| 1. |
- Ahmed, Ammar, et al.
(författare)
-
Learning from the few : Fine-grained approach to pediatric wrist pathology recognition on a limited dataset
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 181
-
Tidskriftsartikel (refereegranskat)abstract
- Wrist pathologies, particularly fractures common among children and adolescents, present a critical diagnostic challenge. While X-ray imaging remains a prevalent diagnostic tool, the increasing misinterpretation rates highlight the need for more accurate analysis, especially considering the lack of specialized training among many surgeons and physicians. Recent advancements in deep convolutional neural networks offer promise in automating pathology detection in trauma X-rays. However, distinguishing subtle variations between pediatric wrist pathologies in X-rays remains challenging. Traditional manual annotation, though effective, is laborious, costly, and requires specialized expertise. In this paper, we address the challenge of pediatric wrist pathology recognition with a fine-grained approach, aimed at automatically identifying discriminative regions in X-rays without manual intervention. We refine our fine-grained architecture through ablation analysis and the integration of LION. Leveraging Grad-CAM, an explainable AI technique, we highlight these regions. Despite using limited data, reflective of real-world medical study constraints, our method consistently outperforms state-of-the-art image recognition models on both augmented and original (challenging) test sets. Our proposed refined architecture achieves an increase in accuracy of 1.06% and 1.25% compared to the baseline method, resulting in accuracies of 86% and 84%, respectively. Moreover, our approach demonstrates the highest fracture sensitivity of 97%, highlighting its potential to enhance wrist pathology recognition.
|
|
| 2. |
- Ali, Subhan, et al.
(författare)
-
The enlightening role of explainable artificial intelligence in medical & healthcare domains : A systematic literature review
- 2023
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 166
-
Tidskriftsartikel (refereegranskat)abstract
- In domains such as medical and healthcare, the interpretability and explainability of machine learning and artificial intelligence systems are crucial for building trust in their results. Errors caused by these systems, such as incorrect diagnoses or treatments, can have severe and even life-threatening consequences for patients. To address this issue, Explainable Artificial Intelligence (XAI) has emerged as a popular area of research, focused on understanding the black-box nature of complex and hard-to-interpret machine learning models. While humans can increase the accuracy of these models through technical expertise, understanding how these models actually function during training can be difficult or even impossible. XAI algorithms such as Local Interpretable Model-Agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP) can provide explanations for these models, improving trust in their predictions by providing feature importance and increasing confidence in the systems. Many articles have been published that propose solutions to medical problems by using machine learning models alongside XAI algorithms to provide interpretability and explainability. In our study, we identified 454 articles published from 2018- 2022 and analyzed 93 of them to explore the use of these techniques in the medical domain.
|
|
| 3. |
|
|
| 4. |
|
|
| 5. |
- Buddenkotte, Thomas, et al.
(författare)
-
Calibrating ensembles for scalable uncertainty quantification in deep learning-based medical image segmentation
- 2023
-
Ingår i: Computers in Biology and Medicine. - : Elsevier Ltd. - 0010-4825 .- 1879-0534. ; 163
-
Tidskriftsartikel (refereegranskat)abstract
- Uncertainty quantification in automated image analysis is highly desired in many applications. Typically, machine learning models in classification or segmentation are only developed to provide binary answers; however, quantifying the uncertainty of the models can play a critical role for example in active learning or machine human interaction. Uncertainty quantification is especially difficult when using deep learning-based models, which are the state-of-the-art in many imaging applications. The current uncertainty quantification approaches do not scale well in high-dimensional real-world problems. Scalable solutions often rely on classical techniques, such as dropout, during inference or training ensembles of identical models with different random seeds to obtain a posterior distribution. In this paper, we present the following contributions. First, we show that the classical approaches fail to approximate the classification probability. Second, we propose a scalable and intuitive framework for uncertainty quantification in medical image segmentation that yields measurements that approximate the classification probability. Third, we suggest the usage of k-fold cross-validation to overcome the need for held out calibration data. Lastly, we motivate the adoption of our method in active learning, creating pseudo-labels to learn from unlabeled images and human–machine collaboration.
|
|
| 6. |
- Chelebian, Eduard, et al.
(författare)
-
DEPICTER : Deep representation clustering for histology annotation
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 170
-
Tidskriftsartikel (refereegranskat)abstract
- Automatic segmentation of histopathology whole -slide images (WSI) usually involves supervised training of deep learning models with pixel -level labels to classify each pixel of the WSI into tissue regions such as benign or cancerous. However, fully supervised segmentation requires large-scale data manually annotated by experts, which can be expensive and time-consuming to obtain. Non -fully supervised methods, ranging from semi -supervised to unsupervised, have been proposed to address this issue and have been successful in WSI segmentation tasks. But these methods have mainly been focused on technical advancements in algorithmic performance rather than on the development of practical tools that could be used by pathologists or researchers in real -world scenarios. In contrast, we present DEPICTER (Deep rEPresentatIon ClusTERing), an interactive segmentation tool for histopathology annotation that produces a patch -wise dense segmentation map at WSI level. The interactive nature of DEPICTER leverages self- and semi -supervised learning approaches to allow the user to participate in the segmentation producing reliable results while reducing the workload. DEPICTER consists of three steps: first, a pretrained model is used to compute embeddings from image patches. Next, the user selects a number of benign and cancerous patches from the multi -resolution image. Finally, guided by the deep representations, label propagation is achieved using our novel seeded iterative clustering method or by directly interacting with the embedding space via feature space gating. We report both real-time interaction results with three pathologists and evaluate the performance on three public cancer classification dataset benchmarks through simulations. The code and demos of DEPICTER are publicly available at https://github.com/eduardchelebian/depicter.
|
|
| 7. |
- De Silva, Kushan, et al.
