signals. A comprehensive literature search was conducted across PubMed, Embase, Web of Science, IEEE Xplore, and ACM Digital Library in accordance with PRISMA 2020 guidelines. Twenty-five studies met the inclusion criteria. Extracted data included ECG input types, model architectures, preprocessing methods, feature sets, validation strategies, and performance metrics. Most studies used small, single-site, cross-sectional datasets, with sample sizes ranging from 24 to over 190,000 individuals. ECG preprocessing methods varied widely, including filtering, normalization, and decomposition. Features were extracted from time, frequency, morphological, and non-linear domains, though formal feature selection was applied inconsistently. ML and DL models reported high internal accuracy (

>

90%) but most lacked external validation and subgroup performance assessments. Notably, no study specifically focused on rural or underserved populations, and only one provided open-source code. AI-based ECG analysis demonstrates strong potential for detecting diabetes; however, current research is limited by generalizability issues, lack of standardized methods, poor external validation, and insufficient transparency. Future studies should prioritize rigorous validation, reproducibility, fairness audits, and applications in rural and underserved settings to ensure equitable and clinically viable deployment of these models.

糖尿病是一项重大的全球健康挑战，由于传统筛查方法的局限性，许多人仍未被诊断出来。基于人工智能（AI）的心电图（ECG）分析为糖尿病的早期检测提供了一种有前途的、无创的方法。本系统综述旨在批判性地评估机器学习（ML）和深度学习（DL）模型，这些模型用于使用ECG信号进行糖尿病和前驱糖尿病的无创预测。根据PRISMA 2020指南，在PubMed、Embase、Web of Science、IEEE explore和ACM数字图书馆进行了全面的文献检索。25项研究符合纳入标准。提取的数据包括心电输入类型、模型架构、预处理方法、特征集、验证策略和性能指标。大多数研究使用小型、单站点、横断面数据集，样本量从24人到19万人不等。心电预处理方法多种多样，包括滤波、归一化和分解。从时间、频率、形态和非线性领域提取特征，尽管形式特征选择的应用不一致。ML和DL模型报告了较高的内部准确度（约90%），但大多数模型缺乏外部验证和亚组性能评估。值得注意的是，没有一项研究专门针对农村或服务不足的人群，只有一项研究提供了开源代码。基于人工智能的心电图分析显示出检测糖尿病的强大潜力；然而，目前的研究受到可泛性问题、缺乏标准化方法、外部验证差和透明度不足的限制。未来的研究应优先考虑严格的验证、可重复性、公平性审计以及在农村和服务不足地区的应用，以确保这些模型的公平和临床可行的部署。

{"title":"Artificial intelligence approaches for non-invasive diabetes prediction using ECG signals: A systematic review","authors":"Kiruthika Balakrishnan , Durgadevi Velusamy , Karthikeyan Ramasamy , Hana E. Hinkle , Holly J. Hudson , Ram Bilas Pachori , Hikmat Khan","doi":"10.1016/j.cmpb.2026.109264","DOIUrl":"10.1016/j.cmpb.2026.109264","url":null,"abstract":"<div><div>Diabetes is a major global health challenge, with many individuals remaining undiagnosed due to the limitations of traditional screening methods. Artificial intelligence (AI)-based electrocardiogram (ECG) analysis offers a promising, non-invasive approach for the early detection of diabetes. This systematic review aims to critically evaluate machine learning (ML) and deep learning (DL) models developed for non-invasive prediction of diabetes and prediabetes using ECG signals. A comprehensive literature search was conducted across PubMed, Embase, Web of Science, IEEE Xplore, and ACM Digital Library in accordance with PRISMA 2020 guidelines. Twenty-five studies met the inclusion criteria. Extracted data included ECG input types, model architectures, preprocessing methods, feature sets, validation strategies, and performance metrics. Most studies used small, single-site, cross-sectional datasets, with sample sizes ranging from 24 to over 190,000 individuals. ECG preprocessing methods varied widely, including filtering, normalization, and decomposition. Features were extracted from time, frequency, morphological, and non-linear domains, though formal feature selection was applied inconsistently. ML and DL models reported high internal accuracy (<span><math><mo>></mo></math></span>90%) but most lacked external validation and subgroup performance assessments. Notably, no study specifically focused on rural or underserved populations, and only one provided open-source code. AI-based ECG analysis demonstrates strong potential for detecting diabetes; however, current research is limited by generalizability issues, lack of standardized methods, poor external validation, and insufficient transparency. Future studies should prioritize rigorous validation, reproducibility, fairness audits, and applications in rural and underserved settings to ensure equitable and clinically viable deployment of these models.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"278 ","pages":"Article 109264"},"PeriodicalIF":4.8,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146123997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Corrigendum to “Energy loss minimization-based side branch flow model for FFR calculation based on intracoronary images” [Computer Methods and Programs in Biomedicine 269 (2025) 108872] “基于冠状动脉内图像的FFR计算的基于能量损失最小化的侧分支血流模型”的勘误表[生物医学计算机方法和程序269(2025)108872]。

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-05-01 Epub Date: 2026-02-11 DOI: 10.1016/j.cmpb.2026.109278

Xiangling Lai , Xiaofei Xue , Zhifan Gao , Baihong Xie , Heye Zhang , Dong Yong , Haibo Jia , Xiujian Liu

引用次数: 0

Effects of tibial fracture-induced gait alterations on healing outcomes: Implications for patient-specific rehabilitation strategies 胫骨骨折引起的步态改变对愈合结果的影响：对患者特定康复策略的影响

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-05-01 Epub Date: 2026-02-06 DOI: 10.1016/j.cmpb.2026.109281

Qianjun Ding , Lunjian Li , Saeed Miramini , Shuangmin Shi , Peter Ebeling , Yongping Wei , Lihai Zhang

Background and objective

Partial weight-bearing (PWB) exercise is critical in fracture rehabilitation. However, the effects of tibial fracture-induced gait alterations on musculoskeletal loading conditions, particularly regarding body weights (BWs) and walking speeds, remain unclear. These patient-specific gait deviations can substantially modify mechanical stimuli at the fracture site, thereby affecting healing trajectories. This study aims to evaluate how fracture-altered gait mechanics influence loading conditions during rehabilitation to inform patient-specific strategies.

