[Computerized management as an adjunct to rhinomanometry].

Determining the cause of seizures is a significant medical problem, as misdiagnosis can result in increased morbidity and even mortality of patients. The reported research evaluates the efficacy of using an artificial neural network (ANN) for determining epileptic seizure occurrences for patients with lateralized bursts of theta (LBT) EEGs. Training and test cases are acquired from examining records of 1,500 consecutive adult seizure patients. The small resulting pool of 92 patients with LBT EEGs requires using a jack-knife procedure for developing the ANN categorization models. The ANNs are evaluated for accuracy, specificity, and sensitivity on classification of each patient into the correct two-group categorization: epileptic seizure or non-epileptic seizure. The original ANN model using eight variables produces a categorization accuracy of 62%. Following a modified factor analysis, an ANN model utilizing just four of the original variables achieves a categorization accuracy of 68%.

Diagnosis, Computer-Assisted↗

PubMed2001

Computer-assisted visualization of arteriovenous malformations on the home personal computer.

OBJECTIVE: Arteriovenous malformations (AVMs) are difficult lesions to treat, partly because it is difficult to formulate a three-dimensional mental image of the nidus and its supplying arteries, draining veins, and arteries of passage. Our purpose is to develop personal computer software that allows better visualization of complex, three-dimensional, connected vascular anatomy for surgical planning. METHODS: Vessels are defined from magnetic resonance angiograms and are symbolically linked to form vascular trees. The nidus of the AVM is also defined by magnetic resonance angiography. These representations of the nidus and vasculature can be viewed together in a software program that allows the user to color-code groups of vessels or to selectively turn connected groups of vessels "off" to avoid obscuring the part of the image that the user wants to observe. Structures can be viewed from any angle. The vessels can also be shown intersecting any magnetic resonance angiogram slice or superimposed upon digital subtraction angiograms obtained from the same patient. RESULTS: We report results from two patients with AVMs in which our representations were compared with the findings during surgery. Our three-dimensional vascular trees correctly depicted the relationship of the nidus to feeding vessels in three dimensions. We show findings in an additional, unoperated patient for whom vessel trees were created from three-dimensional digital subtraction angiography data and compared with a volume rendering of the original data set. CONCLUSION: Computer-assisted, three-dimensional visualizations of complex vascular anatomy can be helpful in planning the surgical excision of AVMs. Software programs that produce these images can provide important information that is difficult to obtain by traditional techniques. This imaging method is also applicable to guidance of endovascular procedures and removal of complex tumors.

Diagnosis, Computer-Assisted↗

PubMed2001

[Reporting echocardiography exams with the G8-Cardio ANMCO software].

BACKGROUND: The availability of a common computerized program for echocardiographic study archiving and reporting at national and/or international level could make it possible to standardize the echo reports of different echocardiographic laboratories, and to use the wealth of data thus obtainable with echocardiography, and to exploit its capillary territorial distribution, with the aim of collecting echocardiographic data in a standard format for epidemiological, scientific and administrative purposes. METHODS: To develop such a software, an ad hoc joint National Association of Hospital Cardiologists and Italian Society of Echocardiography task force worked in conjunction with the Italian Branch of Agilent Technologies to standardize the phraseology of accepted echocardiographic terms and of the quantitative parameters derived from transthoracic and transesophageal echocardiographic examination at rest as well as during exercise and pharmacological stress, and to develop an ad hoc software. This echocardiographic study archiving and reporting program is part of the whole G8-Cardio ANMCO software developed to computerize the whole cardiological chart. The software has been developed by Agilent Technologies to provide a fast, easy-access and easy to use report generator for the non-computer specialist using DBMS Oracle 7.3 database and Power Builder 5.0 to develop a user-friendly interface. RESULTS: The number of qualitative and quantitative variables contained in the program is 733 for echocardiography at rest, while it depends on the stressor and on the length of the examination for the stress echo (dipyridamole 214-384, dobutamine 236-406, exercise 198-392). The program was tested and refined in our laboratory between November 1999 and May 2000. During this time period, 291 resting and 56 stress echocardiographic studies were reported and recorded in a database. On average, each resting echocardiographic study lasting 10 +/- 4 (range 5-17) min was recorded using 50 +/- 11 (range 33-67) variables and 41,566 bytes of hard-disk memory space. Stress echocardiographic studies, each lasting 7 +/- 5 (range 5-21) min, were recorded using 143 +/- 74 (range 38-194) variables and 38,531 bytes of hard-disk memory space. CONCLUSIONS: To our knowledge this software represents the first experience of a common computerized program for echo archiving and reporting carried out at national level.

Diagnosis, Computer-Assisted↗