optimization algorithm
Dotaz
Zobrazit nápovědu
The A* - Algorithm for heuristic search is applied to construct a Neural Network structure (NS) that optimally fits the structure of data to be learned. In this way, the user of Neural Networks (NN) is able to avoid the empirical testing of different structures. The method given here is applied to the recognition of different patterns derived from the EEG of an epileptic patient.
BACKGROUND: Device-based algorithms offer the potential for automated optimization of cardiac resynchronization therapy (CRT), but the process for accepting them into clinical use is currently still ad-hoc, rather than based on pre-clinical and clinical testing of specific features of validity. We investigated how the QuickOpt-guided VV delay (VVD) programming performs against the clinical and engineering heuristic of QRS complex shortening by CRT. METHODS: A prospective, 2-center study enrolled 37 consecutive patients with CRT. QRS complex duration (QRSd) was assessed during intrinsic atrioventricular conduction, synchronous biventricular pacing, and biventricular pacing with QuickOpt-proposed VVD. The measurements were done manually by electronic calipers in signal-averaged and magnified 12-lead QRS complexes. RESULTS: Native QRSd was 174 ± 22 ms. Biventricular pacing with empiric AVD and synchronous VVD resulted in QRSd 156 ± 20 ms, a significant narrowing from the baseline QRSd by 17 ± 27 ms, P = 0.0003. In 36 of 37 patients, the QuickOpt algorithm recommended left ventricular preexcitation with VVD of 42 ± 18 ms (median 40 ms; interquartile range 30-55 ms, P <0.00001). QRSd in biventricular pacing with QuickOpt-based VVD was significantly longer compared with synchronous biventricular pacing (168 ± 25 ms vs. 156 ± 20 ms; difference 12 ± 11ms; P <0.00001). This prolongation correlated with the absolute VVD value (R = 0.66, P <0.00001). CONCLUSIONS: QuickOpt algorithm systematically favours a left-preexcitation VVD which translates into a significant prolongation of the QRSd compared to synchronous biventricular pacing. There is no reason to believe that a manipulation that systematically widens QRSd should be considered to optimize physiology. Device-based CRT optimization algorithms should undergo systematic mechanistic pre-clinical evaluation in various scenarios before they are tested in large clinical studies.
Objective.The current practices of designing neural networks rely heavily on subjective judgment and heuristic steps, often dictated by the level of expertise possessed by architecture designers. To alleviate these challenges and streamline the design process, we propose an automatic method, a novel approach to enhance the optimization of neural network architectures for processing intracranial electroencephalogram (iEEG) data.Approach.We present a genetic algorithm, which optimizes neural network architecture and signal pre-processing parameters for iEEG classification.Main results.Our method improved the macroF1 score of the state-of-the-art model in two independent datasets, from St. Anne's University Hospital (Brno, Czech Republic) and Mayo Clinic (Rochester, MN, USA), from 0.9076 to 0.9673 and from 0.9222 to 0.9400 respectively.Significance.By incorporating principles of evolutionary optimization, our approach reduces the reliance on human intuition and empirical guesswork in architecture design, thus promoting more efficient and effective neural network models. The proposed method achieved significantly improved results when compared to the state-of-the-art benchmark model (McNemar's test,p≪ 0.01). The results indicate that neural network architectures designed through machine-based optimization outperform those crafted using the subjective heuristic approach of a human expert. Furthermore, we show that well-designed data preprocessing significantly affects the models' performance.
- MeSH
- elektroencefalografie metody MeSH
- elektrokortikografie * MeSH
- lidé MeSH
- neuronové sítě * MeSH
- počítačové zpracování signálu MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
Diet, stress, genetics, and a sedentary lifestyle may all contribute to heart disease rates. Although recent studies propose comprehensive automated diagnostic systems, these systems tend to focus on one aspect, such as feature selection, prioritization, or predictive accuracy. A more complete approach that considers all of these factors can improve the efficiency of a cardiac prediction system. This study uses an appropriate strategy to overcome potential network design problems, design challenges, overfitting, and lack of robustness that can interfere with system performance. The research introduces an ideally designed deep trust network called ID-DTN to improve system performance. The Ruzzo-Tompa method is used to eliminate noncontributory features. The Seagull Optimization Algorithm (SOA) is introduced to optimize the trust depth network to achieve optimal network design. The study scrutinizes the deep trust network (ID-DTN) and the restricted Boltzmann machine (RBM) and sheds light on the system's operation. This proposal can optimize both network architecture and feature selection, which is the main novelty. The proposed method is analyzed using the below-mentioned metrics: Matthew's correlation coefficient, F1 score, accuracy, sensitivity, specificity, and accuracy. ID-DTN performs well compared to other state-of-the-art methods. The validation results confirm that the proposed method improves the prediction accuracy to 97.11% and provides reliable recommendations for patients with cardiovascular disease.
