Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.07.007
Sijia Yan , Xi Ming , Jiaying Wu, Xiaojian Zhu, Mi Zhou, Yi Xiao
Objective
Chimeric antigen receptor T (CAR-T) cell therapy has demonstrated significant efficacy in the treatment of hematologic malignancies, yet it can lead to severe complications. Intracranial hemorrhage (ICH), although rare, is a particularly lethal outcome. Current research on the risk factors and underlying mechanisms of ICH post-CAR-T therapy remains limited.
Methods
This retrospective study analyzed 10 cases of ICH among 2255 patients who underwent CAR-T therapy at Tongji Hospital in Wuhan from January 2015 to December 2024. We collected data on baseline characteristics, CAR-T treatment parameters, inflammatory markers, coagulation function, and clinical outcomes.
Results
The incidence of ICH was 0.44 %, with the median time to onset being 25 days following infusion. All affected patients exhibited thrombocytopenia, and 80 % experienced coagulopathy. Levels of C-reactive protein and Interleukin-6 were significantly elevated before and after treatment. The mortality rate was 90 %, with nine patients succumbing to respiratory and circulatory failure associated with ICH.
Conclusion
The development of ICH after CAR-T therapy is strongly linked to thrombocytopenia, coagulopathy, and the inflammatory response triggered by cytokine release syndrome (CRS). Early identification and proactive management of high-risk patients could potentially improve outcomes. Further prospective studies are necessary to confirm risk prediction models and enhance therapeutic strategies.
{"title":"Fatal complication unveiled: analyzing intracerebral hemorrhage risk after CAR-T therapy in hematologic malignancies","authors":"Sijia Yan , Xi Ming , Jiaying Wu, Xiaojian Zhu, Mi Zhou, Yi Xiao","doi":"10.1016/j.hest.2025.07.007","DOIUrl":"10.1016/j.hest.2025.07.007","url":null,"abstract":"<div><h3>Objective</h3><div>Chimeric antigen receptor T (CAR-T) cell therapy has demonstrated significant efficacy in the treatment of hematologic malignancies, yet it can lead to severe complications. Intracranial hemorrhage (ICH), although rare, is a particularly lethal outcome. Current research on the risk factors and underlying mechanisms of ICH post-CAR-T therapy remains limited.</div></div><div><h3>Methods</h3><div>This retrospective study analyzed 10 cases of ICH among 2255 patients who underwent CAR-T therapy at Tongji Hospital in Wuhan from January 2015 to December 2024. We collected data on baseline characteristics, CAR-T treatment parameters, inflammatory markers, coagulation function, and clinical outcomes.</div></div><div><h3>Results</h3><div>The incidence of ICH was 0.44 %, with the median time to onset being 25 days following infusion. All affected patients exhibited thrombocytopenia, and 80 % experienced coagulopathy. Levels of C-reactive protein and Interleukin-6 were significantly elevated before and after treatment. The mortality rate was 90 %, with nine patients succumbing to respiratory and circulatory failure associated with ICH.</div></div><div><h3>Conclusion</h3><div>The development of ICH after CAR-T therapy is strongly linked to thrombocytopenia, coagulopathy, and the inflammatory response triggered by cytokine release syndrome (CRS). Early identification and proactive management of high-risk patients could potentially improve outcomes. Further prospective studies are necessary to confirm risk prediction models and enhance therapeutic strategies.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 212-219"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183871","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.07.001
Yan Fu , Xiao Hu , Xiao-San Wu , Lei Zhu , Meng-Qiu Zhang , Yun-He Xia , Fang-Jian Zhu , Zi-Jie Wang , Chuan-Qin Fang , Qi Li
Objective
Ultra-early hematoma growth (uHG) has been shown to predict hematoma expansion (HE) and poor outcomes in intracerebral hemorrhage (ICH) patients. Gastrointestinal bleeding, a severe ICH complication, has been found to correlate with hematoma volumes. This study analyzes the correlation between uHG and the risk of post-ICH gastrointestinal bleeding.
