User Tools

Site Tools


machine_learning

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
machine_learning [2018/10/21 10:03]
administrador
machine_learning [2018/11/10 17:16] (current)
administrador
Line 7: Line 7:
 ML methods are already widely applied in multiple aspects of our daily lives, although this is not always obvious to the casual observer; common examples are email spam filters, search suggestions,​ online shopping suggestions,​ and speech recognition in smartphones ML methods are already widely applied in multiple aspects of our daily lives, although this is not always obvious to the casual observer; common examples are email spam filters, search suggestions,​ online shopping suggestions,​ and speech recognition in smartphones
 ((Jordan MI, Mitchell TM. Machine learning: trends, perspectives,​ and prospects. Science . 2015;​349(6245):​255-260.)). ((Jordan MI, Mitchell TM. Machine learning: trends, perspectives,​ and prospects. Science . 2015;​349(6245):​255-260.)).
----- 
-ML is increasingly tested in neurosurgical applications and even demonstrated to emulate the performance of clinical [[expert]]s 
-((Mariak Z, Swiercz M, Krejza J, Lewko J, Lyson T. Intracranial pressure processing with artificial neural networks: classification of signal properties. Acta Neurochir (Wien) . 2000;​142(4):​407-411;​ discussion 411-402.)) 
-((Nucci CG, De Bonis P, Mangiola A et al.   ​Intracranial pressure wave morphological classification:​ automated analysis and clinical validation. Acta Neurochir (Wien) . 2016;​158(3):​581-588;​ discussion 588.)) 
-((Sieben G, Praet M, Roels H, Otte G, Boullart L, Calliauw L. The development of a decision support system for the pathological diagnosis of human cerebral tumours based on a neural network classifier. Acta Neurochir (Wien) . 1994;​129(3-4):​193-197.)) 
-((Mathew B, Norris D, Mackintosh I, Waddell G. Artificial intelligence in the prediction of operative findings in low back surgery. Brit J Neurosurg . 1989;​3(2):​161-170.)) 
-((Arle JE, Perrine K, Devinsky O, Doyle WK. Neural network analysis of preoperative variables and outcome in epilepsy surgery. J Neurosurg . 1999;​90(6):​998-1004.)) 
-((Gazit T, Andelman F, Glikmann-Johnston Y et al.   ​Probabilistic machine learning for the evaluation of presurgical language dominance. J Neurosurg . 2016;​125(2):​1-13.)) 
-((Shi HY, Hwang SL, Lee KT, Lin CL. In-hospital mortality after traumatic brain injury surgery: a nationwide population-based comparison of mortality predictors used in artificial neural network and logistic regression models. J Neurosurg . 2013;​118(4):​746-752.)) 
-((Azimi P, Mohammadi HR. Predicting endoscopic third ventriculostomy success in childhood hydrocephalus:​ an artificial neural network analysis. J Neurosurg Pediatr . 2014;​13(4):​426-432.)) 
-((Azimi P, Mohammadi H. Prediction of successful ETV outcome in childhood hydrocephalus:​ an artificial neural networks analysis. J Neurosurg . 2015;​122(6):​426-432.)) 
-((Chang K, Zhang B, Guo X et al.   ​Multimodal imaging patterns predict survival in recurrent glioblastoma patients treated with bevacizumab. Neuro-oncology . 2016;​18(12):​1680-1687.)) 
-((Jones TL, Byrnes TJ, Yang G, Howe FA, Bell BA, Barrick TR. Brain tumor classification using the diffusion tensor image segmentation (D-SEG) technique. Neuro-oncology . 2015;​17(3):​466-476.)) 
-((Macyszyn L, Akbari H, Pisapia JM et al.   ​Imaging patterns predict patient survival and molecular subtype in glioblastoma via machine learning techniques. Neuro-oncology . 2016;​18(3):​417-425.)) 
-((Teplyuk NM, Mollenhauer B, Gabriely G et al.   ​MicroRNAs in cerebrospinal fluid identify glioblastoma and metastatic brain cancers and reflect disease activity. Neuro-oncology . 2012;​14(6):​689-700.)) 
-((Zhang B, Chang K, Ramkissoon S et al.   ​Multimodal MRI features predict isocitrate dehydrogenase genotype in high-grade gliomas. Neuro-oncology . 2017;​19(1):​109-117.)) 
-((Fouke SJ, Weinberger K, Kelsey M et al.   A machine-learning-based classifier for predicting a multi-parametric probability map of active tumor extent within glioblastoma multiforme. Neuro-oncology . 2012;​14:​vi124-vi125.)) 
-((Kim LM, Commean P, Boyd A et al.   ​Predicting the location and probability of viable tumor within glioblastoma multiforme with multiparametric magnetic resonance imaging. Neuro-oncology . 2012;​14:​vi120-vi128.)) 
-((Orphanidou-Vlachou E, Vlachos N, Davies N, Arvanitis T, Grundy R, Peet A. Texture analysis of T1-and t2-weighted magnetic resonance images to discriminate posterior fossa tumors in children. Neuro-oncology . 2014;​16:​i123-i126.)) 
-((Rayfield C, Swanson K. Predicting the response to treatment in GBM: Machine learning on clinical images. Neuro-oncology . 2015;​17:​v167.)) 
-((Akbari H, Macyszyn L, Da X et al.   ​Imaging surrogates of infiltration obtained via multiparametric imaging pattern analysis predict subsequent location of recurrence of glioblastoma. Neurosurgery . 2016;​78(4):​572-580.)) 
-((Mitchell TJ, Hacker CD, Breshears JD et al.   A novel data-driven approach to preoperative mapping of functional cortex using resting-state functional magnetic resonance imaging. Neurosurgery . 2013;​73(6):​969-982;​ discussion 982-963.)) 
-((Oermann EK, Kress MA, Collins BT et al.   ​Predicting survival in patients with brain metastases treated with radiosurgery using artificial neural networks. Neurosurgery . 2013;​72(6):​944-951;​ discussion 952.)) 
-((Taghva A. An automated navigation system for deep brain stimulator placement using hidden Markov models. Neurosurgery . 2010;66(3 Suppl Operative):​108-117;​ discussion 117.)) 
-((Dumont TM, Rughani AI, Tranmer BI. Prediction of symptomatic cerebral vasospasm after aneurysmal subarachnoid hemorrhage with an artificial neural network: feasibility and comparison with logistic regression models. World Neurosurg . 2011;​75(1):​57-63;​ discussion 25-58.)). 
----- 
-Automated analysis of radiological data for diagnosis, segmentation,​ or outcome prediction could, be one of the first ML applications that finds its way to actual clinical practice 
-((Obermeyer Z, Emanuel EJ. Predicting the future - big data, machine learning, and clinical medicine. N Engl J Med . 2016;​375(13):​1216-1219.)). 
----- 
-Current [[outcome]]s [[prediction]] [[tool]]s are largely based on and limited by [[regression]] methods. Utilization of [[machine learning]] (ML) methods that can handle multiple diverse inputs could strengthen predictive abilities and improve patient outcomes. [[Inpatient]] [[length of stay]] (LOS) is one such outcome that serves as a surrogate for patient [[disease]] severity and resource utilization. 
  