(författare)
-
Clinical notes as prognostic markers of mortality associated with diabetes mellitus following critical care : A retrospective cohort analysis using machine learning and unstructured big data
- 2021
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 132
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Clinical notes are ubiquitous resources offering potential value in optimizing critical care via data mining technologies. Objective: To determine the predictive value of clinical notes as prognostic markers of 1-year all-cause mortality among people with diabetes following critical care. Materials and methods: Mortality of diabetes patients were predicted using three cohorts of clinical text in a critical care database, written by physicians (n = 45253), nurses (159027), and both (n = 204280). Natural language processing was used to pre-process text documents and LASSO-regularized logistic regression models were trained and tested. Confusion matrix metrics of each model were calculated and AUROC estimates between models were compared. All predictive words and corresponding coefficients were extracted. Outcome probability associated with each text document was estimated. Results: Models built on clinical text of physicians, nurses, and the combined cohort predicted mortality with AUROC of 0.996, 0.893, and 0.922, respectively. Predictive performance of the models significantly differed from one another whereas inter-rater reliability ranged from substantial to almost perfect across them. Number of predictive words with non-zero coefficients were 3994, 8159, and 10579, respectively, in the models of physicians, nurses, and the combined cohort. Physicians & rsquo; and nursing notes, both individually and when combined, strongly predicted 1-year all-cause mortality among people with diabetes following critical care. Conclusion: Clinical notes of physicians and nurses are strong and novel prognostic markers of diabetes-associated mortality in critical care, offering potentially generalizable and scalable applications. Clinical text-derived personalized risk estimates of prognostic outcomes such as mortality could be used to optimize patient care.
|
|
| 8. |
- Faryna, Khrystyna, et al.
(författare)
-
Automatic data augmentation to improve generalization of deep learning in H&E stained histopathology
- 2024
-
Ingår i: Computers in Biology and Medicine. - : PERGAMON-ELSEVIER SCIENCE LTD. - 0010-4825 .- 1879-0534. ; 170
-
Tidskriftsartikel (refereegranskat)abstract
- In histopathology practice, scanners, tissue processing, staining, and image acquisition protocols vary from center to center, resulting in subtle variations in images. Vanilla convolutional neural networks are sensitive to such domain shifts. Data augmentation is a popular way to improve domain generalization. Currently, state-of-the-art domain generalization in computational pathology is achieved using a manually curated set of augmentation transforms. However, manual tuning of augmentation parameters is time-consuming and can lead to sub-optimal generalization performance. Meta-learning frameworks can provide efficient ways to find optimal training hyper-parameters, including data augmentation. In this study, we hypothesize that an automated search of augmentation hyper-parameters can provide superior generalization performance and reduce experimental optimization time. We select four state-of-theart automatic augmentation methods from general computer vision and investigate their capacity to improve domain generalization in histopathology. We analyze their performance on data from 25 centers across two different tasks: tumor metastasis detection in lymph nodes and breast cancer tissue type classification. On tumor metastasis detection, most automatic augmentation methods achieve comparable performance to state-of-the-art manual augmentation. On breast cancer tissue type classification, the leading automatic augmentation method significantly outperforms state-of-the-art manual data augmentation.
|
|
| 9. |
- Guarrasi, Valerio, et al.
(författare)
-
Multi-objective optimization determines when, which and how to fuse deep networks : an application to predict COVID-19 outcomes
- 2023
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 154
-
Tidskriftsartikel (refereegranskat)abstract
- The COVID-19 pandemic has caused millions of cases and deaths and the AI-related scientific community, after being involved with detecting COVID-19 signs in medical images, has been now directing the efforts towards the development of methods that can predict the progression of the disease. This task is multimodal by its very nature and, recently, baseline results achieved on the publicly available AIforCOVID dataset have shown that chest X-ray scans and clinical information are useful to identify patients at risk of severe outcomes. While deep learning has shown superior performance in several medical fields, in most of the cases it considers unimodal data only. In this respect, when, which and how to fuse the different modalities is an open challenge in multimodal deep learning. To cope with these three questions here we present a novel approach optimizing the setup of a multimodal end-to-end model. It exploits Pareto multi-objective optimization working with a performance metric and the diversity score of multiple candidate unimodal neural networks to be fused. We test our method on the AIforCOVID dataset, attaining state-of-the-art results, not only outperforming the baseline performance but also being robust to external validation. Moreover, exploiting XAI algorithms we figure out a hierarchy among the modalities and we extract the features’ intra-modality importance, enriching the trust on the predictions made by the model.
|
|
| 10. |
- Guo, Xiaoyi, et al.
(författare)
-
Random Fourier features-based sparse representation classifier for identifying DNA-binding proteins
- 2022
-
Ingår i: Computers in Biology and Medicine. - London : Elsevier. - 0010-4825 .- 1879-0534. ; 151
-
Tidskriftsartikel (refereegranskat)abstract
- DNA-binding proteins (DBPs) protect DNA from nuclease hydrolysis, inhibit the action of RNA polymerase,prevents replication and transcription from occurring simultaneously on a piece of DNA. Most of theconventional methods for detecting DBPs are biochemical methods, but the time cost is high. In recent years,a variety of machine learning-based methods that have been used on a large scale for large-scale screeningof DBPs. To improve the prediction performance of DBPs, we propose a random Fourier features-based sparserepresentation classifier (RFF-SRC), which randomly map the features into a high-dimensional space to solvenonlinear classification problems. And ?2,1-matrix norm is introduced to get sparse solution of model. Toevaluate performance, our model is tested on several benchmark data sets of DBPs and 8 UCI data sets. RFF-SRCachieves better performance in experimental results. © 2022 Elsevier Ltd.
|
|
| 11. |
- Hagberg, Eva, et al.
(författare)
-
Semi-supervised learning with natural language processing for right ventricle classification in echocardiography—a scalable approach
- 2022
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 143
-
Tidskriftsartikel (refereegranskat)abstract
- We created a deep learning model, trained on text classified by natural language processing (NLP), to assess right ventricular (RV) size and function from echocardiographic images. We included 12,684 examinations with corresponding written reports for text classification. After manual annotation of 1489 reports, we trained an NLP model to classify the remaining 10,651 reports. A view classifier was developed to select the 4-chamber or RV-focused view from an echocardiographic examination (n = 539). The final models were two image classification models trained on the predicted labels from the combined manual annotation and NLP models and the corresponding echocardiographic view to assess RV function (training set n = 11,008) and size (training set n = 9951. The text classifier identified impaired RV function with 99% sensitivity and 98% specificity and RV enlargement with 98% sensitivity and 98% specificity. The view classification model identified the 4-chamber view with 92% accuracy and the RV-focused view with 73% accuracy. The image classification models identified impaired RV function with 93% sensitivity and 72% specificity and an enlarged RV with 80% sensitivity and 85% specificity; agreement with the written reports was substantial (both κ = 0.65). Our findings show that models for automatic image assessment can be trained to classify RV size and function by using model-annotated data from written echocardiography reports. This pipeline for auto-annotation of the echocardiographic images, using a NLP model with medical reports as input, can be used to train an image-assessment model without manual annotation of images and enables fast and inexpensive expansion of the training dataset when needed. © 2022
|
|
| 12. |
- Hassan, Sk Sarif, et al.