Method
We prospectively recruited consecutive ICH patients. uHG was identified as the hematoma volume on the baseline computed tomography (CT) scan, divided by time from onset to CT (ml/h). Univariate and multivariate logistic regression models identified risk factors for gastrointestinal bleeding. Predictive models were formulated and evaluated using receiver operating characteristic (ROC) curves.
Results
Our study involved 138 patients, and 52 (37.7 %) had gastrointestinal bleeding after ICH. Multivariate analysis revealed uHG (aOR: 1.061; 95 % CI: 1.002–1.123; p = 0.041) as an independent predictor of gastrointestinal bleeding in ICH patients. uHG also correlated significantly with 90-day outcomes (aOR: 1.109; 95 % CI: 1.012–1.215; p = 0.026). ROC analysis showed the uHG − incorporated predictive model had strong predictive power for gastrointestinal bleeding.
Conclusion
uHG was linked to higher risks of gastrointestinal bleeding and 90-day poor outcomes among primary ICH patients.
目的超早期血肿生长(uHG)已被证明可预测脑出血(ICH)患者血肿扩张(HE)和预后不良。胃肠道出血,一种严重的脑出血并发症,已被发现与血肿体积相关。本研究分析了uHG与脑出血后消化道出血风险的相关性。方法前瞻性招募连续的脑出血患者。uHG被确定为基线计算机断层扫描(CT)上的血肿体积,除以从发病到CT的时间(ml/h)。单因素和多因素logistic回归模型确定了胃肠道出血的危险因素。采用受试者工作特征(ROC)曲线建立预测模型并进行评估。结果本研究共纳入138例患者,其中52例(37.7%)发生脑出血。多因素分析显示,uHG (aOR: 1.061; 95% CI: 1.002-1.123; p = 0.041)是脑出血患者胃肠道出血的独立预测因子。uHG也与90天预后显著相关(aOR: 1.109; 95% CI: 1.012-1.215; p = 0.026)。ROC分析显示uHG -纳入的预测模型对胃肠道出血具有较强的预测能力。结论:hg与原发性脑出血患者胃肠道出血风险增高和90天预后不良相关。
{"title":"Ultra-early hematoma growth predicts gastrointestinal bleeding after spontaneous intracerebral hemorrhage","authors":"Yan Fu , Xiao Hu , Xiao-San Wu , Lei Zhu , Meng-Qiu Zhang , Yun-He Xia , Fang-Jian Zhu , Zi-Jie Wang , Chuan-Qin Fang , Qi Li","doi":"10.1016/j.hest.2025.07.001","DOIUrl":"10.1016/j.hest.2025.07.001","url":null,"abstract":"<div><h3>Objective</h3><div>Ultra-early hematoma growth (uHG) has been shown to predict hematoma expansion (HE) and poor outcomes in intracerebral hemorrhage (ICH) patients. Gastrointestinal bleeding, a severe ICH complication, has been found to correlate with hematoma volumes. This study analyzes the correlation between uHG and the risk of post-ICH gastrointestinal bleeding.</div></div><div><h3>Method</h3><div>We prospectively recruited consecutive ICH patients. uHG was identified as the hematoma volume on the baseline computed tomography (CT) scan, divided by time from onset to CT (ml/h). Univariate and multivariate logistic regression models identified risk factors for gastrointestinal bleeding. Predictive models were formulated and evaluated using receiver operating characteristic (ROC) curves.</div></div><div><h3>Results</h3><div>Our study involved 138 patients, and 52 (37.7 %) had gastrointestinal bleeding after ICH. Multivariate analysis revealed uHG (aOR: 1.061; 95 % CI: 1.002–1.123; <em>p</em> = 0.041) as an independent predictor of gastrointestinal bleeding in ICH patients. uHG also correlated significantly with 90-day outcomes (aOR: 1.109; 95 % CI: 1.012–1.215; <em>p</em> = 0.026). ROC analysis showed the uHG − incorporated predictive model had strong predictive power for gastrointestinal bleeding.</div></div><div><h3>Conclusion</h3><div>uHG was linked to higher risks of gastrointestinal bleeding and 90-day poor outcomes among primary ICH patients.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 206-211"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183870","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.08.003
Jingjing Wang , Shenhao Chen , Ziyang Zhang , Ruoru Wu , Ye Li
Objective
The management of spontaneous intracerebral hemorrhage(ICH) is a topic of debate. While craniotomy has been the traditional treatment, minimally invasive techniques like robot-guided stereotactic catheter drainage with urokinase(MISCD-UK) have emerged as alternatives. This study compares MISCD-UK with conventional craniotomy hematoma evacuation(CHE).