Method

A gait-based biomechanical model was developed to simulate PWB walking in patients with tibial fractures, based on changes in ground reaction forces (GRFs) compared to the uninjured limb. Fracture-induced effects on joint and muscle loading rates were quantified and integrated into the fracture healing model, simulating mesenchymal stem cell (MSC)-mediated tissue differentiation under various walking speeds and BWs.

Results

Tibial fracture-induced gait changes increased peak loading rates at knee and ankle joints by 0.9–1.3 BW/s and 0.8–1.2 BW/s, respectively, as speed rises from 0.6 to 1.0v₀(v₀ = 5 km/h). The tibialis posterior and rectus femoris exhibited the largest increase in peak loading rates, by 92 % and 220 %, respectively. Ignoring fracture effect could underestimate the mechanical stimulation and risk of fracture non-union. There exists an optimal walking speed for patient-specific BW to promote endochondral ossification, while controlling fibrous tissue formation (e.g., walking at 0.6 v₀ for a 65 kg patient under 30 % PWB).

Conclusion

This study offers clinically relevant insights to assist physiotherapists in prescribing effective, patient-specific gait rehabilitation strategies to enhance tibial fracture healing during early recovery.

背景与目的部分负重训练在骨折康复中起着至关重要的作用。然而，胫骨骨折引起的步态改变对肌肉骨骼负荷状况的影响，特别是对体重（BWs）和步行速度的影响尚不清楚。这些患者特有的步态偏差可以极大地改变骨折部位的机械刺激，从而影响愈合轨迹。本研究旨在评估骨折改变的步态力学如何影响康复期间的负荷条件，以告知患者特定的策略。方法建立基于步态的生物力学模型，模拟胫骨骨折患者的地面反作用力（GRFs）与未受伤肢体的变化。骨折诱导对关节和肌肉负荷率的影响被量化并整合到骨折愈合模型中，模拟不同步行速度和体重下间充质干细胞（MSC）介导的组织分化。结果胫骨骨折引起的步态变化，随着速度从0.6 ~ 1.0v0（v0 = 5 km/h）增加，膝关节和踝关节峰值负荷率分别增加0.9 ~ 1.3 BW/s和0.8 ~ 1.2 BW/s。胫骨后肌和股直肌的峰值负荷率分别增加了92%和220%，增幅最大。忽略骨折效应会低估机械刺激和骨折不愈合的风险。针对不同体重的患者，存在一个最佳的步行速度，以促进软骨内骨化，同时控制纤维组织的形成（例如，体重65公斤、体重30%以下的患者以0.6 v0的速度行走）。结论本研究提供了临床相关的见解，有助于物理治疗师在早期康复过程中制定有效的、针对患者的步态康复策略，以促进胫骨骨折愈合。

{"title":"Effects of tibial fracture-induced gait alterations on healing outcomes: Implications for patient-specific rehabilitation strategies","authors":"Qianjun Ding , Lunjian Li , Saeed Miramini , Shuangmin Shi , Peter Ebeling , Yongping Wei , Lihai Zhang","doi":"10.1016/j.cmpb.2026.109281","DOIUrl":"10.1016/j.cmpb.2026.109281","url":null,"abstract":"<div><h3>Background and objective</h3><div>Partial weight-bearing (PWB) exercise is critical in fracture rehabilitation. However, the effects of tibial fracture-induced gait alterations on musculoskeletal loading conditions, particularly regarding body weights (BWs) and walking speeds, remain unclear. These patient-specific gait deviations can substantially modify mechanical stimuli at the fracture site, thereby affecting healing trajectories. This study aims to evaluate how fracture-altered gait mechanics influence loading conditions during rehabilitation to inform patient-specific strategies.</div></div><div><h3>Method</h3><div>A gait-based biomechanical model was developed to simulate PWB walking in patients with tibial fractures, based on changes in ground reaction forces (GRFs) compared to the uninjured limb. Fracture-induced effects on joint and muscle loading rates were quantified and integrated into the fracture healing model, simulating mesenchymal stem cell (MSC)-mediated tissue differentiation under various walking speeds and BWs.</div></div><div><h3>Results</h3><div>Tibial fracture-induced gait changes increased peak loading rates at knee and ankle joints by 0.9–1.3 BW/s and 0.8–1.2 BW/s, respectively, as speed rises from 0.6 to 1.0v<sub>0</sub>(v<sub>0</sub> = 5 km/h). The tibialis posterior and rectus femoris exhibited the largest increase in peak loading rates, by 92 % and 220 %, respectively. Ignoring fracture effect could underestimate the mechanical stimulation and risk of fracture non-union. There exists an optimal walking speed for patient-specific BW to promote endochondral ossification, while controlling fibrous tissue formation (<em>e.g.,</em> walking at 0.6 <em>v</em><sub>0</sub> for a 65 kg patient under 30 % PWB).</div></div><div><h3>Conclusion</h3><div>This study offers clinically relevant insights to assist physiotherapists in prescribing effective, patient-specific gait rehabilitation strategies to enhance tibial fracture healing during early recovery.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"278 ","pages":"Article 109281"},"PeriodicalIF":4.8,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146186280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Stable EEG source estimation for standardized Kalman filter using rate-of-change tracking 基于变化率跟踪的标准化卡尔曼滤波稳定脑电源估计。

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-05-01 Epub Date: 2026-02-06 DOI: 10.1016/j.cmpb.2026.109280

Joonas Lahtinen

Background and objective:

Localization of brain activity is an important guiding tool for presurgical evaluation and treatment planning, particularly in the context of various brain-related diseases such as epilepsy. Since brain activity is highly dynamic and arises from the firing of neuronal networks, advanced spatiotemporal modeling is needed to capture this time-varying behavior accurately.

Methods:

A new parameter tuning and model utilizing the rate-of-change of brain activity distribution were developed to improve the filtering-parametrization-stability of the otherwise accurate estimation of the recently introduced Standardized Kalman filter. Namely, we propose a backward-differentiation-based measurement model for the rate of change. Simulated and real non-invasive electroencephalography data, along with realistic head models from two real subjects, were used in time-evolution tracking and localization experiments focusing on somatosensory evoked potentials. The method was compared to the original Standardized Kalman filter and Standardized Low-resolution Brain Electromagnetic Tomography (sLORETA).

Results:

Results indicate that the proposed parametrization yields high localization accuracy, as the original Standardized Kalman filtering localizes 7 and 6 out of 8, and sLORETA found 8 and 6 of the literature-defined originators of short latency somatosensory evoked potentials. The proposed standardized filtering method identified 7 of 8 expected originators. The change-rate-based model exhibits greater tracking stability than filtering without it against changes in filtering parameters. In addition, the method is more stable against badly set model parameters.