(författare)
-
A unique view of SARS-CoV-2 through the lens of ORF8 protein
- 2021
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 133
-
Tidskriftsartikel (refereegranskat)abstract
- Immune evasion is one of the unique characteristics of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) attributed to its ORF8 protein. This protein modulates the adaptive host immunity through down regulation of MHC-1 (Major Histocompatibility Complex) molecules and innate immune responses by surpassing the host's interferon-mediated antiviral response. To understand the host's immune perspective in reference to the ORF8 protein, a comprehensive study of the ORF8 protein and mutations possessed by it have been performed. Chemical and structural properties of ORF8 proteins from different hosts, such as human, bat, and pangolin, suggest that the ORF8 of SARS-CoV-2 is much closer to ORF8 of Bat RaTG13-CoV than to that of Pangolin-CoV. Eighty-seven mutations across unique variants of ORF8 in SARS-CoV-2 can be grouped into four classes based on their predicted effects (Hussain et al., 2021) [1]. Based on the geo-locations and timescale of sample collection, a possible flow of mutations was built. Furthermore, conclusive flows of amalgamation of mutations were found upon sequence similarity analyses and consideration of the amino acid conservation phylogenies. Therefore, this study seeks to highlight the uniqueness of the rapidly evolving SARS-CoV-2 through the ORF8.
|
|
| 13. |
- Ibrahim, Mahmoud A. A., et al.
(författare)
-
In silico drug discovery of major metabolites from spices as SARS-CoV-2 main protease inhibitors
- 2020
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 126
-
Tidskriftsartikel (refereegranskat)abstract
- Coronavirus Disease 2019 (COVID-19) is an infectious illness caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), originally identified in Wuhan, China (December 2019) and has since expanded into a pandemic. Here, we investigate metabolites present in several common spices as possible inhibitors of COVID-19. Specifically, 32 compounds isolated from 14 cooking seasonings were examined as inhibitors for SARS-CoV-2 main protease (MPrn), which is required for viral multiplication. Using a drug discovery approach to identify possible antiviral leads, in silico molecular docking studies were performed. Docking calculations revealed a high potency of salvianolic acid A and curcumin as MPr inhibitors with binding energies of 9.7 and 9.2 kcal/mol, respectively. Binding mode analysis demonstrated the ability of salvianolic acid A and curcumin to form nine and six hydrogen bonds, respectively with amino acids proximal to MPr 's active site. Stabilities and binding affinities of the two identified natural spices were calculated over 40 ns molecular dynamics simulations and compared to an antiviral protease inhibitor (lopinavir). Molecular mechanics-generalized Born surface area energy calculations revealed greater salvianolic acid A affinity for the enzyme over curcumin and lopinavir with energies of 44.8, 34.2 and 34.8 kcal/mol, respectively. Using a STRING database, protein-protein interactions were identified for salvianolic acid A included the biochemical signaling genes ACE, MAPK14 and ESR1; and for curcumin, EGFR and TNF. This study establishes salvianolic acid A as an in silico natural product inhibitor against the SARS-CoV-2 main protease and provides a promising inhibitor lead for in vitro enzyme testing.
|
|
| 14. |
- Jiang, Linfeng, et al.
(författare)
-
RMAU-Net : Residual Multi-Scale Attention U-Net For liver and tumor segmentation in CT images
- 2023
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 158
-
Tidskriftsartikel (refereegranskat)abstract
- Liver cancer is one of the leading causes of cancer-related deaths worldwide. Automatic liver and tumor segmentation are of great value in clinical practice as they can reduce surgeons' workload and increase the probability of success in surgery. Liver and tumor segmentation is a challenging task because of the different sizes, shapes, blurred boundaries of livers and lesions, and low-intensity contrast between organs within patients. To address the problem of fuzzy livers and small tumors, we propose a novel Residual Multi-scale Attention U-Net (RMAU-Net) for liver and tumor segmentation by introducing two modules, i.e., Res-SE-Block and MAB. The Res-SE-Block can mitigate the problem of gradient disappearance by residual connection and enhance the quality of representations by explicitly modeling the interdependencies and feature recalibration between the channels of features. The MAB can exploit rich multi-scale feature information and capture inter -channel and inter-spatial relationships of features simultaneously. In addition, a hybrid loss function, that combines focal loss and dice loss, is designed to improve segmentation accuracy and speed up convergence. We evaluated the proposed method on two publicly available datasets, i.e., LiTS and 3D-IRCADb. Our proposed method achieved better performance than the other state-of-the-art methods, with dice scores of 0.9552 and 0.9697 for LiTS and 3D-IRCABb liver segmentation, and dice scores of 0.7616 and 0.8307 for LiTS and 3D-IRCABb liver tumor segmentation.
|
|
| 15. |
|
|
| 16. |
- Lindström, H. Jonathan G., 1992-, et al.
(författare)
-
Interplay of mutations, alternate mechanisms, and treatment breaks in leukaemia : Understanding and implications studied with stochastic models
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 169
-
Tidskriftsartikel (refereegranskat)abstract
- Bcr-Abl1 kinase domain mutations are the most prevalent cause of treatment resistance in chronic myeloid leukaemia (CML). Alternate resistance pathways nevertheless exist, and cell line experiments show certain patterns in the gain, and loss, of some of these alternate adaptations. These adaptations have clinical consequences when the tumour develops mechanisms that are beneficial to its growth under treatment, but slow down its growth when not treated. The results of temporarily halting treatment in CML have not been widely discussed in the clinic and there is no robust theoretical model that could suggest when such a pause in therapy can be tolerated. We constructed a dynamic model of how mechanisms such as Bcr-Abl1 overexpression and drug transporter upregulation evolve to produce resistance in cell lines, and investigate its behaviour subject to different treatment schedules, in particular when the treatment is paused ('drug holiday'). Our study results suggest that the presence of additional resistance mechanisms creates an environment which favours mutations that are either preexisting or occur late during treatment. Importantly, the results suggest the existence of tumour drug addiction, where cancer cells become dependent on the drug for (optimal) survival, which could be exploited through a treatment holiday. All simulation code is available at https://github.com/Sandalmoth/dual-adaptation.
|
|
| 17. |
- Liu, Qingshan, et al.