-To develop a novel method to systematically rank, select, and combine ML algorithms to build a model that predicts LOS following [[craniotomy]] for [[brain tumor]].+===== Machine learning in neurosurgery =====
  
-A training ​[[dataset]] of 41 222 patients who underwent craniotomy for brain tumor was created from the [[National Inpatient Sample]]. Twenty-nine ML algorithms were trained on 26 preoperative variables to predict LOS. Trained algorithms were ranked by calculating the root mean square logarithmic error (RMSLE) and top performing algorithms combined to form an ensemble. The ensemble was externally validated using a dataset of 4592 patients from the [[National Surgical Quality Improvement Program]]. Additional analyses identified variables that most strongly influence the ensemble model predictions.+see [[Machine learning in neurosurgery]].
  
-The ensemble model predicted LOS with RMSLE of .555 (95% confidence interval, .553-.557) on internal validation and .631 on external validation. Nonelective surgery, preoperative pneumonia, sodium abnormality,​ or weight loss, and non-White race were the strongest predictors of increased LOS. 
- 
-An ML ensemble model predicts LOS with good performance on internal and external validation, and yields clinical insights that may potentially improve patient outcomes. This systematic ML method can be applied to a broad range of clinical problems to improve patient care 
-((Muhlestein WE, Akagi DS, Davies JM, Chambless LB. Predicting Inpatient Length ​ 
-of Stay After Brain Tumor Surgery: Developing [[Machine Learning]] Ensembles to 
-Improve Predictive Performance. Neurosurgery. 2018 Aug 3. doi: 
-10.1093/​neuros/​nyy343. [Epub ahead of print] PubMed PMID: 30113665. 
-)). 
----- 
- 
-A systematic search was performed in the PubMed and Embase databases as of August 2016 to review all studies comparing the performance of various ML approaches with that of clinical experts in neurosurgical literature. 
- 
-Twenty-three studies were identified that used ML algorithms for diagnosis, presurgical planning, or outcome prediction in neurosurgical patients. Compared to clinical experts, ML models demonstrated a median absolute improvement in accuracy and area under the receiver operating curve of 13% (interquartile range 4-21%) and 0.14 (interquartile range 0.07-0.21), respectively. In 29 (58%) of the 50 outcome measures for which a P-value was provided or calculated, ML models outperformed clinical experts (P < .05). In 18 of 50 (36%), no difference was seen between ML and expert performance (P > .05), while in 3 of 50 (6%) clinical experts outperformed ML models (P < .05). All 4 studies that compared clinicians assisted by ML models vs clinicians alone demonstrated a better performance in the first group. 
- 
-Senders et al., conclude that ML models have the potential to augment the decision-making capacity of clinicians in neurosurgical applications;​ however, significant hurdles remain associated with creating, validating, and deploying ML models in the clinical setting. Shifting from the preconceptions of a human-vs-machine to a human-and-machine paradigm could be essential to overcome these hurdles 
-((Senders JT, Arnaout O, Karhade AV, Dasenbrock HH, Gormley WB, Broekman ML, 
-Smith TR. Natural and Artificial Intelligence in Neurosurgery:​ A Systematic 
-Review. Neurosurgery. 2018 Aug 1;​83(2):​181-192. doi: 10.1093/​neuros/​nyx384. 
-PubMed PMID: 28945910. 
-)). 
----- 
- 
- 
-Lazaridis et al., and others, have developed [[predictive model]]s based on [[machine learning]] from continuous time series of [[intracranial pressure]] and [[partial pressure of brain tissue oxygen]]. These models provide accurate predictions of physiologic crises events in a timely fashion, offering the opportunity for an earlier application of targeted interventions.They review the rationale for prediction, discuss available predictive models with examples, and offer suggestions for their future prospective testing in conjunction with preventive clinical [[algorithm]]s 