Conclusions:

With new model parametrization, the studied standardized methodologies provide assumably accurate and stable estimations to explore the location and dynamical properties of cortical and subcortical brain activity. The results showing correct sub-thalamic localization demonstrate the significant potential of these methods in the guidance of stereo-electroencephalography sensor placements or the placement of implant electrodes for deep-brain stimulation.

背景与目的：脑活动定位是术前评估和治疗计划的重要指导工具，特别是在各种脑相关疾病如癫痫的情况下。由于大脑活动是高度动态的，并且是由神经元网络的放电引起的，因此需要先进的时空建模来准确地捕捉这种时变行为。方法：开发了一种新的参数调整和模型，利用大脑活动分布的变化率来提高最近引入的标准化卡尔曼滤波器的滤波-参数化-稳定性，否则会准确估计。也就是说，我们提出了一个基于后向微分的变化率测量模型。模拟和真实的无创脑电图数据，以及来自两名真实受试者的真实头部模型，用于时间进化跟踪和定位实验，重点关注体感诱发电位。将该方法与原始的标准化卡尔曼滤波和标准化低分辨率脑电磁断层扫描（sLORETA）进行了比较。结果表明，所提出的参数化具有较高的定位精度，因为原始的标准化卡尔曼滤波定位了8个中的7个和6个，而sLORETA发现了8个和6个文献定义的短潜伏期体感诱发电位的起源。提出的标准化过滤方法确定了8个预期发起者中的7个。对于滤波参数的变化，基于变化率的模型比没有变化率的模型表现出更好的跟踪稳定性。此外，该方法在模型参数设置较差的情况下具有较好的稳定性。结论：通过新的模型参数化，所研究的标准化方法为探索大脑皮层和皮层下活动的位置和动态特性提供了假设的准确和稳定的估计。显示正确的丘脑下定位的结果表明，这些方法在指导立体脑电图传感器的放置或植入电极的放置以进行深部脑刺激方面具有重要的潜力。

{"title":"Stable EEG source estimation for standardized Kalman filter using rate-of-change tracking","authors":"Joonas Lahtinen","doi":"10.1016/j.cmpb.2026.109280","DOIUrl":"10.1016/j.cmpb.2026.109280","url":null,"abstract":"<div><h3>Background and objective:</h3><div>Localization of brain activity is an important guiding tool for presurgical evaluation and treatment planning, particularly in the context of various brain-related diseases such as epilepsy. Since brain activity is highly dynamic and arises from the firing of neuronal networks, advanced spatiotemporal modeling is needed to capture this time-varying behavior accurately.</div></div><div><h3>Methods:</h3><div>A new parameter tuning and model utilizing the rate-of-change of brain activity distribution were developed to improve the filtering-parametrization-stability of the otherwise accurate estimation of the recently introduced Standardized Kalman filter. Namely, we propose a backward-differentiation-based measurement model for the rate of change. Simulated and real non-invasive electroencephalography data, along with realistic head models from two real subjects, were used in time-evolution tracking and localization experiments focusing on somatosensory evoked potentials. The method was compared to the original Standardized Kalman filter and Standardized Low-resolution Brain Electromagnetic Tomography (sLORETA).</div></div><div><h3>Results:</h3><div>Results indicate that the proposed parametrization yields high localization accuracy, as the original Standardized Kalman filtering localizes 7 and 6 out of 8, and sLORETA found 8 and 6 of the literature-defined originators of short latency somatosensory evoked potentials. The proposed standardized filtering method identified 7 of 8 expected originators. The change-rate-based model exhibits greater tracking stability than filtering without it against changes in filtering parameters. In addition, the method is more stable against badly set model parameters.</div></div><div><h3>Conclusions:</h3><div>With new model parametrization, the studied standardized methodologies provide assumably accurate and stable estimations to explore the location and dynamical properties of cortical and subcortical brain activity. The results showing correct sub-thalamic localization demonstrate the significant potential of these methods in the guidance of stereo-electroencephalography sensor placements or the placement of implant electrodes for deep-brain stimulation.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"278 ","pages":"Article 109280"},"PeriodicalIF":4.8,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146177338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust multimodal mental workload classification: A cross-physiological condition machine learning approach 鲁棒多模态脑力工作负荷分类：一种跨生理条件机器学习方法

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-04-01 Epub Date: 2026-01-14 DOI: 10.1016/j.cmpb.2026.109251

Anais Pontiggia , Michael Quiquempoix , Pierre Fabries , Vincent Beauchamps , Clémentine Jacques , Mathias Guillard , Pascal Van Beers , Carine Malle , Danielle Gomez-Merino , Nathalie Koulmann , Mounir Chennaoui , Fabien Sauvet , HYPSOM Investigator Group

Background and Objective

Aircraft pilots can be faced with a high mental workload (MW) combined with moderate hypoxia and sleep restriction. We aimed to assess the cross-validation of a machine learning-based MW predictive model under hypoxia and/or sleep restriction. Secondly, we developed a robust predictive model using multimodal physiological parameters to improve the validity across different physiological conditions.

Methods

Seventeen healthy participants were randomly exposed to three 12-minute periods of increased MW (low, medium, and high) in a 4-condition crossover design: sleep restriction (SR, <3 h Total Sleep Time, TST) vs. habitual sleep (HS, >6 h TST), hypoxia (HY, 2 h, F_IO₂=13.6%, ∼3500 m) vs. normoxia (NO, F_IO₂=21%). MW levels were designed using the Multi-Attribute Test Battery (MATB)-II with an additional auditory Oddball-like task. Six machine learning classifiers were compared. Features selection (from EEG, ECG, respiratory and eye tracking sensors) was performed using backward Recursive Feature Elimination (RFE).

Results

The best models for 1-minute MW levels classification on HSNO were K-Nearest Neighbors (KNN, F1 score = 80.3 ± 8.9%), Support Vector Machine (SVM, 77.8 ± 10.3%) and Random Forest (RF, 75.7 ± 9.1%). Exposure to sleep restriction and/or hypoxia decreased models’ performance (F1 <35%). KNN and RF models, in particular those including EEG and eye tracking, trained on All-Conditions performed well across conditions (F1 scores = 77.4 ± 7.8% and 70.7 ± 10.2%).