(författare)
-
Health warning based on 3R ECG Sample's combined features and LSTM
- 2023
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 162
-
Tidskriftsartikel (refereegranskat)abstract
- Most researches use the fixed-length sample to identify ECG abnormalities based on MIT ECG dataset, which leads to information loss. To address this problem, this paper proposes a method for ECG abnormality detection and health warning based on ECG Holter of PHIA and 3R-TSH-L method. The 3R-TSH-L method is implemented by:(1) getting 3R ECG samples using Pan-Tompkins method and using volatility to obtain high-quality raw ECG data; (2) extracting combination features including time-domain features, frequency domain features and time-frequency domain features; (3) using LSTM for classification, training and testing the algorithm based on the MIT-BIH dataset, and obtaining relatively optimal features as spliced normalized fusion features including kurtosis, skewness and RR interval time domain features, STFT-based sub-band spectrum features, and harmonic ratio features. The ECG data were collected using the self-developed ECG Holter (PHIA) on 14 subjects, aged between 24 and 75 including both male and female, to build the ECG dataset (ECG-H). The algorithm was transferred to the ECG-H dataset, and a health warning assessment model based on abnormal ECG rate and heart rate variability weighting was proposed. Experiments show that 3R-TSH-L method proposed in the paper has a high accuracy of 98.28% for the detection of ECG abnormalities of MIT-BIH dataset and a good transfer learning ability of 95.66% accuracy for ECG-H. The health warning model was also testified to be reasonable. The key technique of the ECG Holter of PHIA and the method 3R-TSH-L proposed in this paper is expected to be widely used in family-oriented healthcare.
|
|
| 18. |
- Malik, Shairyar, et al.
(författare)
-
A novel hybrid meta-heuristic contrast stretching technique for improved skin lesion segmentation
- 2022
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 151
-
Tidskriftsartikel (refereegranskat)abstract
- The high precedence of epidemiological examination of skin lesions necessitated the well-performing efficient classification and segmentation models. In the past two decades, various algorithms, especially machine/deep learning-based methods, replicated the classical visual examination to accomplish the above-mentioned tasks. These automated streams of models demand evident lesions with less background and noise affecting the region of interest. However, even after the proposal of these advanced techniques, there are gaps in achieving the efficacy of matter. Recently, many preprocessors proposed to enhance the contrast of lesions, which further aided the skin lesion segmentation and classification tasks. Metaheuristics are the methods used to support the search space optimisation problems. We propose a novel Hybrid Metaheuristic Differential Evolution-Bat Algorithm (DE-BA), which estimates parameters used in the brightness preserving contrast stretching transformation function. For extensive experimentation we tested our proposed algorithm on various publicly available databases like ISIC 2016, 2017, 2018 and PH2, and validated the proposed model with some state-of-the-art already existing segmentation models. The tabular and visual comparison of the results concluded that DE-BA as a preprocessor positively enhances the segmentation results.
|
|
| 19. |
|
|
| 20. |
- Nyholm, Joel, et al.
(författare)
-
Prediction of dementia based on older adults’ sleep disturbances using machine learning
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 171
-
Tidskriftsartikel (refereegranskat)abstract
- Background: The most common degenerative condition in older adults is dementia, which can be predicted using a number of indicators and whose progression can be slowed down. One of the indicators of an increased risk of dementia is sleep disturbances. This study aims to examine if machine learning can predict dementia and which sleep disturbance factors impact dementia.Methods: This study uses five machine learning algorithms (gradient boosting, logistic regression, gaussian naive Bayes, random forest and support vector machine) and data on the older population (60+) in Sweden from the Swedish National Study on Ageing and Care — Blekinge (). Each algorithm uses 10-fold stratified cross-validation to obtain the results, which consist of the Brier score for checking accuracy and the feature importance for examining the factors which impact dementia. The algorithms use 16 features which are on personal and sleep disturbance factors.Results: Logistic regression found an association between dementia and sleep disturbances. However, it is slight for the features in the study. Gradient boosting was the most accurate algorithm with 92.9% accuracy, 0.926 f1-score, 0.974 ROC AUC and 0.056 Brier score. The significant factors were different in each machine learning algorithm. If the person sleeps more than two hours during the day, their sex, education level, age, waking up during the night and if the person snores are the variables that most consistently have the highest feature importance in all algorithms.Conclusion: There is an association between sleep disturbances and dementia, which machine learning algorithms can predict. Furthermore, the risk factors for dementia are different across the algorithms, but sleep disturbances can predict dementia.
|
|
| 21. |
- Ou, Jiajie, et al.
(författare)
-
ResTransUnet: An effective network combined with Transformer and U-Net for liver segmentation in CT scans
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 177
-
Tidskriftsartikel (refereegranskat)abstract
- Liver segmentation is a fundamental prerequisite for the diagnosis and surgical planning of hepatocellular carcinoma. Traditionally, the liver contour is drawn manually by radiologists using a slice-by-slice method. However, this process is time-consuming and error-prone, depending on the radiologist's experience. In this paper, we propose a new end-to-end automatic liver segmentation framework, named ResTransUNet, which exploits the transformer's ability to capture global context for remote interactions and spatial relationships, as well as the excellent performance of the original U-Net architecture. The main contribution of this paper lies in proposing a novel fusion network that combines Unet and Transformer architectures. In the encoding structure, a dual-path approach is utilized, where features are extracted separately using both convolutional neural networks (CNNs) and Transformer networks. Additionally, an effective feature enhancement unit is designed to transfer the global features extracted by the Transformer network to the CNN for feature enhancement. This model aims to address the drawbacks of traditional Unet-based methods, such as feature loss during encoding and poor capture of global features. Moreover, it avoids the disadvantages of pure Transformer models, which suffer from large parameter sizes and high computational complexity. The experimental results on the LiTS2017 dataset demonstrate remarkable performance for our proposed model, with Dice coefficients, volumetric overlap error (VOE), and relative volume difference (RVD) values for liver segmentation reaching 0.9535, 0.0804, and −0.0007, respectively. Furthermore, to further validate the model's generalization capability, we conducted tests on the 3Dircadb, Chaos, and Sliver07 datasets. The experimental results demonstrate that the proposed method outperforms other closely related models with higher liver segmentation accuracy. In addition, significant improvements can be achieved by applying our method when handling liver segmentation with small and discontinuous liver regions, as well as blurred liver boundaries. The code is available at the website: https://github.com/Jouiry/ResTransUNet.
|
|
| 22. |
- Panda, Rishab, et al.
