RESUMO
The quantitative analysis of pulsed-chemical exchange saturation transfer (CEST) using a full model-based method is computationally challenging, as it involves dealing with varying RF values in pulsed saturation. A power equivalent continuous approximation of B1 power was usually applied to accelerate the analysis. In line with recent consensus recommendations from the CEST community for pulsed-CEST at 3T, particularly recommending a high RF saturation power (B1 = 2.0 µT) for the clinical application in brain tumors, this technical note investigated the feasibility of using average power (AP) as the continuous approximation. The simulated results revealed excellent performance of the AP continuous approximation in low saturation power scenarios, but discrepancies were observed in the z-spectra for the high saturation power cases. Cautions should be taken, or it may lead to inaccurate fitted parameters, and the difference can be more than 10% in the high saturation power cases.
RESUMO
This project employs artificial intelligence, including machine learning and deep learning, to assess COVID-19 readmission risk in Malaysia. It offers tools to mitigate healthcare resource strain and enhance patient outcomes. This study outlines a methodology for classifying COVID-19 readmissions. It starts with dataset description and pre-processing, while the data balancing was computed through Random Oversampling, Borderline SMOTE, and Adaptive Synthetic Sampling. Nine machine learning and ten deep learning techniques are applied, with five-fold cross-validation for evaluation. Optuna is used for hyperparameter selection, while the consistency in training hyperparameters is maintained. Evaluation metrics encompass accuracy, AUC, and training/inference times. Results were based on stratified five-fold cross-validation and different data-balancing methods. Notably, CatBoost consistently excelled in accuracy and AUC across all tables. Using ROS, CatBoost achieved the highest accuracy (0.9882 ± 0.0020) with an AUC of 1.0000 ± 0.0000. CatBoost maintained its superiority in BSMOTE and ADASYN as well. Deep learning approaches performed well, with SAINT leading in ROS and TabNet leading in BSMOTE and ADASYN. Decision Tree ensembles like Random Forest and XGBoost consistently showed strong performance.
RESUMO
Respiratory rate (RR) is a critical vital sign that can provide valuable insights into various medical conditions, including pneumonia. Unfortunately, manual RR counting is often unreliable and discontinuous. Current RR estimation algorithms either lack the necessary accuracy or demand extensive window sizes. In response to these challenges, this study introduces a novel method for continuously estimating RR from photoplethysmogram (PPG) with a reduced window size and lower processing requirements. To evaluate and compare classical and deep learning algorithms, this study leverages the BIDMC and CapnoBase datasets, employing the Respiratory Rate Estimation (RRest) toolbox. The optimal classical techniques combination on the BIDMC datasets achieves a mean absolute error (MAE) of 1.9 breaths/min. Additionally, the developed neural network model utilises convolutional and long short-term memory layers to estimate RR effectively. The best-performing model, with a 50% train-test split and a window size of 7 s, achieves an MAE of 2 breaths/min. Furthermore, compared to other deep learning algorithms with window sizes of 16, 32, and 64 s, this study's model demonstrates superior performance with a smaller window size. The study suggests that further research into more precise signal processing techniques may enhance RR estimation from PPG signals.
RESUMO
Syncope is a transient loss of consciousness with rapid onset. The aims of the study were to systematically evaluate available machine learning (ML) algorithm for supporting syncope diagnosis to determine their performance compared to existing point scoring protocols. We systematically searched IEEE Xplore, Web of Science, and Elsevier for English articles (Jan 2011 - Sep 2021) on individuals aged five and above, employing ML algorithms in syncope detection with Head-up titl table test (HUTT)-monitored hemodynamic parameters and reported metrics. Extracted data encompassed subject count, age range, syncope protocols, ML type, hemodynamic parameters, and performance metrics. Of the 6301 studies initially identified, 10 studies, involving 1205 participants aged 5 to 82 years, met the inclusion criteria, and formed the basis for it. Selected studies must use ML algorithms in syncope detection with hemodynamic parameters recorded throughout HUTT. The overall ML algorithm performance achieved a sensitivity of 88.8% (95% CI: 79.4-96.1%), specificity of 81.5% (95% CI: 69.8-92.8%) and accuracy of 85.8% (95% CI: 78.6-92.8%). Machine learning improves syncope diagnosis compared to traditional scoring, requiring fewer parameters. Future enhancements with larger databases are anticipated. Integrating ML can curb needless admissions, refine diagnostics, and enhance the quality of life for syncope patients.