Methods
A retrospective review of 60 ICH patients treated at Xuanwu Hospital(2023–2025) was conducted, with 30 patients undergoing MISCD-UK and 30 undergoing CHE. Primary outcomes included operative metrics, complications, and 3-month functional outcomes(modified Rankin Scale [mRS], Barthel Index [BI]).
Results
MISCD-UK significantly reduced operative duration(58 vs. 257.5 min, P < 0.001) and blood loss (5 mL vs. 225 mL, P < 0.001). ICU and hospital stays were shorter with MISCD-UK (ICU: 8 vs. 16.5 days, P = 0.005; total hospital stay: 9.5 vs. 17 days, P = 0.012). However, 3-month functional outcomes were similar (mRS 0–3: 36.7 % MISCD-UK vs. 50 % CHE, P = 0.297). Multivariable analysis showed that admission GCS, intraventricular hemorrhage, and anticoagulant use were key predictors of outcomes, while the surgical approach didn’t impact functional recovery.
Conclusion
MISCD-UK provides perioperative advantages over craniotomy but doesn’t improve 3-month functional outcomes, suggesting early neurologic severity and hemorrhage characteristics are crucial for prognosis.
目的自发性脑出血(ICH)的治疗是一个争论不休的话题。虽然开颅术一直是传统的治疗方法,但像机器人引导的尿激酶立体定向导管引流(MISCD-UK)这样的微创技术已经成为替代方案。本研究比较了MISCD-UK与传统开颅血肿清除术(CHE)。方法回顾性分析宣武医院2023-2025年收治的脑出血患者60例,其中30例行MISCD-UK, 30例行CHE。主要结局包括手术指标、并发症和3个月功能结局(改良Rankin量表[mRS]、Barthel指数[BI])。结果miscd - uk显著缩短手术时间(58分钟vs. 257.5分钟,P < 0.001)和出血量(5 mL vs. 225 mL, P < 0.001)。MISCD-UK的ICU和住院时间较短(ICU: 8天对16.5天,P = 0.005;总住院时间:9.5天对17天,P = 0.012)。然而,3个月的功能结果相似(mRS 0-3: 36.7% MISCD-UK vs 50% CHE, P = 0.297)。多变量分析显示,入院GCS、脑室内出血和抗凝剂使用是预后的关键预测因素,而手术方式对功能恢复没有影响。结论miscd - uk术围手术期优于开颅术,但不能改善3个月的功能预后,提示早期神经系统严重程度和出血特征对预后至关重要。
{"title":"Robot-guided catheter drainage with urokinase versus craniotomy hematoma evacuation for spontaneous intracerebral hemorrhage: A retrospective study","authors":"Jingjing Wang , Shenhao Chen , Ziyang Zhang , Ruoru Wu , Ye Li","doi":"10.1016/j.hest.2025.08.003","DOIUrl":"10.1016/j.hest.2025.08.003","url":null,"abstract":"<div><h3>Objective</h3><div>The management of spontaneous intracerebral hemorrhage(ICH) is a topic of debate. While craniotomy has been the traditional treatment, minimally invasive techniques like robot-guided stereotactic catheter drainage with urokinase(MISCD-UK) have emerged as alternatives. This study compares MISCD-UK with conventional craniotomy hematoma evacuation(CHE).</div></div><div><h3>Methods</h3><div>A retrospective review of 60 ICH patients treated at Xuanwu Hospital(2023–2025) was conducted, with 30 patients undergoing MISCD-UK and 30 undergoing CHE. Primary outcomes included operative metrics, complications, and 3-month functional outcomes(modified Rankin Scale [mRS], Barthel Index [BI]).</div></div><div><h3>Results</h3><div>MISCD-UK significantly reduced operative duration(58 vs. 257.5 min, <em>P</em> < 0.001) and blood loss (5 mL vs. 225 mL, <em>P</em> < 0.001). ICU and hospital stays were shorter with MISCD-UK (ICU: 8 vs. 16.5 days, <em>P</em> = 0.005; total hospital stay: 9.5 vs. 17 days, <em>P</em> = 0.012). However, 3-month functional outcomes were similar (mRS 0–3: 36.7 % MISCD-UK vs. 50 % CHE, <em>P</em> = 0.297). Multivariable analysis showed that admission GCS, intraventricular hemorrhage, and anticoagulant use were key predictors of outcomes, while the surgical approach didn’t impact functional recovery.</div></div><div><h3>Conclusion</h3><div>MISCD-UK provides perioperative advantages over craniotomy but doesn’t improve 3-month functional outcomes, suggesting early neurologic severity and hemorrhage characteristics are crucial for prognosis.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 220-226"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183831","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}
Clazosentan, selective endothelin A receptor (ETA) antagonist, has effectively prevented vasospasm and delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage (aSAH). However, pulmonary complications are recognized adverse events, and predictive factors remain unclear.
Methods
Between April 2022 and June 2024, 66 patients with aSAH were treated at our institution. The patients who underwent obliterative intervention for ruptured aneurysm and received clazosentan were included. Patients were stratified into those who developed pulmonary complications related to clazosentan (PCrC) and those who did not (non-PCrC). Relevant demographic, laboratory, and radiographical data were analyzed.
Results
Of the 51 patients analyzed, PCrC occurred in 10 patients, and a median onset interval of 3 days post-administration. The mean age of the PCrC group was significantly higher than that of the non-PCrC group (p = 0.049). Additionally, serum sodium levels (p = 0.023), brain natriuretic peptide concentration (p = 0.010) and cardiothoracic ratio (p = 0.002) on admission were significantly elevated in the PCrC group.
Conclusion
The plausible mechanism underlying PCrC involves the selective inhibition of ETA receptors by clazosentan, and inverse activation of ETB receptors. This activation may disrupt sodium equilibrium in alveoli, contributing to PCrC. Patients with preexisting conditions such as subclinical heart failure or relative hypernatremia may be particularly susceptible to PCrC.