(författare)
-
Network analysis of chromophore binding site in LOV domain
- 2023
-
Ingår i: Computers in Biology and Medicine. - : PERGAMON-ELSEVIER SCIENCE LTD. - 0010-4825 .- 1879-0534. ; 161
-
Tidskriftsartikel (refereegranskat)abstract
- Photoreceptor proteins are versatile toolbox for developing biosensors for optogenetic applications. These molecular tools get activated upon illumination of blue light, which in turn offers a non-invasive method for gaining high spatiotemporal resolution and precise control of cellular signal transduction. The Light-Oxygen-Voltage (LOV) domain family of proteins is a well-recognized system for constructing optogenetic devices. Translation of these proteins into efficient cellular sensors is possible by tuning their photochemistry lifetime. However, the bottleneck is the need for more understanding of the relationship between the protein environment and photocycle kinetics. Significantly, the effect of the local environment also modulates the electronic structure of chromophore, which perturbs the electrostatic and hydrophobic interaction within the binding site. This work highlights the critical factors hidden in the protein networks, linking with their experimental photocycle kinetics. It presents an opportunity to quantitatively examine the alternation in chromophore's equilibrium geometry and identify details which have substantial implications in designing synthetic LOV constructs with desirable photocycle efficiency.
|
|
| 23. |
- Qu, Zhiguo, et al.
(författare)
-
Quantum conditional generative adversarial network based on patch method for abnormal electrocardiogram generation
- 2023
-
Ingår i: Computers in Biology and Medicine. - Oxford : Elsevier. - 0010-4825 .- 1879-0534. ; 166, s. 1-13
-
Tidskriftsartikel (refereegranskat)abstract
- To address the scarcity and class imbalance of abnormal electrocardiogram (ECG) databases, which are crucial in AI-driven diagnostic tools for potential cardiovascular disease detection, this study proposes a novel quantum conditional generative adversarial algorithm (QCGAN-ECG) for generating abnormal ECG signals. The QCGAN-ECG constructs a quantum generator based on patch method. In this method, each sub-generator generates distinct features of abnormal heartbeats in different segments. This patch-based generative algorithm conserves quantum resources and makes QCGAN-ECG practical for near-term quantum devices. Additionally, QCGAN-ECG introduces quantum registers as control conditions. It encodes information about the types and probability distributions of abnormal heartbeats into quantum registers, rendering the entire generative process controllable. Simulation experiments on Pennylane demonstrated that the QCGAN-ECG could generate completely abnormal heartbeats with an average accuracy of 88.8%. Moreover, the QCGAN-ECG can accurately fit the probability distribution of various abnormal ECG data. In the anti-noise experiments, the QCGAN-ECG showcased outstanding robustness across various levels of quantum noise interference. These results demonstrate the effectiveness and potential applicability of the QCGAN-ECG for generating abnormal ECG signals, which will further promote the development of AI-driven cardiac disease diagnosis systems. The source code is available at github.com/VanSWK/QCGAN_ECG. © 2023 The Author(s)
|
|
| 24. |
- Razvadauskas, Haroldas, et al.
(författare)
-
Exploring classical machine learning for identification of pathological lung auscultations
- 2024
-
Ingår i: Computers in Biology and Medicine. - Oxford : Elsevier. - 0010-4825 .- 1879-0534. ; 168
-
Tidskriftsartikel (refereegranskat)abstract
- The use of machine learning in biomedical research has surged in recent years thanks to advances in devices and artificial intelligence. Our aim is to expand this body of knowledge by applying machine learning to pulmonary auscultation signals. Despite improvements in digital stethoscopes and attempts to find synergy between them and artificial intelligence, solutions for their use in clinical settings remain scarce. Physicians continue to infer initial diagnoses with less sophisticated means, resulting in low accuracy, leading to suboptimal patient care. To arrive at a correct preliminary diagnosis, the auscultation diagnostics need to be of high accuracy. Due to the large number of auscultations performed, data availability opens up opportunities for more effective sound analysis. In this study, digital 6-channel auscultations of 45 patients were used in various machine learning scenarios, with the aim of distinguishing between normal and abnormal pulmonary sounds. Audio features (such as fundamental frequencies F0-4, loudness, HNR, DFA, as well as descriptive statistics of log energy, RMS and MFCC) were extracted using the Python library Surfboard. Windowing, feature aggregation, and concatenation strategies were used to prepare data for machine learning algorithms in unsupervised (fair-cut forest, outlier forest) and supervised (random forest, regularized logistic regression) settings. The evaluation was carried out using 9-fold stratified cross-validation repeated 30 times. Decision fusion by averaging the outputs for a subject was also tested and found to be helpful. Supervised models showed a consistent advantage over unsupervised ones, with random forest achieving a mean AUC ROC of 0.691 (accuracy 71.11%, Kappa 0.416, F1-score 0.675) in side-based detection and a mean AUC ROC of 0.721 (accuracy 68.89%, Kappa 0.371, F1-score 0.650) in patient-based detection. © Copyright 2024 Elsevier B.V., All rights reserved
|
|
| 25. |
- Sahebi, Golnaz, et al.