Conclusion

Our results highlight the need for training MW models under different physiological constraints and using multimodal datasets to improve robustness. (NCT05563688)

背景与目的飞机飞行员可能面临高精神负荷（MW），并伴有中度缺氧和睡眠限制。我们的目的是评估缺氧和/或睡眠限制下基于机器学习的脑卒中预测模型的交叉验证。其次，建立了基于多模态生理参数的鲁棒预测模型，提高了不同生理条件下的有效性。方法17名健康参与者随机暴露于3个12分钟的增加MW（低、中、高）4条件交叉设计：睡眠限制（SR， 3小时总睡眠时间，TST） vs习惯性睡眠（HS， 6小时总睡眠时间），缺氧（HY， 2小时，FIO2=13.6%, ~ 3500 m） vs正常缺氧（NO, FIO2=21%）。使用多属性测试电池(MATB)-II设计MW级别，并附加一个听觉怪异任务。比较了六种机器学习分类器。采用反向递归特征消除（RFE）对EEG、ECG、呼吸和眼动传感器进行特征选择。结果对HSNO进行1分钟MW水平分类的最佳模型为k近邻（KNN， F1评分为80.3±8.9%）、支持向量机（SVM, 77.8±10.3%）和随机森林（RF, 75.7±9.1%）。睡眠限制和/或缺氧会降低模特的表现（F1 <35%）。在所有条件下训练的KNN和RF模型，特别是包括脑电图和眼动追踪的模型，在所有条件下都表现良好（F1得分= 77.4±7.8%和70.7±10.2%）。结论我们的研究结果表明，需要在不同的生理约束下训练MW模型，并使用多模态数据集来提高鲁棒性。(NCT05563688)

{"title":"Robust multimodal mental workload classification: A cross-physiological condition machine learning approach","authors":"Anais Pontiggia , Michael Quiquempoix , Pierre Fabries , Vincent Beauchamps , Clémentine Jacques , Mathias Guillard , Pascal Van Beers , Carine Malle , Danielle Gomez-Merino , Nathalie Koulmann , Mounir Chennaoui , Fabien Sauvet , HYPSOM Investigator Group","doi":"10.1016/j.cmpb.2026.109251","DOIUrl":"10.1016/j.cmpb.2026.109251","url":null,"abstract":"<div><h3>Background and Objective</h3><div>Aircraft pilots can be faced with a high mental workload (MW) combined with moderate hypoxia and sleep restriction. We aimed to assess the cross-validation of a machine learning-based MW predictive model under hypoxia and/or sleep restriction. Secondly, we developed a robust predictive model using multimodal physiological parameters to improve the validity across different physiological conditions.</div></div><div><h3>Methods</h3><div>Seventeen healthy participants were randomly exposed to three 12-minute periods of increased MW (low, medium, and high) in a 4-condition crossover design: sleep restriction (SR, <3 h Total Sleep Time, TST) vs. habitual sleep (HS, >6 h TST), hypoxia (HY, 2 h, F<sub>I</sub>O<sub>2</sub>=13.6%, ∼3500 m) vs. normoxia (NO, F<sub>I</sub>O<sub>2</sub>=21%). MW levels were designed using the Multi-Attribute Test Battery (MATB)-II with an additional auditory Oddball-like task. Six machine learning classifiers were compared. Features selection (from EEG, ECG, respiratory and eye tracking sensors) was performed using backward Recursive Feature Elimination (RFE).</div></div><div><h3>Results</h3><div>The best models for 1-minute MW levels classification on HSNO were K-Nearest Neighbors (KNN, F1 score = 80.3 ± 8.9%), Support Vector Machine (SVM, 77.8 ± 10.3%) and Random Forest (RF, 75.7 ± 9.1%). Exposure to sleep restriction and/or hypoxia decreased models’ performance (F1 <35%). KNN and RF models, in particular those including EEG and eye tracking, trained on All-Conditions performed well across conditions (F1 scores = 77.4 ± 7.8% and 70.7 ± 10.2%).</div></div><div><h3>Conclusion</h3><div>Our results highlight the need for training MW models under different physiological constraints and using multimodal datasets to improve robustness. (NCT05563688)</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"277 ","pages":"Article 109251"},"PeriodicalIF":4.8,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhancing accuracy and explainability in colorectal lesion classification with attention-supervised Vision Transformers 使用注意监督视觉变压器提高结直肠病变分类的准确性和可解释性

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-04-01 Epub Date: 2026-01-21 DOI: 10.1016/j.cmpb.2026.109260

Luca Carlini , Luca Di Stefano , Chiara Lena , Davide Massimi , Tommy Rizkala , Cesare Hassan , Elena De Momi

Objective:

Accurate assessment of colorectal lesion morphology during colonoscopy is essential for guiding treatment and estimating cancer risk. The Paris classification is widely adopted for this purpose but suffers from substantial inter-observer variability, while Vision Transformers (ViTs) can base their decisions on diffuse, off-lesion attention patterns that are hard to interpret. This study investigates whether directly supervising ViT attention maps with expert lesion annotations can concurrently improve Paris classification performance and model explainability.

Method:

We propose a Lesion-Focused Attention Loss (

L_{LFA}

), an attention-supervised pretraining objective that uses expert polyp bounding boxes to focus last-layer [CLS] attention on annotated lesion regions, followed by standard cross-entropy fine-tuning.

L_{LFA}

is applied to six ViT architectures and evaluated on the public SUN dataset for binary (0-I vs. 0-II) and three-class (0-Ip, 0-Is, 0-IIa) Paris classification. Performance is assessed using frame-wise accuracy and the

AttIn

, we additionally perform an ablation study against a Grad-CAM consistency baseline.

Results:

Attention-supervised pretraining yields consistent gains in both accuracy and lesion-focused attention. Across the six ViTs, adding

L_{LFA}

improves three-class accuracy by up to 7 percentage points. In a detailed ablation on ViT-B/16,

L_{LFA}

outperforms a Grad-CAM consistency baseline by about 5–13 percentage points across the 2-class and 3-class tasks, and

χ^{2}

tests confirm a significant association between high

AttIn

and correct predictions.

Conclusion:

Direct supervision of ViT attention with

L_{LFA}

leverages expert knowledge to jointly boost Paris classification accuracy and spatial interpretability, and compares favourably with Grad-CAM–based explanation regularisation. The source code and dataset splits are publicly available at https://github.com/LucaCarlini/SUNDatasetPretraining.