- MeSH
- algoritmy * MeSH
- lidé MeSH
- nemoci srdce * diagnóza MeSH
- neuronové sítě MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
554 s.
- Klíčová slova
- struktury chemické-aktivita biologická,
- Konspekt
- Biochemie. Molekulární biologie. Biofyzika
- NLK Obory
- chemie, klinická chemie
- biochemie
Lacerace plíce je jedním ze závažných důsledků jak tupých, tak penetrujících traumat hrudníku. Autoři referují o 16 případech těchto poranění diagnostikovaných a operačně léčených na vlastním pracovišti v uplynulých pěti letech. Ze čtyř pětin se jednalo o následek tupého úrazu hrudníku, a to převážně v důsledku dopravních nehod. U 12 osob byla plicní lacerace součástí závažného polytraumatu, polovina z nich si vyžádala anatomickou plicní resekci. Naopak všechna tři penetrující poranění bylo možno ošetřit pouze suturou, resp. klínovitou plicní resekcí, jedenkrát s využitím miniinvazivní metodiky. Pooperační komplikace přežívajících nemocných byly řídké a nevynutily si operační revizi. Během operace či bezprostředně po ní zemřeli 4 muži (25 %), vždy pro mnohočetná poranění neslučitelná se životem. Míra poškození plicního parenchymu byla v souboru přímo úměrná velikosti zraňujícího násilí, hluboké ruptury byly ve všech případech součástí polytraumatu. Optimálním operačním přístupem, především při vícečetných a hlubokých laceracích, se jeví různé typy torakotomie, indikace pro videotorakoskopii jsou řídké. V rámci vlastního ošetření plicních ruptur se upřednostňují záchovné zákroky. Ve sdělení jsou diskutovány i další aspekty diagnostiky a léčby lacerací plic.
Lung laceration is one of the most serious consequences of both blunt, and penetrating trauma of the thorax. The authors described 16 cases diagnosed and surgically treated at their clinic over last five years. In about 80 % of cases, the laceration was a consequence of blunt trauma to the thorax, caused mainly by a traffic accident. In 12 patients lung laceration was part of serious polytrauma, and anatomic lung resection was necessary in one half of these cases. In contrast, all 3 cases of penetrating trauma could be treated only using sutures, or by wedge-shaped lung resection in one patient using a miniinvasive method. Postoperative complications in surviving patients were rare and operative revision was unnecessary. During or immediately after operation 4 male patients died (25 %), in all cases from multiple trauma incompatible with life. The extent of lung parenchyma damage in our group was proportional to the extent of force of the injury. Deep ruptures were in all cases a part of polytrauma. The optimal surgical method, espe-cially in multiple and deep lacerations, seems to be various types of thoracotomy, indications for videothoracoscopy occur rarely. In the frame-work for the proper treatment of lung ruptures, saving procedures are preferred. Other aspects of diagnoses and therapy of lung lacerations are also discussed.
