The TOPOSCOUT expert system for stroke localization.

Spitzer, Klaus; Thie, Andreas; Caplan, Louis R.; Kunze, Klaus

doi:10.1161/01.str.20.9.1195

Cited by 17 publications

(7 citation statements)

References 28 publications

(9 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Agreement rates for the neurologic examination are higher for some findings than others, but vary even more than for history (21-92%). Another computerized system for the diagnosis and localization of stroke by vascular territory showed a high rate of agreement for patients with hemispheric lesions compared with data from the Hamburg Stroke Data Bank.21, 22 We believe that this study is the first to validate a questionnaire and algorithm against a n external panel of expert neurologists. Further, agreement rates may be improved by neurologists consulting one another before rendering an opinion1a or by additional training.…”

Section: Discussionmentioning

confidence: 77%

Validation of the ACAS TIA/stroke algorithm

Karanjia¹,

Nelson²,

Lefkowitz³

et al. 1997

Neurology

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 77%

Validation of the ACAS TIA/stroke algorithm

Karanjia¹,

Nelson²,

Lefkowitz³

et al. 1997

Neurology

View full text Add to dashboard Cite

show abstract

“…hemisphere vs, brainstem) involved and have not attempted the degree of precision utilized here. Even so, the accuracy of these rule-based systems has been in the range of 50-75% (Reggia et al 1984, Spitzer et al 1989. While it is difficult to compare those efforts directly with the results obtained for the diagnostic problem-solving systems evaluated here, it is evident from the results of this study that abductive models of the type we examined can be very powerful, regardless of the plausibility criteria used.…”

Section: Discussionmentioning

confidence: 83%

An experimental study of criteria for hypothesis plausibility

Tuhrim

Reggia

Goodall

1991

Journal of Experimental & Theoretical Artificial Intelligen

View full text Add to dashboard Cite

Abductive diagnostic problem-solving systems use causal relations to infer plausible diagnostic hypotheses. An important but controversial issue for such models is what characteristics should define the most plausible hypotheses. While there are theoretical predictions relevant to this issue; there are almost no empirical data on which to base rational decisions. Accordingly, this study examines four different criteria of hypothesis plausibility in diagnosing the site of brain damage in 100 medical patients. The criteria examined are (1) naive minimal cardinality, (2) irredundancy, (3) most probable (Bayesian), and (4) minimal cardinality when adjacency relations are taken into account. Model performance when these different hypothesis plausibility criteria are used confirms the previously predicted inadequacy of minimal cardinality. It also indicates that irredundancy ('minimality'), the criterion most widely used in current AI models, is not useful in this setting because of the large number of alternative, implausible hypotheses it produces. The most interesting result is that a modified minimal cardinality criterion produces the best hypotheses when measured as the ratio of agreements with human experts per hypothesis generated. In addition, comparing the results of this study to two previous rule-based systems for a similar application indicates that abductive diagnostic systems can be very powerful as application programs. These results, useful in themselves, underscore the need for more systematic empirical studies of abductive problem-solving models.

show abstract

“…We set out to create a tool to assist in the localization of neurologic lesions that would overcome limitations of other neuroanatomical diagnostic programs. Most studies of computer-assisted neuroanatomical pattern recognition programs were designed in the 1980's and 1990's, and thus employ outdated computer technology [8][9][10] . More recently, Kamo et al 11 developed a sophisticated tool demonstrating that computer memory could be valuable in identifying causative lesions based on neuroanatomical pathways.…”

Section: Discussionmentioning

confidence: 99%

Web-Based Software to Assist in the Localization of Neuroanatomical Lesions

Lewis

Strike

Doja

et al. 2011

Can. J. Neurol. Sci.

View full text Add to dashboard Cite

Localization of neurological lesions is an essential skill in neurology requiring the analysis of clinical history and physical examination findings that lead neurologists to a particular lesion. Medical students are often overwhelmed by this concept, so much so that Jozefowicz 1 coined the term "neurophobia" to describe student attitudes towards the study of neurology. Risdale et al 2 suggested that brief and unfocussed neurology training in medical school contributes to neurophobia. They emphasized the need for the development of innovations in neurology education to improve competency among medical students.Computer technology is an integral component of modern medical school curricula 3 . Due to the complexity of neuro-ABSTRACT: Objective: To evaluate the educational effectiveness of a novel, web-based neuroanatomical localization application. Methods: A prototype version of a neuroanatomical localization application was developed, limited to lesions involving Cranial Nerve (CN) VII. Second year medical students at the University of Ottawa were recruited to participate in the study. Participants were exposed to a didactic teaching session on CN VII anatomy. They were subsequently randomized to two groups -one group was granted access to the localization application (the "intervention group"), while the other group was given a booklet of standard textbook resources (the "control group"). Participants then completed a case-based multiple choice test on localization of neurologic lesions associated with CN VII, followed by a questionnaire regarding the experience. Results: Thirty-nine students volunteered to participate. Twenty were randomized to the intervention group and 19 to the control group. There was a mean test score difference of 1.3 (CI. 95 = 0.2, 2.3) that was significantly higher in the intervention group when compared to the control group. Significance was determined by a Wilcoxon rank test (p = 0.028). Questionnaire results were similar for both groups, showing an overall favourable evaluation of the localization application. Conclusions: The results support our hypotheses that students using the application would perform better on the multiple choice question (MCQ) test and there would be an overall preference for its use. The demonstrated educational benefit of the application, in addition to the demand for such a resource expressed by the participants, warrant further investigation into the development of a neurological localization application. RÉSUMÉ: Logiciel Web pour aider à localiser les lésions neuroanatomiques.Objectif : Le but de l'étude était d'évaluer l'efficacité au point de vue didactique d'une application nouvelle d'un logiciel Web de localisation neuroanatomique. Méthode : Un prototype d'une application de localisation neuroanatomique, limité aux lésions du septième nerf crânien (VII NC), a été développé. Des étudiants de deuxième année de médecine de l'Université d'Ottawa ont été recrutés pour participer à cette étude. Les participants étaient exposés à des sessions d'enseignemen...

show abstract

The TOPOSCOUT expert system for stroke localization.

Cited by 17 publications

References 28 publications

Validation of the ACAS TIA/stroke algorithm

Validation of the ACAS TIA/stroke algorithm

An experimental study of criteria for hypothesis plausibility

Web-Based Software to Assist in the Localization of Neuroanatomical Lesions

Contact Info

Product

Resources

About