目的：在结肠镜检查中准确评估结直肠病变形态对指导治疗和评估癌变风险至关重要。巴黎分类被广泛用于此目的，但存在大量观察者之间的可变性，而视觉变形器（ViTs）可以根据难以解释的弥漫性，非病变注意模式做出决策。本研究探讨了直接监督ViT注意图与专家病变注释是否可以同时提高Paris分类性能和模型的可解释性。方法：我们提出了一种病灶聚焦注意力损失（LLFA），这是一种注意力监督的预训练目标，它使用专家息肉边界盒将最后一层[CLS]注意力集中在注释的病变区域上，然后进行标准的交叉熵微调。LLFA应用于六个ViT架构，并在公共SUN数据集上对二进制（0-I vs. 0-II）和三级（0-Ip, 0-Is, 0-IIa） Paris分类进行了评估。使用逐帧精度和AttIn对性能进行评估，我们还针对Grad-CAM一致性基线进行了消融研究。结果：注意监督的预训练在准确性和病灶集中注意力方面都有一致的收获。在六个vit中，添加LLFA将三级精度提高了7个百分点。在ViT-B/16的详细消融中，LLFA在2级和3级任务中优于Grad-CAM一致性基线约5-13个百分点，χ2检验证实了高AttIn与正确预测之间的显着关联。结论：利用LLFA对ViT注意力进行直接监督，利用专家知识共同提高了Paris分类精度和空间可解释性，且优于基于grad - cam的解释正则化。源代码和数据集拆分可以在https://github.com/LucaCarlini/SUNDatasetPretraining上公开获得。

{"title":"Enhancing accuracy and explainability in colorectal lesion classification with attention-supervised Vision Transformers","authors":"Luca Carlini , Luca Di Stefano , Chiara Lena , Davide Massimi , Tommy Rizkala , Cesare Hassan , Elena De Momi","doi":"10.1016/j.cmpb.2026.109260","DOIUrl":"10.1016/j.cmpb.2026.109260","url":null,"abstract":"<div><h3>Objective:</h3><div>Accurate assessment of colorectal lesion morphology during colonoscopy is essential for guiding treatment and estimating cancer risk. The Paris classification is widely adopted for this purpose but suffers from substantial inter-observer variability, while Vision Transformers (ViTs) can base their decisions on diffuse, off-lesion attention patterns that are hard to interpret. This study investigates whether directly supervising ViT attention maps with expert lesion annotations can concurrently improve Paris classification performance and model explainability.</div></div><div><h3>Method:</h3><div>We propose a Lesion-Focused Attention Loss (<span><math><msub><mrow><mi>L</mi></mrow><mrow><mtext>LFA</mtext></mrow></msub></math></span>), an attention-supervised pretraining objective that uses expert polyp bounding boxes to focus last-layer <span>[CLS]</span> attention on annotated lesion regions, followed by standard cross-entropy fine-tuning. <span><math><msub><mrow><mi>L</mi></mrow><mrow><mtext>LFA</mtext></mrow></msub></math></span> is applied to six ViT architectures and evaluated on the public SUN dataset for binary (0-I vs. 0-II) and three-class (0-Ip, 0-Is, 0-IIa) Paris classification. Performance is assessed using frame-wise accuracy and the <span><math><mi>AttIn</mi></math></span>, we additionally perform an ablation study against a Grad-CAM consistency baseline.</div></div><div><h3>Results:</h3><div>Attention-supervised pretraining yields consistent gains in both accuracy and lesion-focused attention. Across the six ViTs, adding <span><math><msub><mrow><mi>L</mi></mrow><mrow><mtext>LFA</mtext></mrow></msub></math></span> improves three-class accuracy by up to 7 percentage points. In a detailed ablation on ViT-B/16, <span><math><msub><mrow><mi>L</mi></mrow><mrow><mtext>LFA</mtext></mrow></msub></math></span> outperforms a Grad-CAM consistency baseline by about 5–13 percentage points across the 2-class and 3-class tasks, and <span><math><msup><mrow><mi>χ</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> tests confirm a significant association between high <span><math><mi>AttIn</mi></math></span> and correct predictions.</div></div><div><h3>Conclusion:</h3><div>Direct supervision of ViT attention with <span><math><msub><mrow><mi>L</mi></mrow><mrow><mtext>LFA</mtext></mrow></msub></math></span> leverages expert knowledge to jointly boost Paris classification accuracy and spatial interpretability, and compares favourably with Grad-CAM–based explanation regularisation. The source code and dataset splits are publicly available at <span><span>https://github.com/LucaCarlini/SUNDatasetPretraining</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"277 ","pages":"Article 109260"},"PeriodicalIF":4.8,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024467","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Visual prompt tuning for task-flexible medical image synthesis 任务灵活的医学图像合成的视觉提示调整

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-04-01 Epub Date: 2026-01-09 DOI: 10.1016/j.cmpb.2026.109244

Jonghun Kim, Hyunjin Park

Background and objective

: Medical image synthesis has broad applications in modality-to-modality translation, denoising, and super-resolution, when specific modalities are missing, various types of noise occur, or resolution discrepancies exist across modalities. The traditional approach requires a separate model for each task, making it inefficient, and nearly impossible to accommodate various tasks in medical image synthesis.

Methods:

We introduce a task-agnostic medical image-synthesis model utilizing prompt tuning that leverages a diffusion model and prompt tuning to fine-tune large capacity pretrained models efficiently. Our method can handle multiple tasks that cover various input–output combinations in a single model.

Results:

Our method can perform denoising, translation, super-resolution, and tumor inpainting tasks for brain MRI and abdominal CT. Through quantitative and qualitative evaluations, we demonstrate that our model achieves the best performance in terms of FID scores across all evaluated tasks. Our multi-task model achieves a PSNR of 25.76 and an SSIM of 0.908 for T1-to-T2 translation; a PSNR of 30.30 and an SSIM of 0.932 for denoising; a PSNR of 29.24 and an SSIM of 0.874 for super-resolution; and an FID of 16.18 with an LPIPS of 0.090 for tumor inpainting.

Conclusions:

We proposed a method that enables task-agnostic medical image synthesis, allowing for the specification of the desired synthesis task, modality, and organ of the target image via prompt tuning. Our method can be extended to other modalities and organs. The code is available at https://github.com/jongdory/VPT-Med.