(författare)
-
GeFeS : A generalized wrapper feature selection approach for optimizing classification performance
- 2020
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 125
-
Tidskriftsartikel (refereegranskat)abstract
- In this paper, we propose a generalized wrapper-based feature selection, called GeFeS, which is based on a parallel new intelligent genetic algorithm (GA). The proposed GeFeS works properly under different numerical dataset dimensions and sizes, carefully tries to avoid overfitting and significantly enhances classification accuracy. To make the GA more accurate, robust and intelligent, we have proposed a new operator for features weighting, improved the mutation and crossover operators, and integrated nested cross-validation into the GA process to properly validate the learning model. The k-nearest neighbor (kNN) classifier is utilized to evaluate the goodness of selected features. We have evaluated the efficiency of GeFeS on various datasets selected from the UCI machine learning repository. The performance is compared with state-of-the-art classification and feature selection methods. The results demonstrate that GeFeS can significantly generalize the proposed multi-population intelligent genetic algorithm under different sizes of two-class and multi-class datasets. We have achieved the average classification accuracy of 95.83%, 97.62%, 99.02%, 98.51%, and 94.28% while reducing the number of features from 56 to 28, 34 to 18, 279 to 135, 30 to 16, and 19 to 9 under lung cancer, dermatology, arrhythmia, WDBC, and hepatitis, respectively.
|
|
| 26. |
- Salehi, Amir M., et al.
(författare)
-
Reiterative modeling of combined transcriptomic and proteomic features refines and improves the prediction of early recurrence in squamous cell carcinoma of head and neck
- 2022
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 149
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Patients with squamous cell carcinoma of the head and neck (SCCHN) have a high-risk of recurrence. We aimed to develop machine learning methods to identify transcriptomic and proteomic features that provide accurate classification models for predicting risk of early recurrence in SCCHN patients.Methods: Clinical, genomic, transcriptomic and proteomic features distinguishing recurrence risk were examined in SCCHN patients from The Cancer Genome Atlas (TCGA). Recurrence within one year after treatment was classified as high-risk and no recurrence as low-risk.Results: No significant differences in individual clinicopathological characteristics, mutation profiles or mRNA expression patterns were seen between the groups using conventional statistical analysis. Using the machine learning algorithm, extreme gradient boosting (XGBoost), ten proteins (RAD50, 4E-BP1, MYH11, MAP2K1, BECN1, NF2, RAB25, ERRFI1, KDR, SERPINE1) and five mRNAs (PLAUR, DKK1, AXIN2, ANG and VEGFA) made the greatest contribution to classification. These features were used to build improved models in XGBoost, achieving the best discrimination performance when combining transcriptomic and proteomic data, providing an accuracy of 0.939 and an Area Under the ROC Curve (AUC) of 0.951.Conclusions: This study highlights machine learning to identify transcriptomic and proteomic factors that play important roles in predicting risk of recurrence in patients with SCCHN and to develop such models by iterative cycles to enhance their accuracy, thereby aiding the introduction of personalized treatment regimens.
|
|
| 27. |
|
|
| 28. |
- Schilcher, Jörg, 1978-, et al.
(författare)
-
Fusion of electronic health records and radiographic images for a multimodal deep learning prediction model of atypical femur fractures
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 168
-
Tidskriftsartikel (refereegranskat)abstract
- Atypical femur fractures (AFF) represent a very rare type of fracture that can be difficult to discriminate radiologically from normal femur fractures (NFF). AFFs are associated with drugs that are administered to prevent osteoporosis-related fragility fractures, which are highly prevalent in the elderly population. Given that these fractures are rare and the radiologic changes are subtle currently only 7% of AFFs are correctly identified, which hinders adequate treatment for most patients with AFF. Deep learning models could be trained to classify automatically a fracture as AFF or NFF, thereby assisting radiologists in detecting these rare fractures. Historically, for this classification task, only imaging data have been used, using convolutional neural networks (CNN) or vision transformers applied to radiographs. However, to mimic situations in which all available data are used to arrive at a diagnosis, we adopted an approach of deep learning that is based on the integration of image data and tabular data (from electronic health records) for 159 patients with AFF and 914 patients with NFF. We hypothesized that the combinatorial data, compiled from all the radiology departments of 72 hospitals in Sweden and the Swedish National Patient Register, would improve classification accuracy, as compared to using only one modality. At the patient level, the area under the ROC curve (AUC) increased from 0.966 to 0.987 when using the integrated set of imaging data and seven pre-selected variables, as compared to only using imaging data. More importantly, the sensitivity increased from 0.796 to 0.903. We found a greater impact of data fusion when only a randomly selected subset of available images was used to make the image and tabular data more balanced for each patient. The AUC then increased from 0.949 to 0.984, and the sensitivity increased from 0.727 to 0.849.These AUC improvements are not large, mainly because of the already excellent performance of the CNN (AUC of 0.966) when only images are used. However, the improvement is clinically highly relevant considering the importance of accuracy in medical diagnostics. We expect an even greater effect when imaging data from a clinical workflow, comprising a more diverse set of diagnostic images, are used.
|
|
| 29. |
- Shamohammadi, Hossein, et al.
(författare)
-
3D numerical simulation of hot airflow in the human nasal cavity and trachea
- 2022
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825 .- 1879-0534. ; 147, s. 105702-
-
Tidskriftsartikel (refereegranskat)abstract
- Background and objective: The primary function of the human respiratory system is gas and moisture exchange, and conditioning inhaled air to prevent damage to the lungs and alveoli. In a fire incident, exposed soft tissues contract and the respiratory system may be severely damaged, possibly leading to respiratory failure and even respiratory arrest. The purpose of this study is to numerically simulate hot airflow in the human upper airway and trachea to investigate heat and moisture transfer and induced thermal injuries. Methods: For analysis, the airflow is assumed to be laminar and steady, and simulations have been carried out at volume flow rates of 5 and 10 L/min, inlet temperatures of 70-240 ?C, and relative humidity up to 40%. The mucous layer and surrounding tissues are incorporated into the conducting zone of the model. The blood perfusion is considered at different rates up to 5(Kg/m3.s) to regulate the temperature, and the vapor concentration is coupled with the energy equation. Results: The temperature and humidity distribution on the airway wall were calculated for all the studied conditions in order to find the mild and severe burn for different inhaled air temperatures. At the inlet temperatures of 70 and 100 ?C, there are mild burns in several nasal cavity regions. At the higher temperatures of 160 and 200 ?C, these areas suffer from severe burns and mild burns occur at the superior parts and nasopharynx. Rapid evaporation and tissue destruction will be observed if anyone breathes the 240 ?C air. Conclusions: The results show that the hot inlet temperatures drop below 44 ?C when passing through the upper airway, and the lower airway was not affected. Increasing the inlet temperature from 70 to 240 ?C extends the burns from mild to severe and the affected areas from the beginning of the nasal cavity to the pharynx.
|
|
| 30. |
|
|
| 31. |
|
|
| 32. |
- Singh, Ram, et al.