{"title":"Perioperative sodium overload is associated with the occurrence of pulmonary complications related to clazosentan in patients with subarachnoid hemorrhage followed by obliterative management for ruptured aneurysm: A retrospective observational study","authors":"Koki Mitani , Takeshi Miyata , Wataru Shiraishi , Yu Abekura , Yuji Agawa , Tomoya Ogawa , Takao Morita , Wataru Yoshizaki , Ryo Hamamoto , Takashi Nagahori , Yusuke Nakazawa , Yukiko Inamori , Taketo Hatano","doi":"10.1016/j.hest.2025.02.003","DOIUrl":"10.1016/j.hest.2025.02.003","url":null,"abstract":"<div><h3>Objective</h3><div>Clazosentan, selective endothelin A receptor (ET<sub>A</sub>) antagonist, has effectively prevented vasospasm and delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage (aSAH). However, pulmonary complications are recognized adverse events, and predictive factors remain unclear.</div></div><div><h3>Methods</h3><div>Between April 2022 and June 2024, 66 patients with aSAH were treated at our institution. The patients who underwent obliterative intervention for ruptured aneurysm and received clazosentan were included. Patients were stratified into those who developed pulmonary complications related to clazosentan (PCrC) and those who did not (non-PCrC). Relevant demographic, laboratory, and radiographical data were analyzed.</div></div><div><h3>Results</h3><div>Of the 51 patients analyzed, PCrC occurred in 10 patients, and a median onset interval of 3 days post-administration. The mean age of the PCrC group was significantly higher than that of the non-PCrC group (<em>p</em> = 0.049). Additionally, serum sodium levels (<em>p</em> = 0.023), brain natriuretic peptide concentration (<em>p</em> = 0.010) and cardiothoracic ratio (<em>p</em> = 0.002) on admission were significantly elevated in the PCrC group.</div></div><div><h3>Conclusion</h3><div>The plausible mechanism underlying PCrC involves the selective inhibition of ET<sub>A</sub> receptors by clazosentan, and inverse activation of ET<sub>B</sub> receptors. This activation may disrupt sodium equilibrium in alveoli, contributing to PCrC. Patients with preexisting conditions such as subclinical heart failure or relative hypernatremia may be particularly susceptible to PCrC.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 227-236"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.07.002
Selorm Adablanu , Utpal Barman , Dulumani Das
Stroke remains a leading cause of death and disability worldwide, necessitating improved diagnostic tools for early detection and classification. Machine learning (ML) techniques have shown promise in addressing this critical healthcare challenge by enabling efficient analysis of stroke-related data. However, the lack of standardized datasets, limited real-time clinical applicability, and the complexity of model interpretability hinder broader adoption. This review critically examines 34 research articles published between 2014 and 2025, focusing on traditional ML, deep learning, transfer learning, and hybrid approaches for stroke detection and classification. Key findings highlight that Traditional ML models such as Support Vector Machines (SVM) and Random Forests (RF) have been widely used but show limitations in high-dimensional medical imaging tasks. Conversely, advanced deep learning models, such as EEG-DenseNet and ResNet50, excel in stroke segmentation and classification tasks, while hybrid methods demonstrate potential for improving accuracy through ensemble strategies. The review also underscores the challenges of dataset scarcity, ethical concerns, and integration barriers in clinical settings. Recommendations for future research include developing more representative datasets, advancing explainable AI methods, and exploring real-time implementation frameworks to bridge the gap between research and clinical practice.