RESUMO
Background: When an ischemic stroke happens, it triggers a complex signalling cascade that may eventually lead to neuronal cell death if no reperfusion. Recently, the relayed nuclear Overhauser enhancement effect at -1.6 ppm [NOE(-1.6 ppm)] has been postulated may allow for a more in-depth analysis of the ischemic injury. This study assessed the potential utility of NOE(-1.6 ppm) in an ischemic stroke model. Methods: Diffusion-weighted imaging, perfusion-weighted imaging, and chemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI) data were acquired from five rats that underwent scans at 9.4 T after middle cerebral artery occlusion. Results: The apparent diffusion coefficient (ADC), cerebral blood flow (CBF), and apparent exchange-dependent relaxations (AREX) at 3.5 ppm and NOE(-1.6 ppm) were quantified. AREX(3.5 ppm) and NOE(-1.6 ppm) were found to be hypointense and exhibited different signal patterns within the ischemic tissue. The NOE(-1.6 ppm) deficit areas were equal to or larger than the ADC deficit areas, but smaller than the AREX(3.5 ppm) deficit areas. This suggested that NOE(-1.6 ppm) might further delineate the acidotic tissue estimated using AREX(3.5 ppm). Since NOE(-1.6 ppm) is closely related to membrane phospholipids, NOE(-1.6 ppm) potentially highlighted at-risk tissue affected by lipid peroxidation and membrane damage. Altogether, the ADC/NOE(-1.6 ppm)/AREX(3.5 ppm)/CBF mismatches revealed four zones of increasing sizes within the ischemic tissue, potentially reflecting different pathophysiological information. Conclusions: Using CEST coupled with ADC and CBF, the ischemic tissue may thus potentially be separated into four zones to better understand the pathophysiology after stroke and improve ischemic tissue fate definition. Further verification of the potential utility of NOE(-1.6 ppm) may therefore lead to a more precise diagnosis.
RESUMO
Debates persist regarding the impact of Stain Normalization (SN) on recent breast cancer histopathological studies. While some studies propose no influence on classification outcomes, others argue for improvement. This study aims to assess the efficacy of SN in breast cancer histopathological classification, specifically focusing on Invasive Ductal Carcinoma (IDC) grading using Convolutional Neural Networks (CNNs). The null hypothesis asserts that SN has no effect on the accuracy of CNN-based IDC grading, while the alternative hypothesis suggests the contrary. We evaluated six SN techniques, with five templates selected as target images for the conventional SN techniques. We also utilized seven ImageNet pre-trained CNNs for IDC grading. The performance of models trained with and without SN was compared to discern the influence of SN on classification outcomes. The analysis unveiled a p-value of 0.11, indicating no statistically significant difference in Balanced Accuracy Scores between models trained with StainGAN-normalized images, achieving a score of 0.9196 (the best-performing SN technique), and models trained with non-normalized images, which scored 0.9308. As a result, we did not reject the null hypothesis, indicating that we found no evidence to support a significant discrepancy in effectiveness between stain-normalized and non-normalized datasets for IDC grading tasks. This study demonstrates that SN has a limited impact on IDC grading, challenging the assumption of performance enhancement through SN.
Assuntos
Neoplasias da Mama , Carcinoma Ductal de Mama , Carcinoma Ductal , Humanos , Feminino , Neoplasias da Mama/diagnóstico por imagem , Neoplasias da Mama/patologia , Mama/patologia , Redes Neurais de Computação , Coloração e Rotulagem , Carcinoma Ductal de Mama/patologiaRESUMO
Background: Renal cancer is one of the leading causes of cancer-related deaths worldwide, and early detection of renal cancer can significantly improve the patients' survival rate. However, the manual analysis of renal tissue in the current clinical practices is labor-intensive, prone to inter-pathologist variations and easy to miss the important cancer markers, especially in the early stage. Methods: In this work, we developed deep convolutional neural network (CNN) based heterogeneous ensemble models for automated analysis of renal histopathological images without detailed annotations. The proposed method would first segment the histopathological tissue into patches with different magnification factors, then classify the generated patches into normal and tumor tissues using the pre-trained CNNs and lastly perform the deep ensemble learning to determine the final classification. The heterogeneous ensemble models consisted of CNN models from five deep learning architectures, namely VGG, ResNet, DenseNet, MobileNet, and EfficientNet. These CNN models were fine-tuned and used as base learners, they exhibited different performances and had great diversity in histopathological image analysis. The CNN models with superior classification accuracy (Acc) were then selected to undergo ensemble learning for the final classification. The performance of the investigated ensemble approaches was evaluated against the state-of-the-art literature. Results: The performance evaluation demonstrated the superiority of the proposed best performing ensembled model: five-CNN based weighted averaging model, with an Acc (99%), specificity (Sp) (98%), F1-score (F1) (99%) and area under the receiver operating characteristic (ROC) curve (98%) but slightly inferior recall (Re) (99%) compared to the literature. Conclusions: The outstanding robustness of the developed ensemble model with a superiorly high-performance scores in the evaluated metrics suggested its reliability as a diagnosis system for assisting the pathologists in analyzing the renal histopathological tissues. It is expected that the proposed ensemble deep CNN models can greatly improve the early detection of renal cancer by making the diagnosis process more efficient, and less misdetection and misdiagnosis; subsequently, leading to higher patients' survival rate.