(författare)
-
Impact of quarantine on fractional order dynamical model of Covid-19
- 2022
-
Ingår i: Computers in Biology and Medicine. - Oxford : Elsevier. - 0010-4825 .- 1879-0534. ; 151, Part A
-
Tidskriftsartikel (refereegranskat)abstract
- In this paper, a Covid-19 dynamical transmission model of a coupled non-linear fractional differential equation in the Atangana-Baleanu Caputo sense is proposed. The basic dynamical transmission features of the proposed system are briefly discussed. The qualitative as well as quantitative results on the existence and uniqueness of the solutions are evaluated through the fixed point theorem. The Ulam-Hyers stability analysis of the suggested system is established. The two-step Adams-Bashforth-Moulton (ABM) numerical method is employed to find its numerical solution. The numerical simulation is performed to accesses the impact of various biological parameters on the dynamics of Covid-19 disease. © 2022 The Author(s). Published by Elsevier Ltd. All rights reserved.
|
|
| 33. |
- Wang, Gang, et al.
(författare)
-
US2Mask : Image-to-mask generation learning via a conditional GAN for cardiac ultrasound image segmentation
- 2024
-
Ingår i: Computers in Biology and Medicine. - Oxford : Elsevier. - 0010-4825 .- 1879-0534. ; 172, s. 1-13
-
Tidskriftsartikel (refereegranskat)abstract
- Cardiac ultrasound (US) image segmentation is vital for evaluating clinical indices, but it often demands a large dataset and expert annotations, resulting in high costs for deep learning algorithms. To address this, our study presents a framework utilizing artificial intelligence generation technology to produce multi-class RGB masks for cardiac US image segmentation. The proposed approach directly performs semantic segmentation of the heart's main structures in US images from various scanning modes. Additionally, we introduce a novel learning approach based on conditional generative adversarial networks (CGAN) for cardiac US image segmentation, incorporating a conditional input and paired RGB masks. Experimental results from three cardiac US image datasets with diverse scan modes demonstrate that our approach outperforms several state-of-the-art models, showcasing improvements in five commonly used segmentation metrics, with lower noise sensitivity. Source code is available at https://github.com/energy588/US2mask. © 2024 Elsevier Ltd
|
|
| 34. |
|
|
| 35. |
- Yang, Jingmei, et al.
(författare)
-
Synergy and antagonism between azacitidine and FLT3 inhibitors
- 2024
-
Ingår i: Computers in Biology and Medicine. - : Elsevier. - 0010-4825 .- 1879-0534. ; 169, s. 107889-107889
-
Tidskriftsartikel (refereegranskat)abstract
- Synergetic interactions between drugs can make a drug combination more effective. Alternatively, they may allow to use lower concentrations and thus avoid toxicities or side effects that not only cause discomfort but might also reduce the overall survival. Here, we studied whether synergy exists between agents that are used for treatment of acute myeloid leukaemia (AML). Azacitidine is a demethylation agent that is used in the treatment of AML patients that are unfit for aggressive chemotherapy. An activating mutation in the FLT3 gene is common in AML patients and in the absence of specific treatment makes prognosis worse. FLT3 inhibitors may be used in such cases. We sought to determine whether combination of azacitidine with a FLT3 inhibitor (gilteritinib, quizartinib, LT-850-166, FN-1501 or FF-10101) displayed synergy or antagonism. To this end, we calculated dose–response matrices of these drug combinations from experiments in human AML cells and subsequently analysed the data using a novel consensus scoring algorithm. The results show that combinations that involved non-covalent FLT3 inhibitors, including the two clinically approved drugs gilteritinib and quizartinib were antagonistic. On the other hand combinations with the covalent inhibitor FF-10101 had some range of concentrations where synergy was observed.
|
|
| 36. |
|
|
| 37. |
- Henriksson, Mikael, et al.
(författare)
-
Short-term reproducibility of parameters characterizing atrial fibrillatory waves
- 2020
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825. ; 117
-
Tidskriftsartikel (refereegranskat)abstract
- Objective: To study reproducibility of f-wave parameters in terms of inter- and intrapatient variation. Approach: Five parameters are investigated: dominant atrial frequency (DAF), f-wave amplitude, phase dispersion, spectral organization, and spatiotemporal variability. For each parameter, the variance ratio R, defined as the ratio between inter- and intrapatient variance, is computed; a larger R corresponds to better stability and reproducibility. The study population consists of 20 high-quality ECGs recorded from patients with atrial fibrillation (11/9 paroxysmal/persistent). Main results: The well-established parameters DAF and f-wave amplitude were associated with considerably larger R-values (13.1 and 21.0, respectively) than phase dispersion (2.4), spectral organization (2.4), andspatiotemporal variability (2.7). The use of an adaptive harmonic frequency tracker to estimate the DAF resulted in a larger R (13.1) than did block-based maximum likelihood estimation (6.3). Significance: This study demonstrates a noticeable difference in reproducibility among f-wave parameters, a resultwhich should be taken into account when performing f-wave analysis.
|
|
| 38. |
- McGlinn, Kris, et al.
(författare)
-
FAIRVASC: A semantic web approach to rare disease registry integration
- 2022
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825. ; 145
-
Tidskriftsartikel (refereegranskat)abstract
- Rare disease data is often fragmented within multiple heterogeneous siloed regional disease registries, each containing a small number of cases. These data are particularly sensitive, as low subject counts make the identification of patients more likely, meaning registries are not inclined to share subject level data outside their registries. At the same time access to multiple rare disease datasets is important as it will lead to new research opportunities and analysis over larger cohorts. To enable this, two major challenges must therefore be overcome. The first is to integrate data at a semantic level, so that it is possible to query over registries and return results which are comparable. The second is to enable queries which do not take subject level data from the registries. To meet the first challenge, this paper presents the FAIRVASC ontology to manage data related to the rare disease anti-neutrophil cytoplasmic antibody (ANCA) associated vasculitis (AAV), which is based on the harmonisation of terms in seven European data registries. It has been built upon a set of key clinical questions developed by a team of experts in vasculitis selected from the registry sites and makes use of several standard classifications, such as Systematized Nomenclature of Medicine - Clinical Terms (SNOMED-CT) and Orphacode. It also presents the method for adding semantic meaning to AAV data across the registries using the declarative Relational to Resource Description Framework Mapping Language (R2RML). To meet the second challenge a federated querying approach is presented for accessing aggregated and pseudonymized data, and which supports analysis of AAV data in a manner which protects patient privacy. For additional security the federated querying approach is augmented with a method for auditing queries (and the uplift process) using the provenance ontology (PROV-O) to track when queries and changes occur and by whom. The main contribution of this work is the successful application of semantic web technologies and federated queries to provide a novel infrastructure that can readily incorporate additional registries, thus providing access to harmonised data relating to unprecedented numbers of patients with rare disease, while also meeting data privacy and security concerns.
|
|
| 39. |
- Olsson, Alexander E., et al.