{"title":"Advancing deep learning for automated stroke detection: a review","authors":"Selorm Adablanu , Utpal Barman , Dulumani Das","doi":"10.1016/j.hest.2025.07.002","DOIUrl":"10.1016/j.hest.2025.07.002","url":null,"abstract":"<div><div>Stroke remains a leading cause of death and disability worldwide, necessitating improved diagnostic tools for early detection and classification. Machine learning (ML) techniques have shown promise in addressing this critical healthcare challenge by enabling efficient analysis of stroke-related data. However, the lack of standardized datasets, limited real-time clinical applicability, and the complexity of model interpretability hinder broader adoption. This review critically examines 34 research articles published between 2014 and 2025, focusing on traditional ML, deep learning, transfer learning, and hybrid approaches for stroke detection and classification. Key findings highlight that Traditional ML models such as Support Vector Machines (SVM) and Random Forests (RF) have been widely used but show limitations in high-dimensional medical imaging tasks. Conversely, advanced deep learning models, such as EEG-DenseNet and ResNet50, excel in stroke segmentation and classification tasks, while hybrid methods demonstrate potential for improving accuracy through ensemble strategies. The review also underscores the challenges of dataset scarcity, ethical concerns, and integration barriers in clinical settings. Recommendations for future research include developing more representative datasets, advancing explainable AI methods, and exploring real-time implementation frameworks to bridge the gap between research and clinical practice.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 247-260"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183908","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.01.001
Yu-tian Liu , Na Li , Chuan-yuan Tao
{"title":"Rupture of cerebral cavernous malformations due to cerebral venous sinus compression","authors":"Yu-tian Liu , Na Li , Chuan-yuan Tao","doi":"10.1016/j.hest.2025.01.001","DOIUrl":"10.1016/j.hest.2025.01.001","url":null,"abstract":"","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 265-267"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183910","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.04.003
Yuren Hu , Zengqiang Yan , Zhuo Kuang , Xianbo Deng , Li Yu
Objective
The segmentation of intracerebral hemorrhage (ICH) lesions in brain CT scans is of paramount importance for the diagnosis and treatment of stroke. Given the tremendous challenge of pixel-wise annotation in intracerebral hemorrhage, weakly supervised segmentation for ICH based on image-level labels has drawn great attention. Typical methods constructed based on convolutional neural networks often suffer from insufficient global perception, making it difficult to address ICH lesion diversity. Therefore, vision transformer, building pair-wise global dependency, becomes a popular alternative. Unfortunately, the data-hungry nature of vision transformer hinders its full exploitation given relatively limited medical imaging data, resulting in over-smoothing issue.