RESUMO
Gastric cancer is a leading cause of cancer-related deaths worldwide, underscoring the need for early detection to improve patient survival rates. The current clinical gold standard for detection is histopathological image analysis, but this process is manual, laborious, and time-consuming. As a result, there has been growing interest in developing computer-aided diagnosis to assist pathologists. Deep learning has shown promise in this regard, but each model can only extract a limited number of image features for classification. To overcome this limitation and improve classification performance, this study proposes ensemble models that combine the decisions of several deep learning models. To evaluate the effectiveness of the proposed models, we tested their performance on the publicly available gastric cancer dataset, Gastric Histopathology Sub-size Image Database. Our experimental results showed that the top 5 ensemble model achieved state-of-the-art detection accuracy in all sub-databases, with the highest detection accuracy of 99.20% in the 160 × 160 pixels sub-database. These results demonstrated that ensemble models could extract important features from smaller patch sizes and achieve promising performance. Overall, our proposed work could assist pathologists in detecting gastric cancer through histopathological image analysis and contribute to early gastric cancer detection to improve patient survival rates.
RESUMO
Cardiac health diseases are one of the key causes of death around the globe. The number of heart patients has considerably increased during the pandemic. Therefore, it is crucial to assess and analyze the medical and cardiac images. Deep learning architectures, specifically convolutional neural networks have profoundly become the primary choice for the assessment of cardiac medical images. The left ventricle is a vital part of the cardiovascular system where the boundary and size perform a significant role in the evaluation of cardiac function. Due to automatic segmentation and good promising results, the left ventricle segmentation using deep learning has attracted a lot of attention. This article presents a critical review of deep learning methods used for the left ventricle segmentation from frequently used imaging modalities including magnetic resonance images, ultrasound, and computer tomography. This study also demonstrates the details of the network architecture, software, and hardware used for training along with publicly available cardiac image datasets and self-prepared dataset details incorporated. The summary of the evaluation matrices with results used by different researchers is also presented in this study. Finally, all this information is summarized and comprehended in order to assist the readers to understand the motivation and methodology of various deep learning models, as well as exploring potential solutions to future challenges in LV segmentation.
Assuntos
Aprendizado Profundo , Cardiopatias , Humanos , Ventrículos do Coração/diagnóstico por imagem , Coração , Redes Neurais de Computação , Imageamento por Ressonância Magnética , Processamento de Imagem Assistida por Computador/métodosRESUMO
Amide proton transfer (APT) imaging, a variant of chemical exchange saturation transfer MRI, has shown promise in detecting ischemic tissue acidosis following impaired aerobic metabolism in animal models and in human stroke patients due to the sensitivity of the amide proton exchange rate to changes in pH within the physiological range. Recent studies have demonstrated the possibility of using APT-MRI to detect acidosis of the ischemic penumbra, enabling the assessment of stroke severity and risk of progression, monitoring of treatment progress, and prognostication of clinical outcome. This paper reviews current APT imaging methods actively used in ischemic stroke research and explores the clinical aspects of ischemic stroke and future applications for these methods.