(författare)
-
Automatic discovery of resource-restricted Convolutional Neural Network topologies for myoelectric pattern recognition
- 2020
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825. ; 120
-
Tidskriftsartikel (refereegranskat)abstract
- Convolutional Neural Networks (CNNs) have been subject to extensive attention in the pattern recognition literature due to unprecedented performance in tasks of information extraction from unstructured data. Whereas available methods for supervised training of a CNN with a given network topology are well-defined with rigorous theoretical justification, procedures for the initial selection of topology are currently not. Work incorporating selection of the CNN topology has instead substantially been guided by the domain-specific expertise of the creator(s), followed by iterative improvement via empirical evaluation. This limitation of methodology is restricting in the pursuit of naturally controlled muscle-computer interfaces, where CNNs have been identified as a promising research avenue but effective topology selection heuristics are lacking. With the goal of mitigating ambiguities in topology selection, this paper presents a systematic approach wherein we apply a novel evolutionary algorithm to search a space of candidate topologies. Furthermore, we constrain the search-space by excluding topologies with excessive inference-time computational complexity, making the obtained results implementable in embedded systems. In contrast to manual topology design, our algorithm requires the user to only specify a relatively small set of intuitive hyperparameters. To validate our approach, we use it in order to create topologies for myoelectric pattern recognition via movement decoding of surface electromyography signals. By collating offline classification accuracies obtained from experiments on a collection of publicly available databases, we demonstrate that our method generates computationally lightweight topologies with performance comparable to those of available alternatives.
|
|
| 40. |
- Patel, Mitesh, et al.
(författare)
-
Effects of Deep Brain Stimulation on Postural Control in Parkinson's Disease
- 2020
-
Ingår i: Computers in Biology and Medicine. - : Elsevier BV. - 0010-4825. ; 122
-
Tidskriftsartikel (refereegranskat)abstract
- The standard approach to the evaluation of tremor in medical practice is subjective scoring. The objective of this study was to show that signal processing of physiological data, that are known to be altered by tremor in Parkinson's disease (PD), can quantify the postural dynamics and the effects of DBS. We measured postural control and its capacity to adapt to balance perturbations with a force platform and perturbed balance by altering visual feedback and using pseudo-random binary sequence perturbations (PRBS) of different durations. Our signal processing involved converting the postural control data into spectral power with Fast-Fourier Transformation across a wide bandwidth and then subdividing this into three bands (0–4 Hz, 4–7 Hz and 7–25 Hz). We quantified the amount of power in each bandwidth. From 25 eligible participants, 10 PD participants (9 males, mean age 63.8 years) fulfilled the inclusion criteria; idiopathic PD responsive to L-Dopa; >1 year use of bilateral STN stimulation. Seventeen controls (9 males, mean age 71.2 years) were studied for comparison. Participants with PD were assessed after overnight withdrawal of anti-PD medications. Postural control was measured with a force platform during quiet stance (35 s) and during PRBS calf muscle vibration that perturbed stance (200 s). Tests were performed with eyes open and eyes closed and with DBS ON and DBS OFF. The balance perturbation period was divided into five sequential 35-s periods to assess the subject's ability to address postural imbalance using adaptation. The signal processing analyses revealed that DBS did not significantly change the dynamics of postural control in the 0–4 Hz spectral power but the device reduced the use of spectral power >4 Hz; a finding that was present in both anteroposterior and lateral directions, during vibration, and more so in eyes open tests. Visual feedback, which usually improves postural stability, was less effective in participants with PD with DBS OFF across all postural sway frequencies during quiet stance and during balance perturbations. The expected adaptation of postural control was found in healthy participants between the first and last balance perturbation period. However, adaptation was almost abolished across all spectral frequencies in both the anteroposterior and lateral directions, with both eyes open and eyes closed and DBS ON and OFF in participants with PD. To conclude, this study revealed that DBS altered the spectral frequency dynamics of postural control in participants through a reduction of the power used >4 Hz. Moreover, DBS tended to increase the stabilizing effect of vision across all spectral bands. However, the signal processing analyses also revealed that DBS was not able to restore adaptive motor control abilities in PD.
|
|
| 41. |
- Åkesson, Julius, et al.
(författare)
-
Random effects during training : Implications for deep learning-based medical image segmentation
- 2024
-
Ingår i: Computers in Biology and Medicine. - 0010-4825. ; 180
-
Tidskriftsartikel (refereegranskat)abstract
- Background: A single learning algorithm can produce deep learning-based image segmentation models that vary in performance purely due to random effects during training. This study assessed the effect of these random performance fluctuations on the reliability of standard methods of comparing segmentation models. Methods: The influence of random effects during training was assessed by running a single learning algorithm (nnU-Net) with 50 different random seeds for three multiclass 3D medical image segmentation problems, including brain tumour, hippocampus, and cardiac segmentation. Recent literature was sampled to find the most common methods for estimating and comparing the performance of deep learning segmentation models. Based on this, segmentation performance was assessed using both hold-out validation and 5-fold cross-validation and the statistical significance of performance differences was measured using the Paired t-test and the Wilcoxon signed rank test on Dice scores. Results: For the different segmentation problems, the seed producing the highest mean Dice score statistically significantly outperformed between 0 % and 76 % of the remaining seeds when estimating performance using hold-out validation, and between 10 % and 38 % when estimating performance using 5-fold cross-validation. Conclusion: Random effects during training can cause high rates of statistically-significant performance differences between segmentation models from the same learning algorithm. Whilst statistical testing is widely used in contemporary literature, our results indicate that a statistically-significant difference in segmentation performance is a weak and unreliable indicator of a true performance difference between two learning algorithms.
|
|
