Neural network technique for detecting emergency states in neurosurgical patients

Abstract

The problem of reliable detection of life-threatening situations in the neurosurgical patient undergoing treatment in the ICU is still far from reaching a satisfactory solution, although several methods of clinical and instrumental evaluation have recently been developed for the early detection of oncoming signs of danger. Continuous monitoring of intracranial pressure (ICP) provides neurosurgeons with valuable information about the current condition of the patient. However, it is increasingly felt that traditional methods of extracting information from the ICP signal have reached their natural limits, mostly because of difficulties in fitting the appropriate mathematical model to this non-linear and non-stationary process. Successful implementations of artificial neural networks in many medical tasks have encouraged the application of this method of ICP processing. Two problems are considered: the prediction of trends in ICP, and recognition of the configuration of unfavourable symptoms likely to signal danger for the neurosurgical patient. The construction of neural network predictors of ICP trends is based on wavelet pre-processing of the original signal. The approach to the second task involves pre-processing of the ICP with spectral and statistical methods and classification of the extracted features of the current signal on an arbitrarily selected scale of danger.

Time-Series Modeling and Forecasting of Cerebral Pressure–Flow Physiology: A Scoping Systematic Review of the Human and Animal Literature

“…A total of 14 studies [ 56 , 58 , 60 , 62 , 64 , 65 , 69 , 70 , 71 , 73 , 78 , 82 , 90 , 92 ] utilized time-series analysis techniques, including dynamic relationship analysis methods, TFA, and wavelet analysis. A total of 5 studies [ 59 , 80 , 89 , 95 , 96 ] employed variations in AR time-series models, 8 studies [ 55 , 57 , 72 , 74 , 77 , 81 , 85 , 86 ] utilized machine learning models, and 19 studies [ 2 , 3 , 61 , 63 , 66 , 67 , 68 , 75 , 76 , 79 , 83 , 84 , 87 , 88 , 91 , 93 , 94 , 97 , 98 ] carried out comparative model evaluations.…”

Development of Traumatic Brain Injury Associated Intracranial Hypertension Prediction Algorithms: A Narrative Review

“…The first use of ML to specifically predict tIH occurred in 2000 when Mariak and associates, 67 in a continuation of earlier work by Swiercz and co-workers, 37 , 44 described using an ANN algorithm to automatically assign an observed ICP waveform to one of four expert-determined risk classes (good, moderate, serious, and severe). Although no definition of the risk classes was provided, the ANN algorithm was consistent with expert scoring 70% of the time.…”

“…Our final goal is to develop an expert bedside system that can automatically analyse versatile information about the patient and recognise, as early as possible, those combinations of symptoms that correspond to a state of emergency. 37 …”

“…into their algorithm. 37 Using this method, they were able to forecast ICP values 3 min in advance with a high degree of accuracy. 37 However, neither group was able to operationalize the DWT pre-processing with an ANN or RNN analysis approach due to the high computational requirements and the need for stable and artefact-free ICP data.…”

“…A variety of recurrent ML algorithms have also been used for both tIH prediction and ICP forecasting with varying degrees of success. 29 , 37 , 44 , 49 , 50 , 67 , 79 Recurrent algorithms, which include RNNs, LSTMs, and other deep neural networks, are a family of algorithms designed to make predictions from temporal dynamic evolution of the input features. 80 , 81 Temporal dynamics is the concept that driver–response relationships are not always constant through time, but rather are conditioned to varying degrees by both recent and remote past events.…”