Assuntos
Acidose , AVC Isquêmico , Acidente Vascular Cerebral , Animais , Humanos , Prótons , Amidas , Acidente Vascular Cerebral/diagnóstico por imagem , Imageamento por Ressonância Magnética/métodosRESUMO
Computer-aided Invasive Ductal Carcinoma (IDC) grading classification systems based on deep learning have shown that deep learning may achieve reliable accuracy in IDC grade classification using histopathology images. However, there is a dearth of comprehensive performance comparisons of Convolutional Neural Network (CNN) designs on IDC in the literature. As such, we would like to conduct a comparison analysis of the performance of seven selected CNN models: EfficientNetB0, EfficientNetV2B0, EfficientNetV2B0-21k, ResNetV1-50, ResNetV2-50, MobileNetV1, and MobileNetV2 with transfer learning. To implement each pre-trained CNN architecture, we deployed the corresponded feature vector available from the TensorFlowHub, integrating it with dropout and dense layers to form a complete CNN model. Our findings indicated that the EfficientNetV2B0-21k (0.72B Floating-Point Operations and 7.1 M parameters) outperformed other CNN models in the IDC grading task. Nevertheless, we discovered that practically all selected CNN models perform well in the IDC grading task, with an average balanced accuracy of 0.936 ± 0.0189 on the cross-validation set and 0.9308 ± 0.0211on the test set.
Assuntos
Carcinoma Ductal , Redes Neurais de Computação , Humanos , Publicações , Aprendizado de MáquinaRESUMO
Spine surgeries impose risk to the spine's surrounding anatomical and physiological structures especially the spinal cord and the nerve roots. Intraoperative neuromonitoring (IONM) is a technology developed to monitor the integrity of the spinal cord and the nerve roots via the surgery. Transcranial motor evoked potential (TcMEP) (one of the IONM modalities) is adopted to monitor the integrity of the motor pathway of the spinal cord and the motor nerve roots. Recent research suggested that the IONM is conducive as a prognostic tool towards the patient's functional outcome. This paper summarizes the researches of IONM being adopted as a prognostic tool. In addition, this paper highlights the problems associated with the signal parameters as the improvement criteria in the previous researches. Lastly, we review the challenges of TcMEP to achieve a prognostic tool focusing on the factors that could interfere with the generation of a stable TcMEP response. The final section will discuss recommendations for IONM technology to achieve an objective prognostic tool.
Assuntos
Potencial Evocado Motor , Monitorização Neurofisiológica Intraoperatória , Humanos , Procedimentos Neurocirúrgicos , Medula Espinal , Coluna Vertebral/cirurgiaRESUMO
Osteoarthritis (OA), especially knee OA, is the most common form of arthritis, causing significant disability in patients worldwide. Manual diagnosis, segmentation, and annotations of knee joints remain as the popular method to diagnose OA in clinical practices, although they are tedious and greatly subject to user variation. Therefore, to overcome the limitations of the commonly used method as above, numerous deep learning approaches, especially the convolutional neural network (CNN), have been developed to improve the clinical workflow efficiency. Medical imaging processes, especially those that produce 3-dimensional (3D) images such as MRI, possess ability to reveal hidden structures in a volumetric view. Acknowledging that changes in a knee joint is a 3D complexity, 3D CNN has been employed to analyse the joint problem for a more accurate diagnosis in the recent years. In this review, we provide a broad overview on the current 2D and 3D CNN approaches in the OA research field. We reviewed 74 studies related to classification and segmentation of knee osteoarthritis from the Web of Science database and discussed the various state-of-the-art deep learning approaches proposed. We highlighted the potential and possibility of 3D CNN in the knee osteoarthritis field. We concluded by discussing the possible challenges faced as well as the potential advancements in adopting 3D CNNs in this field.
Assuntos
Aprendizado Profundo , Osteoartrite do Joelho , Humanos , Imageamento por Ressonância Magnética , Redes Neurais de Computação , Osteoartrite do Joelho/diagnóstico por imagemRESUMO
Hospital readmission shortly after discharge threatens the quality of patient care and leads to increased medical care costs. In the United States, hospitals with high readmission rates are subject to federal financial penalties. This concern calls for incentives for healthcare facilities to reduce their readmission rates by predicting patients who are at high risk of readmission. Conventional practices involve the use of rule-based assessment scores and traditional statistical methods, such as logistic regression, in developing risk prediction models. The recent advancements in machine learning driven by improved computing power and sophisticated algorithms have the potential to produce highly accurate predictions. However, the value of such models could be overrated. Meanwhile, the use of other flexible models that leverage simple algorithms offer great transparency in terms of feature interpretation, which is beneficial in clinical settings. This work presents an overview of the current trends in risk prediction models developed in the field of readmission. The various techniques adopted by researchers in recent years are described, and the topic of whether complex models outperform simple ones in readmission risk stratification is investigated.