-((Lazaridis C, Rusin CG, Robertson CS. Secondary Brain Injury: Predicting and 
-Preventing Insults. Neuropharmacology. 2018 Jun 6. pii: S0028-3908(18)30279-X. 
-doi: 10.1016/​j.neuropharm.2018.06.005. [Epub ahead of print] Review. PubMed PMID: 
-29885419. 
-)). 
----- 
- 
-Machine [[learning]] (ML) is a domain of [[artificial intelligence]] that allows computer [[algorithm]]s to learn from experience without being explicitly programmed. 
- 
-To summarize neurosurgical applications of ML where it has been compared to clinical expertise, here referred to as "​natural intelligence."​ 
- 
-Two important and rapidly developing scientific movements—data [[reproducibility]] and [[machine learning]]—are central to a recent Neuron paper by Chung et al 
-((Chung JE, Magland JF, Barnett AH, Tolosa VM, Tooker AC, Lee KY, Shah KG, Felix 
-SH, Frank LM, Greengard LF. A Fully Automated Approach to Spike Sorting. Neuron. ​ 
-2017 Sep 13;​95(6):​1381-1394.e6. doi: 10.1016/​j.neuron.2017.08.030. PubMed PMID: 
-28910621; PubMed Central PMCID: PMC5743236. 
-)) 
----- 
- 
- 
-A systematic search was performed in the PubMed and Embase databases as of August 2016 to review all studies comparing the performance of various ML approaches with that of clinical experts in neurosurgical literature. 
- 
-Twenty-three studies were identified that used ML algorithms for diagnosis, presurgical planning, or outcome prediction in neurosurgical patients. Compared to clinical experts, ML models demonstrated a median absolute improvement in accuracy and area under the receiver operating curve of 13% (interquartile range 4-21%) and 0.14 (interquartile range 0.07-0.21), respectively. In 29 (58%) of the 50 outcome measures for which a P -value was provided or calculated, ML models outperformed clinical experts ( P < .05). In 18 of 50 (36%), no difference was seen between ML and expert performance ( P > .05), while in 3 of 50 (6%) clinical experts outperformed ML models ( P < .05). All 4 studies that compared clinicians assisted by ML models vs clinicians alone demonstrated a better performance in the first group. 
- 
- 
-We conclude that ML models have the potential to augment the decision-making capacity of clinicians in neurosurgical applications;​ however, significant hurdles remain associated with creating, validating, and deploying ML models in the clinical setting. Shifting from the preconceptions of a human-vs-machine to a human-and-machine paradigm could be essential to overcome these hurdles 
-((Senders JT, Arnaout O, Karhade AV, Dasenbrock HH, Gormley WB, Broekman ML, 
-Smith TR. Natural and Artificial Intelligence in Neurosurgery:​ A Systematic 
-Review. Neurosurgery. 2017 Sep 7. doi: 10.1093/​neuros/​nyx384. [Epub ahead of 
-print] PubMed PMID: 28945910. 
-)). 
----- 
-Yepes-Calderon et al. presented a [[segmentation]] strategy based on an algorithm that uses four features extracted from the medical images to create a statistical estimator capable of determining ventricular volume. When compared with manual segmentations,​ the correlation was 94% and holds promise for even better accuracy by incorporating the unlimited data available. The volume of any segmentable structure can be accurately determined utilizing the [[machine learning]] strategy presented and runs fully automatically within the PACS 
-((Yepes-Calderon F, Nelson MD, McComb JG. Automatically measuring brain 
-ventricular volume within PACS using artificial intelligence. PLoS One. 2018 Mar  
-15;​13(3):​e0193152. doi: 10.1371/​journal.pone.0193152. eCollection 2018. PubMed 
-PMID: 29543817. 
-)). 
machine_learning.txt · Last modified: 2018/11/10 17:16 by administrador