Methods
In this paper, based on the observation that most patches/tokens tend to build pair-wise dependency with intracerebral hemorrhage lesion, we propose weighted attention fusion (WAF) to fully utilize over-smoothing attention maps produced by ViT under conditions of limited training data. Compared to existing research, no additional parameters or computational complexity is introduced by WAF when incorporating target-relevant information. In addition, to recall low-confidence/-salient regions in segmentation, a patch-erasing re-activation mechanism is proposed by forcing the model to explore more class-specific regions.
Results
Experimental results on three datasets, i.e., INSTANCE2022, LocalBrainCT and BraTS2021 demonstrates the effectiveness of the proposed ICH weakly supervised segmentation framework. Compared to the previous works on the weakly supervised sementation, the proposed architecture obtains the state-of-the-art performance on intracerebral hemorrhage segmentation (Dice of 72.39).
Conclusion
This study focus on weakly supervised intracerebral hemorrhage segmentation, and propose a transformer-based framework with weighted attention fusion module and patch-erasing re-activation mechanism. It achives superior performance than previous methods under various settings.
{"title":"Transformer-based weakly supervised intracerebral hemorrhage segmentation using image-level labels","authors":"Yuren Hu , Zengqiang Yan , Zhuo Kuang , Xianbo Deng , Li Yu","doi":"10.1016/j.hest.2025.04.003","DOIUrl":"10.1016/j.hest.2025.04.003","url":null,"abstract":"<div><h3>Objective</h3><div>The segmentation of intracerebral hemorrhage (ICH) lesions in brain CT scans is of paramount importance for the diagnosis and treatment of stroke. Given the tremendous challenge of pixel-wise annotation in intracerebral hemorrhage, weakly supervised segmentation for ICH based on image-level labels has drawn great attention. Typical methods constructed based on convolutional neural networks often suffer from insufficient global perception, making it difficult to address ICH lesion diversity. Therefore, vision transformer, building pair-wise global dependency, becomes a popular alternative. Unfortunately, the data-hungry nature of vision transformer hinders its full exploitation given relatively limited medical imaging data, resulting in over-smoothing issue.</div></div><div><h3>Methods</h3><div>In this paper, based on the observation that most patches/tokens tend to build pair-wise dependency with intracerebral hemorrhage lesion, we propose weighted attention fusion (WAF) to fully utilize over-smoothing attention maps produced by ViT under conditions of limited training data. Compared to existing research, no additional parameters or computational complexity is introduced by WAF when incorporating target-relevant information. In addition, to recall low-confidence/-salient regions in segmentation, a patch-erasing re-activation mechanism is proposed by forcing the model to explore more class-specific regions.</div></div><div><h3>Results</h3><div>Experimental results on three datasets, <em>i.e.</em>, INSTANCE2022, LocalBrainCT and BraTS2021 demonstrates the effectiveness of the proposed ICH weakly supervised segmentation framework. Compared to the previous works on the weakly supervised sementation, the proposed architecture obtains the state-of-the-art performance on intracerebral hemorrhage segmentation (Dice of 72.39).</div></div><div><h3>Conclusion</h3><div>This study focus on weakly supervised intracerebral hemorrhage segmentation, and propose a transformer-based framework with weighted attention fusion module and patch-erasing re-activation mechanism. It achives superior performance than previous methods under various settings.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 195-205"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01DOI: 10.1016/j.hest.2025.07.005
Elarya Morcos , John H. Zhang , Antoine Sakr , Desislava Doycheva