- MeSH
- algoritmy MeSH
- dopravní nehody MeSH
- dospělí MeSH
- lacerace etiologie chirurgie komplikace MeSH
- lidé středního věku MeSH
- lidé MeSH
- mladiství MeSH
- plíce chirurgie MeSH
- poranění hrudníku diagnóza etiologie chirurgie MeSH
- poškození plic MeSH
- torakotomie využití MeSH
- video-asistovaná chirurgie MeSH
- Check Tag
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- mladiství MeSH
- mužské pohlaví MeSH
- ženské pohlaví MeSH
- Publikační typ
- přehledy MeSH
Cíl. Cílem studie bylo posoudit možnosti IRIS (iterativní rekonstrukce v obrazovém prostoru) algoritmu z hlediska redukce šumu v obraze a efektivní dávky. Metoda. Bylo vyhodnoceno celkem 85 pacientů, kteří podstoupili CT vyšetření (plic, břicha a mozku) s rekonstrukcí snímků pomocí IRIS algoritmu, a podobná skupina 95 pacientů, jejichž snímky byly rekonstruovány filtrovanou zpětnou projekcí (FBP). Referenční součin proudu a expozičního času byl pro skupinu s IRIS algoritmem v původních užívaných protokolech snižován tak, aby kvalita snímku na základě subjektivního posouzení radiologa byla zachována. Obě skupiny byly porovnány z hlediska množství šumu na snímku, poměru signálu a šumu a efektivní dávky, kterou pacient během vyšetření obdrží. Výsledky. Ukázalo se, že u vyšetření břicha je možné radiační zátěž pacienta snížit až o 50 %, aniž by došlo ke zhoršení kvality snímku. Na druhou stranu pro vyšetření mozku se podařilo dosáhnout snížení průměrné efektivní dávky pouze o 23 %. Závěr. Z tohoto důvodu by měla obdobná měření a optimalizace vždy předcházet uvedení nového rekonstrukčního algoritmu do klinické praxe.
Aim. The aim of this study was to assess the possibilities of IRIS (Iterative reconstruction in Image Space) algorithm in noise reduction and potential dose reduction. Method. A total of 85 patients clinically underwent CT examination (chest, abdomen, brain) with IRIS reconstruction and similar groups of 95 patients with Filtered Back Projection (FBP) reconstruction were evaluated. The reference tube current product in original CT protocols was modified for IRIS algorithm according to the subjective image quality. Two reconstruction algorithms were compared in terms of image noise, signal to noise ratio and effective dose. Results. It was demonstrated that with the novel approach the mean effective dose reduction up to 50% for CT examination of the abdomen did not lead to image quality degradation. However, to retain the quality of CT images of the brain it was accomplished mean effective dose reduction only of 23%. Conclusion. For this reason such measurements and similar optimization should precede to introduction of any new reconstruction algorithm into clinical routine.
- Klíčová slova
- CT, efektivní dávka,
- MeSH
- algoritmy MeSH
- artefakty MeSH
- dávka záření MeSH
- diagnostické zobrazování metody přístrojové vybavení trendy MeSH
- lidé MeSH
- počítačová rentgenová tomografie metody trendy MeSH
- počítačové zpracování obrazu metody trendy využití MeSH
- řízení kvality MeSH
- software trendy MeSH
- statistika jako téma MeSH
- tělesná hmotnost MeSH
- výsledky a postupy - zhodnocení (zdravotní péče) MeSH
- Check Tag
- lidé MeSH
Several local search algorithms for real-valued domains (axis parallel line search, Nelder-Mead simplex search, Rosenbrock's algorithm, quasi-Newton method, NEWUOA, and VXQR) are described and thoroughly compared in this article, embedding them in a multi-start method. Their comparison aims (1) to help the researchers from the evolutionary community to choose the right opponent for their algorithm (to choose an opponent that would constitute a hard-to-beat baseline algorithm), (2) to describe individual features of these algorithms and show how they influence the algorithm on different problems, and (3) to provide inspiration for the hybridization of evolutionary algorithms with these local optimizers. The recently proposed Comparing Continuous Optimizers (COCO) methodology was adopted as the basis for the comparison. The results show that in low dimensional spaces, the old method of Nelder and Mead is still the most successful among those compared, while in spaces of higher dimensions, it is better to choose an algorithm based on quadratic modeling, such as NEWUOA or a quasi-Newton method.
- MeSH
- algoritmy * MeSH
- benchmarking * MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
- srovnávací studie MeSH
Four methods for global numerical black box optimization with origins in the mathematical programming community are described and experimentally compared with the state of the art evolutionary method, BIPOP-CMA-ES. The methods chosen for the comparison exhibit various features that are potentially interesting for the evolutionary computation community: systematic sampling of the search space (DIRECT, MCS) possibly combined with a local search method (MCS), or a multi-start approach (NEWUOA, GLOBAL) possibly equipped with a careful selection of points to run a local optimizer from (GLOBAL). The recently proposed "comparing continuous optimizers" (COCO) methodology was adopted as the basis for the comparison. Based on the results, we draw suggestions about which algorithm should be used depending on the available budget of function evaluations, and we propose several possibilities for hybridizing evolutionary algorithms (EAs) with features of the other compared algorithms.