背景和目的：医学图像合成在模态到模态的转换、去噪和超分辨率方面有着广泛的应用，当特定模态缺失时，会出现各种类型的噪声，或者在模态之间存在分辨率差异。传统的方法需要为每个任务单独的模型，使其效率低下，并且几乎不可能适应医学图像合成中的各种任务。方法：我们引入了一个任务不可知的医学图像合成模型，该模型利用扩散模型和提示调整来有效地微调大容量预训练模型。我们的方法可以处理多个任务，涵盖单个模型中的各种输入输出组合。结果：该方法可以完成脑MRI和腹部CT的去噪、平移、超分辨率和肿瘤成像任务。通过定量和定性评估，我们证明我们的模型在所有评估任务的FID得分方面达到了最佳表现。我们的多任务模型实现了t1到t2翻译的PSNR为25.76，SSIM为0.908；去噪的PSNR为30.30，SSIM为0.932；超分辨率的PSNR为29.24，SSIM为0.874；FID为16.18，LPIPS为0.090。结论：我们提出了一种方法，使任务不可知的医学图像合成，允许规范所需的合成任务，模式，并通过及时调整目标图像的器官。我们的方法可以扩展到其他形态和器官。代码可在https://github.com/jongdory/VPT-Med上获得。

{"title":"Visual prompt tuning for task-flexible medical image synthesis","authors":"Jonghun Kim, Hyunjin Park","doi":"10.1016/j.cmpb.2026.109244","DOIUrl":"10.1016/j.cmpb.2026.109244","url":null,"abstract":"<div><h3>Background and objective</h3><div>: Medical image synthesis has broad applications in modality-to-modality translation, denoising, and super-resolution, when specific modalities are missing, various types of noise occur, or resolution discrepancies exist across modalities. The traditional approach requires a separate model for each task, making it inefficient, and nearly impossible to accommodate various tasks in medical image synthesis.</div></div><div><h3>Methods:</h3><div>We introduce a task-agnostic medical image-synthesis model utilizing prompt tuning that leverages a diffusion model and prompt tuning to fine-tune large capacity pretrained models efficiently. Our method can handle multiple tasks that cover various input–output combinations in a single model.</div></div><div><h3>Results:</h3><div>Our method can perform denoising, translation, super-resolution, and tumor inpainting tasks for brain MRI and abdominal CT. Through quantitative and qualitative evaluations, we demonstrate that our model achieves the best performance in terms of FID scores across all evaluated tasks. Our multi-task model achieves a PSNR of 25.76 and an SSIM of 0.908 for T1-to-T2 translation; a PSNR of 30.30 and an SSIM of 0.932 for denoising; a PSNR of 29.24 and an SSIM of 0.874 for super-resolution; and an FID of 16.18 with an LPIPS of 0.090 for tumor inpainting.</div></div><div><h3>Conclusions:</h3><div>We proposed a method that enables task-agnostic medical image synthesis, allowing for the specification of the desired synthesis task, modality, and organ of the target image via prompt tuning. Our method can be extended to other modalities and organs. The code is available at <span><span>https://github.com/jongdory/VPT-Med</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"277 ","pages":"Article 109244"},"PeriodicalIF":4.8,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145924500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Shape matters: Predicting Huntington’s disease using progression modelling 形状很重要：使用进展模型预测亨廷顿氏病。

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-04-01 Epub Date: 2026-01-15 DOI: 10.1016/j.cmpb.2026.109250

Mohsen Ghofrani-Jahromi , Susmita Saha , Adeel Razi , Pubu M. Abeyasinghe , Govinda R. Poudel , Jane S. Paulsen , Sarah J. Tabrizi , Nellie Georgiou-Karistianis

Background

Despite evidence of group-level differences in striatal morphometry among persons with Huntington’s Disease (PwHD), current models of HD progression used for participant selection and assessment of treatment outcomes in clinical trials do not leverage shape information.

Methods

We first validated the capability of a discriminative deep neural network to derive descriptors of shape from all subcortical structures affected by HD, utilizing 2,932 brain scans in 615 PwHD across three longitudinal datasets (TRACK-HD, PREDICT-HD, and IMAGE-HD). We then trained a conditional generative model that used shape descriptors, alongside conventional volumetric, genetic, as well as composite cognitive, motor, and functional features at baseline to predict biomarkers of disease progression at subsequent time points.

Results

We observed that the anatomical shapes of subcortical structures, including putamen, lateral ventricle, pallidum, caudate, thalamus, and accumbens, exhibited strong associations with HD progression, as measured by a commonly used prognostic score. Furthermore, within-stage heterogeneity, along the continuum of disease progression, was better captured: when shape descriptors were aggregated using principal component analysis, they showed a high correlation with disease stage (Spearman’s correlation: ρ = 0.72), compared to volumetric measurements in cubic millimetres (ρ = 0.45). Finally, incorporating subcortical shape into the generative model improved predictive performance, compared to the same model that relied solely on brain volumes.

Conclusion

This study demonstrates that subcortical brain shape is associated with HD progression, enables capturing fine-grained within-stage variability, and improves the predictability of characteristic biomarkers. The findings could potentially optimize future clinical trials through more targeted participant recruitment and more objective post-intervention assessments of treatment efficacy.

背景：尽管有证据表明亨廷顿舞蹈病（PwHD）患者纹状体形态在组水平上存在差异，但目前用于临床试验中参与者选择和治疗结果评估的亨廷顿舞蹈病进展模型并未利用形状信息。方法：我们首先验证了判别深度神经网络从所有受HD影响的皮质下结构中获得形状描述符的能力，利用615名PwHD患者的2,932次脑部扫描，跨越三个纵向数据集（TRACK-HD， PREDICT-HD和IMAGE-HD）。然后，我们训练了一个条件生成模型，该模型使用形状描述符，以及传统的体积、遗传以及复合认知、运动和功能特征作为基线，以预测随后时间点疾病进展的生物标志物。结果：我们观察到皮质下结构的解剖形状，包括壳核、侧脑室、白球、尾状核、丘脑和伏隔核，通过常用的预后评分显示与HD进展有很强的相关性。此外，分期内异质性，沿着疾病进展的连续体，被更好地捕获：当形状描述符使用主成分分析聚合时，与立方毫米的体积测量（ρ = 0.45）相比，它们显示出与疾病分期的高度相关性（Spearman相关性：ρ = 0.72）。最后，与仅依赖脑容量的相同模型相比，将皮层下形状纳入生成模型提高了预测性能。结论：该研究表明，皮质下脑形状与HD进展相关，能够捕获细粒度的期内变异性，并提高特征生物标志物的可预测性。研究结果可能通过更有针对性的参与者招募和更客观的干预后治疗效果评估来优化未来的临床试验。