RESUMO
Amide proton transfer (APT) magnetic resonance imaging (MRI) is a pH-sensitive imaging technique that can potentially complement existing clinical imaging protocol for the assessment of ischemic stroke. This review aims to summarize the developments in the clinical research of APT imaging of ischemic stroke after 17 years of progress since its first preclinical study in 2003. Three electronic databases: PubMed, Scopus, and Cochrane Library were systematically searched for articles reporting clinical studies on APT imaging of ischemic stroke. Only articles in English published between 2003 to 2020 that involved patients presenting ischemic stroke-like symptoms that underwent APT MRI were included. Of 1,093 articles screened, 14 articles met the inclusion criteria with a total of 282 patients that had been scanned using APT imaging. Generally, the clinical studies agreed APT effect to be hypointense in ischemic tissue compared to healthy tissue, allowing for the detection of ischemic stroke. Other uses of APT imaging have also been investigated in the studies, including penumbra identification, predicting long term clinical outcome, and serving as a biomarker for supportive treatment monitoring. The published results demonstrated the potential of APT imaging in these applications, but further investigations and larger trials are needed for conclusive evidence. Future studies are recommended to report the result of asymmetry analysis at 3.5 ppm along with the findings of the study to reduce this contribution to the heterogeneity of experimental methods observed and to facilitate effective comparison of results between studies and centers. In addition, it is important to focus on the development of fast 3D imaging for full volumetric ischemic tissue assessment for clinical translation.
RESUMO
The study of stroke has historically made use of traditional spectroscopy techniques to provide the ground truth for parameters like pH. However, techniques like 31P spectroscopy have limitations, in particular poor temporal and spatial resolution, coupled with a need for a high field strength and specialized coils. More modern magnetic resonance spectroscopy (MRS)-based imaging techniques like chemical exchange saturation transfer (CEST) have been developed to counter some of these limitations but lack the definitive gold standard for pH that 31P spectroscopy provides. In this perspective, both the traditional (31P spectroscopy) and emerging (CEST) techniques in the measurement of pH for ischemic imaging will be discussed. Although each has its own advantages and limitations, it is likely that CEST may be preferable simply due to the hardware, acquisition time and image resolution advantages. However, more experiments on CEST are needed to determine the specificity of endogenous CEST to absolute pH, and 31P MRS can be used to calibrate CEST for pH measurement in the preclinical model to enhance our understanding of the relationship between CEST and pH. Combining the two imaging techniques, one old and one new, we may be able to obtain new insights into stroke physiology that would not be possible otherwise with either alone.
RESUMO
PURPOSE: To assess the correlation and differences between common amide proton transfer (APT) quantification methods in the diagnosis of ischemic stroke. METHODS: Five APT quantification methods, including asymmetry analysis and its variants as well as two Lorentzian model-based methods, were applied to data acquired from six rats that underwent middle cerebral artery occlusion scanned at 9.4T. Diffusion and perfusion-weighted images, and water relaxation time maps were also acquired to study the relationship of these conventional imaging modalities with the different APT quantification methods. RESULTS: The APT ischemic area estimates had varying sizes (Jaccard index: 0.544 ≤ J ≤ 0.971) and had varying correlations in their distributions (Pearson correlation coefficient: 0.104 ≤ r ≤ 0.995), revealing discrepancies in the quantified ischemic areas. The Lorentzian methods produced the highest contrast-to-noise ratios (CNRs; 1.427 ≤ CNR ≤ 2.002), but generated APT ischemic areas that were comparable in size to the cerebral blood flow (CBF) deficit areas; asymmetry analysis and its variants produced APT ischemic areas that were smaller than the CBF deficit areas but larger than the apparent diffusion coefficient deficit areas, though having lower CNRs (0.561 ≤ CNR ≤ 1.083). CONCLUSION: There is a need to further investigate the accuracy and correlation of each quantification method with the pathophysiology using a larger scale multi-imaging modality and multi-time-point clinical study. Future studies should include the magnetization transfer ratio asymmetry results alongside the findings of the study to facilitate the comparison of results between different centers and also the published literature.