Heart failure and stroke are two major cardiovascular conditions that significantly contribute to morbidity and mortality worldwide.1 While the link between heart failure and ischemic stroke is well understood, the relationship between heart failure and hemorrhagic stroke is less clear but gaining increased recognition.2 This review explores the complex mechanisms that may link heart failure to both ischemic and hemorrhagic strokes. Key factors include unstable blood flow, neurohormonal activation, cerebral small vessel disease, and the use of anticoagulation therapy, which, while intended to reduce stroke risk, can inadvertently increase the risk of hemorrhagic stroke.3
Both ischemic and hemorrhagic strokes share overlapping risk factors with heart failure, such as hypertension and impaired blood flow, but the nuanced ways in which heart failure contributes to hemorrhagic stroke risk are still under investigation.3 The review also highlights the importance of clinical outcomes and strategies for managing these risks in heart failure patients, particularly as our understanding of the associations between heart failure and stroke deepens.2 By recognizing these connections, healthcare providers can improve care and implement targeted therapeutic approaches to reduce stroke risk in patients with heart failure.4
{"title":"Link between heart failure and stroke: a comprehensive review","authors":"Elarya Morcos , John H. Zhang , Antoine Sakr , Desislava Doycheva","doi":"10.1016/j.hest.2025.07.005","DOIUrl":"10.1016/j.hest.2025.07.005","url":null,"abstract":"<div><div>Heart failure and stroke are two major cardiovascular conditions that significantly contribute to morbidity and mortality worldwide.<span><span><sup>1</sup></span></span> While the link between heart failure and ischemic stroke is well understood, the relationship between heart failure and hemorrhagic stroke is less clear but gaining increased recognition.<span><span><sup>2</sup></span></span> This review explores the complex mechanisms that may link heart failure to both ischemic and hemorrhagic strokes. Key factors include unstable blood flow, neurohormonal activation, cerebral small vessel disease, and the use of anticoagulation therapy, which, while intended to reduce stroke risk, can inadvertently increase the risk of hemorrhagic stroke.<span><span><sup>3</sup></span></span></div><div>Both ischemic and hemorrhagic strokes share overlapping risk factors with heart failure, such as hypertension and impaired blood flow, but the nuanced ways in which heart failure contributes to hemorrhagic stroke risk are still under investigation.<span><span><sup>3</sup></span></span> The review also highlights the importance of clinical outcomes and strategies for managing these risks in heart failure patients, particularly as our understanding of the associations between heart failure and stroke deepens.<span><span><sup>2</sup></span></span> By recognizing these connections, healthcare providers can improve care and implement targeted therapeutic approaches to reduce stroke risk in patients with heart failure.<span><span><sup>4</sup></span></span></div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 5","pages":"Pages 237-246"},"PeriodicalIF":1.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145183832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-08-01DOI: 10.1016/j.hest.2025.02.007
Suleiman Daoud , Atef F. Hulliel , Rasha S. Mustafa , Almutazballlah Qablan , Yasmeen Jamal Alabdallat , Sa’ed Hasan , Nataly Al-zubi , Adam M. Abdallah , Mohammad Alsharman , Amer Jaradat , Sultan Jarrar , Mohammad A Jamous
Objective
Traumatic brain injury (TBI) and post-traumatic seizures (PTS) have been interconnected since ancient times. Modern research emphasizes their impact on morbidity and mortality, with persistent seizures linked to poor cognitive outcomes. PTS complicates recovery, especially when early seizures indicate severe injury. This study explores how different TBI types affect the development of focal versus generalized seizures to improve clinical management.