{"title":"Shape matters: Predicting Huntington’s disease using progression modelling","authors":"Mohsen Ghofrani-Jahromi , Susmita Saha , Adeel Razi , Pubu M. Abeyasinghe , Govinda R. Poudel , Jane S. Paulsen , Sarah J. Tabrizi , Nellie Georgiou-Karistianis","doi":"10.1016/j.cmpb.2026.109250","DOIUrl":"10.1016/j.cmpb.2026.109250","url":null,"abstract":"<div><h3>Background</h3><div>Despite evidence of group-level differences in striatal morphometry among persons with Huntington’s Disease (PwHD), current models of HD progression used for participant selection and assessment of treatment outcomes in clinical trials do not leverage shape information.</div></div><div><h3>Methods</h3><div>We first validated the capability of a discriminative deep neural network to derive descriptors of shape from all subcortical structures affected by HD, utilizing 2,932 brain scans in 615 PwHD across three longitudinal datasets (TRACK-HD, PREDICT-HD, and IMAGE-HD). We then trained a conditional generative model that used shape descriptors, alongside conventional volumetric, genetic, as well as composite cognitive, motor, and functional features at baseline to predict biomarkers of disease progression at subsequent time points.</div></div><div><h3>Results</h3><div>We observed that the anatomical shapes of subcortical structures, including putamen, lateral ventricle, pallidum, caudate, thalamus, and accumbens, exhibited strong associations with HD progression, as measured by a commonly used prognostic score. Furthermore, within-stage heterogeneity, along the continuum of disease progression, was better captured: when shape descriptors were aggregated using principal component analysis, they showed a high correlation with disease stage (Spearman’s correlation: ρ = 0.72), compared to volumetric measurements in cubic millimetres (ρ = 0.45). Finally, incorporating subcortical shape into the generative model improved predictive performance, compared to the same model that relied solely on brain volumes.</div></div><div><h3>Conclusion</h3><div>This study demonstrates that subcortical brain shape is associated with HD progression, enables capturing fine-grained within-stage variability, and improves the predictability of characteristic biomarkers. The findings could potentially optimize future clinical trials through more targeted participant recruitment and more objective post-intervention assessments of treatment efficacy.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"277 ","pages":"Article 109250"},"PeriodicalIF":4.8,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146017616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Application of artificial intelligence in colonoscopy imaging for polyp analysis-A systematic review 人工智能在息肉分析结肠镜成像中的应用综述。

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-04-01 Epub Date: 2026-01-07 DOI: 10.1016/j.cmpb.2026.109239

Elham amirmohammadi , Ahmad shalbaf , Ali Esteki , Amir sadeghi , Alireza ramazani moghadam , Mina moghtaderi , Pardis ketabi moghadam

The colon is a major component of the digestive system, so early detection of colorectal polyps is essential in preventing colorectal cancer, which is a leading cause of cancer-related death worldwide. While colonoscopy remains the gold standard for polyp detection, its diagnostic accuracy is highly operator-dependent. Recent advances in Deep Learning (DL), a branch of Artificial Intelligence (AI), have shown substantial potential to improve colonoscopy image analysis by enhancing the accuracy, consistency, and objectivity of polyp detection, segmentation, and classification. Artificial intelligence-based systems have significantly reduced inter-observer variability and increased diagnostic efficiency, ultimately transforming the landscape of colorectal lesion assessment. This survey provides a comprehensive and critical analysis of the current status of deep learning applications in colorectal polyp analysis. We systematically review state-of-the-art methodologies across various DL architectures—including Convolutional Neural Networks (CNNs), transformer-based models, and hybrid approaches—and examine their performance on publicly available benchmark datasets. Additionally, we highlight the strengths and limitations of existing techniques, explore the clinical relevance of AI-assisted tools, and identify prevailing challenges such as data imbalance, real-time deployment, and generalizability across diverse populations and colonoscopy devices. By consolidating key advances and outlining future research directions, this review aims to serve as a valuable resource for researchers, clinicians, and developers seeking to leverage deep learning to enhance colorectal polyp detection, diagnosis, and clinical decision-making.

结肠是消化系统的主要组成部分，因此早期发现结肠息肉对于预防结直肠癌至关重要，结直肠癌是全球癌症相关死亡的主要原因。虽然结肠镜检查仍然是息肉检测的金标准，但其诊断准确性高度依赖于操作者。深度学习（DL）是人工智能（AI）的一个分支，其最新进展显示出通过提高息肉检测、分割和分类的准确性、一致性和客观性来改善结肠镜图像分析的巨大潜力。基于人工智能的系统显著降低了观察者之间的差异，提高了诊断效率，最终改变了结直肠病变评估的格局。本调查对深度学习在结肠直肠息肉分析中的应用现状进行了全面和批判性的分析。我们系统地回顾了各种深度学习架构（包括卷积神经网络（cnn）、基于变压器的模型和混合方法）的最新方法，并检查了它们在公开可用的基准数据集上的性能。此外，我们强调现有技术的优势和局限性，探索人工智能辅助工具的临床相关性，并确定当前的挑战，如数据不平衡、实时部署和不同人群和结肠镜检查设备的普遍性。通过整合关键进展并概述未来的研究方向，本综述旨在为研究人员、临床医生和开发人员提供宝贵的资源，以寻求利用深度学习来增强结肠直肠息肉的检测、诊断和临床决策。