Assuntos
Isquemia Encefálica , Neoplasias Encefálicas , AVC Isquêmico , Acidente Vascular Cerebral , Amidas , Animais , Isquemia Encefálica/diagnóstico por imagem , Imageamento por Ressonância Magnética , Prótons , Ratos , Acidente Vascular Cerebral/diagnóstico por imagemRESUMO
Chemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI) holds great potential to provide new metabolic information for clinical applications such as tumor, stroke and Parkinson's Disease diagnosis. Many active research and developments have been conducted to translate this emerging MRI technique for routine clinical applications. In general, there are two CEST quantification techniques: (i) model-free and (ii) model-based techniques. The reliability of these quantification techniques depends heavily on the experimental conditions and quality of the collected data. Errors such as noise may lead to misleading quantification results and thus inaccurate diagnosis when CEST imaging becomes a standard or routine imaging scan in the future. This paper investigates the accuracy and robustness of these quantification techniques under different signal-to-noise (SNR) levels and magnetic field strengths. The quantified CEST effect before and after adding random Gaussian White Noise using model-free and model-based quantification techniques were compared. It was found that the model-free technique consistently yielded larger average percentage error across all tested parameters compared to its model-based counterpart, and that the model-based technique could withstand SNR of about 3 times lower than the model-free technique. When applied on noisy brain tumor, ischemic stroke, and Parkinson's Disease clinical data, the model-free technique failed to produce significant differences between normal and abnormal tissue whereas the model-based technique consistently generated significant differences. Although the model-free technique was less accurate and robust, its simplicity and thus speed would still make it a good approximate when the SNR was high (>50) or when the CEST effect was large and well-defined. For more accurate CEST quantification, model-based techniques should be considered. When SNR was low (<50) and the CEST effect was small such as those acquired from clinical field strength scanners, which are generally 3T and below, model-based techniques should be considered over model-free counterpart to maintain an average percentage error of less than 44% even under very noisy condition as tested in this work.
Assuntos
Imageamento por Ressonância Magnética/métodos , Patologia Molecular/instrumentação , Algoritmos , Neoplasias Encefálicas/diagnóstico , Simulação por Computador , Campos Eletromagnéticos , Humanos , Interpretação de Imagem Assistida por Computador , AVC Isquêmico/diagnóstico , Modelos Químicos , Doença de Parkinson/diagnóstico , Reprodutibilidade dos Testes , Razão Sinal-RuídoRESUMO
Document classification aims to assign one or more classes to a document for ease of management by understanding the content of a document. Hierarchical attention network (HAN) has been showed effective to classify documents that are ambiguous. HAN parses information-intense documents into slices (i.e., words and sentences) such that each slice can be learned separately and in parallel before assigning the classes. However, introducing hierarchical attention approach leads to the redundancy of training parameters which is prone to overfitting. To mitigate the concern of overfitting, we propose a variant of hierarchical attention network using adversarial and virtual adversarial perturbations in 1) word representation, 2) sentence representation and 3) both word and sentence representations. The proposed variant is tested on eight publicly available datasets. The results show that the proposed variant outperforms the hierarchical attention network with and without using random perturbation. More importantly, the proposed variant achieves state-of-the-art performance on multiple benchmark datasets. Visualizations and analysis are provided to show that perturbation can effectively alleviate the overfitting issue and improve the performance of hierarchical attention network.
Assuntos
Aprendizado de Máquina , Redes Neurais de Computação , Algoritmos , Humanos , SoftwareRESUMO
The original concept of the ischaemic penumbra suggested imaging of regional cerebral blood flow and metabolism would be required to identify tissue that may benefit from intervention. Amide proton transfer magnetic resonance imaging, a chemical exchange saturation transfer technique, has been used to derive cerebral intracellular pH in preclinical stroke models and has been proposed as a metabolic marker of ischaemic penumbra. In this proof of principle clinical study, we explored the potential of this pH-weighted magnetic resonance imaging technique at tissue-level. Detailed voxel-wise analysis was performed on data from a prospective cohort of 12 patients with acute ischaemic stroke. Voxels within ischaemic core had a more severe intracellular acidosis than hypoperfused tissue recruited to the final infarct (P < 0.0001), which in turn was more acidotic than hypoperfused tissue that survived (P < 0.0001). In addition, when confined to the grey matter perfusion deficit, intracellular pH (P < 0.0001), but not cerebral blood flow (P = 0.31), differed between tissue that infarcted and tissue that survived. Within the presenting apparent diffusion coefficient lesion, intracellular pH differed between tissue with early apparent diffusion lesion pseudonormalization and tissue with true radiographic recovery. These findings support the need for further investigation of pH-weighted imaging in patients with acute ischaemic stroke.