Methods
A retrospective study at King Abdullah University Hospital (KAUH), Jordan, examined 1934 TBI cases from 2017 to 2023, identifying 118 patients with confirmed post-traumatic seizures.
Results
Among 118 patients, the median age was 21.5 years (range 2–86), and most were male (74.6 %). Trauma types included Intracerebral Hemorrhage (47.5 %), Subarachnoid Hemorrhage (38.1 %), Intraventricular Hemorrhage (17 %), Subdural Hematoma (28.8 %), Epidural Hematoma (12.7 %), and Skull Fractures (38.1 %). Generalized tonic-clonic seizures occurred in 69.5 %, while 30.5 % were focal. At study end, 84.7 % of patients survived. Binary logistic regression revealed that Subarachnoid Hemorrhage (SAH) significantly increased the likelihood of generalized seizures (odds ratio 3.308, p = 0.005). Other trauma types did not significantly affect seizure type.
Conclusion
Subarachnoid hemorrhage (SAH) significantly increases the likelihood of generalized tonic-clonic seizures, suggesting it as a key risk factor. Further research is needed to refine preventive and treatment strategies.
{"title":"The role of subarachnoid hemorrhage in post-traumatic Seizures: Insights from a retrospective study on TBI and seizure patterns","authors":"Suleiman Daoud , Atef F. Hulliel , Rasha S. Mustafa , Almutazballlah Qablan , Yasmeen Jamal Alabdallat , Sa’ed Hasan , Nataly Al-zubi , Adam M. Abdallah , Mohammad Alsharman , Amer Jaradat , Sultan Jarrar , Mohammad A Jamous","doi":"10.1016/j.hest.2025.02.007","DOIUrl":"10.1016/j.hest.2025.02.007","url":null,"abstract":"<div><h3>Objective</h3><div>Traumatic brain injury (TBI) and post-traumatic seizures (PTS) have been interconnected since ancient times. Modern research emphasizes their impact on morbidity and mortality, with persistent seizures linked to poor cognitive outcomes. PTS complicates recovery, especially when early seizures indicate severe injury. This study explores how different TBI types affect the development of focal versus generalized seizures to improve clinical management.</div></div><div><h3>Methods</h3><div>A retrospective study at King Abdullah University Hospital (KAUH), Jordan, examined 1934 TBI cases from 2017 to 2023, identifying 118 patients with confirmed post-traumatic seizures.</div></div><div><h3>Results</h3><div>Among 118 patients, the median age was 21.5 years (range 2–86), and most were male (74.6 %). Trauma types included Intracerebral Hemorrhage (47.5 %), Subarachnoid Hemorrhage (38.1 %), Intraventricular Hemorrhage (17 %), Subdural Hematoma (28.8 %), Epidural Hematoma (12.7 %), and Skull Fractures (38.1 %). Generalized tonic-clonic seizures occurred in 69.5 %, while 30.5 % were focal. At study end, 84.7 % of patients survived. Binary logistic regression revealed that Subarachnoid Hemorrhage (SAH) significantly increased the likelihood of generalized seizures (odds ratio 3.308, p = 0.005). Other trauma types did not significantly affect seizure type.</div></div><div><h3>Conclusion</h3><div>Subarachnoid hemorrhage (SAH) significantly increases the likelihood of generalized tonic-clonic seizures, suggesting it as a key risk factor. Further research is needed to refine preventive and treatment strategies.</div></div>","PeriodicalId":33969,"journal":{"name":"Brain Hemorrhages","volume":"6 4","pages":"Pages 162-165"},"PeriodicalIF":1.3,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144831237","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}