{"title":"Application of artificial intelligence in colonoscopy imaging for polyp analysis-A systematic review","authors":"Elham amirmohammadi , Ahmad shalbaf , Ali Esteki , Amir sadeghi , Alireza ramazani moghadam , Mina moghtaderi , Pardis ketabi moghadam","doi":"10.1016/j.cmpb.2026.109239","DOIUrl":"10.1016/j.cmpb.2026.109239","url":null,"abstract":"<div><div>The colon is a major component of the digestive system, so early detection of colorectal polyps is essential in preventing colorectal cancer, which is a leading cause of cancer-related death worldwide. While colonoscopy remains the gold standard for polyp detection, its diagnostic accuracy is highly operator-dependent. Recent advances in Deep Learning (DL), a branch of Artificial Intelligence (AI), have shown substantial potential to improve colonoscopy image analysis by enhancing the accuracy, consistency, and objectivity of polyp detection, segmentation, and classification. Artificial intelligence-based systems have significantly reduced inter-observer variability and increased diagnostic efficiency, ultimately transforming the landscape of colorectal lesion assessment. This survey provides a comprehensive and critical analysis of the current status of deep learning applications in colorectal polyp analysis. We systematically review state-of-the-art methodologies across various DL architectures—including Convolutional Neural Networks (CNNs), transformer-based models, and hybrid approaches—and examine their performance on publicly available benchmark datasets. Additionally, we highlight the strengths and limitations of existing techniques, explore the clinical relevance of AI-assisted tools, and identify prevailing challenges such as data imbalance, real-time deployment, and generalizability across diverse populations and colonoscopy devices. By consolidating key advances and outlining future research directions, this review aims to serve as a valuable resource for researchers, clinicians, and developers seeking to leverage deep learning to enhance colorectal polyp detection, diagnosis, and clinical decision-making.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"277 ","pages":"Article 109239"},"PeriodicalIF":4.8,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146099529","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A diagnosis tool for early detection and classification of heart disease in individuals using transformer mechanisms 使用变压器机制的个体早期发现和分类心脏病的诊断工具

IF 4.8 2区医学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer methods and programs in biomedicine

Pub Date : 2026-04-01 Epub Date: 2026-01-15 DOI: 10.1016/j.cmpb.2026.109248

Saba Arif , Sang Hyeok Son , Hyun Yong Kim , Sang-Chul Kim , Jong Yun Lee

Background

Heart disease remains a leading cause of mortality, making accurate and efficient prediction tools essential for the general population. In medical diagnosis, deep learning-based approaches have shown significant potential in identifying complex patterns within clinical data.

Methods

This paper proposes a transformer-based model, named the Heart-Doctor for self-detection and classification of heart disease in the general population, addressing the critical challenge of early diagnosis and real-life risk assessment. The proposed model, inspired by the transformer architecture, processes 16 key attributes through a multi-layer transformer encoder, leveraging residual connections and deep feature extraction for improved classification. The proposed method utilizes electronic health records (EHR) from Chungbuk National University Hospital (CBNUH), incorporating symptom-based features along with common clinical attributes such as diagnoses and medical history to enhance predictive performance. The proposed model’s effectiveness is evaluated by implementing and comparing it with Convolutional Neural Network (CNN) and Random Forest (RF) classifiers.

Results

Experimental results demonstrate that the proposed model achieves 99% accuracy, compared to RF (98.79%) and CNN (97.64%), with superior precision, recall, and F1-scores, making it a highly effective tool for multi-class heart disease classification. To ensure real-world applicability, the study also includes the development of an Android-based application that integrates the model for real-time risk assessment, thereby enabling healthcare professionals to make timely and data-driven clinical decisions.

Conclusion

Consequently, the performance of the proposed transformer-based diagnosis model outperforms other models for heart disease classification, allowing individuals to detect their heart disease symptoms early and independently in real life.

背景：心脏病仍然是导致死亡的主要原因，因此准确有效的预测工具对普通人群至关重要。在医学诊断中，基于深度学习的方法在识别临床数据中的复杂模式方面显示出巨大的潜力。方法提出了一种基于变压器的心脏医生模型，用于普通人群心脏病的自我检测和分类，解决了早期诊断和现实生活风险评估的关键挑战。该模型受变压器结构的启发，通过多层变压器编码器处理16个关键属性，利用剩余连接和深度特征提取来改进分类。该方法利用忠北大学医院（CBNUH）的电子健康记录（EHR），将基于症状的特征与诊断和病史等常见临床属性结合起来，以提高预测性能。通过实现并与卷积神经网络（CNN）和随机森林（RF）分类器进行比较，评估了该模型的有效性。结果实验结果表明，与RF（98.79%）和CNN（97.64%）相比，该模型的准确率达到99%，具有更高的准确率、召回率和f1分数，是一种非常有效的多类别心脏病分类工具。为了确保实际应用，该研究还包括基于android的应用程序的开发，该应用程序集成了用于实时风险评估的模型，从而使医疗保健专业人员能够及时做出数据驱动的临床决策。因此，基于变压器的诊断模型的性能优于其他心脏病分类模型，使个体能够在现实生活中早期独立地发现自己的心脏病症状。

{"title":"A diagnosis tool for early detection and classification of heart disease in individuals using transformer mechanisms","authors":"Saba Arif , Sang Hyeok Son , Hyun Yong Kim , Sang-Chul Kim , Jong Yun Lee","doi":"10.1016/j.cmpb.2026.109248","DOIUrl":"10.1016/j.cmpb.2026.109248","url":null,"abstract":"<div><h3>Background</h3><div>Heart disease remains a leading cause of mortality, making accurate and efficient prediction tools essential for the general population. In medical diagnosis, deep learning-based approaches have shown significant potential in identifying complex patterns within clinical data.</div></div><div><h3>Methods</h3><div>This paper proposes a transformer-based model, named the Heart-Doctor for self-detection and classification of heart disease in the general population, addressing the critical challenge of early diagnosis and real-life risk assessment. The proposed model, inspired by the transformer architecture, processes 16 key attributes through a multi-layer transformer encoder, leveraging residual connections and deep feature extraction for improved classification. The proposed method utilizes electronic health records (EHR) from Chungbuk National University Hospital (CBNUH), incorporating symptom-based features along with common clinical attributes such as diagnoses and medical history to enhance predictive performance. The proposed model’s effectiveness is evaluated by implementing and comparing it with Convolutional Neural Network (CNN) and Random Forest (RF) classifiers.</div></div><div><h3>Results</h3><div>Experimental results demonstrate that the proposed model achieves 99% accuracy, compared to RF (98.79%) and CNN (97.64%), with superior precision, recall, and F1-scores, making it a highly effective tool for multi-class heart disease classification. To ensure real-world applicability, the study also includes the development of an Android-based application that integrates the model for real-time risk assessment, thereby enabling healthcare professionals to make timely and data-driven clinical decisions.</div></div><div><h3>Conclusion</h3><div>Consequently, the performance of the proposed transformer-based diagnosis model outperforms other models for heart disease classification, allowing individuals to detect their heart disease symptoms early and independently in real life.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"277 ","pages":"Article 109248"},"PeriodicalIF":4.8,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0