» arhiv
Možnosti:
[Prikaži v obliki za tiskanje]
[Poišči nalogo v zbirki Cobiss]
ID naloge: 72 Letnik: 1999 Predmet: oftalmologija
Ocena hitrih testnih strategij pri racunalniško podprti staticni perimetriji
Avtor: Jernej Pajek
Mentor: doc. dr. sc. Marko Hawlina, dr. med
IZHODIŠCE. Racunalniško podprta staticna perimetrija je diagnosticna metoda, pri kateri s predstavljanjem razlicno svetlih dražljajev in z beleženjem odgovorov preiskovanca dolocamo prag diferencialne svetlobne obcutljivosti (DSO) na vec tockah v vidnem polju. Izmerjene vrednosti so podlaga za izracun statisticnih pokazateljev vidnega polja. Najpomembnejša med njimi sta povprecni odklon izmerjenih vrednosti od normalnih vrednosti (MD) in varianca odklonov (LV), ki raste z razlicnostjo vrednosti DSO v vidnem polju in zato dobro zazna prisotnost omejenih izpadov v vidnem polju (skotomov). Programska oprema, ki omogoca upravljanje perimetra, uporablja za dolocanje praga DSO vec razlicnih strategij. Normalna strategija, ki je standard in se rutinsko uporablja, spreminja svetlost predstavljenih dražljajev v vse manjših korakih po 4, 2 in 1 dB, dokler ne doloci praga na 1dB natancno. Razvoj hitrih strategij (dinamicne strategije in TOP strategije) je omogocil skrajšanje preiskave, ki pri normalni strategiji traja povprecno 15 minut in je zato za preiskovanca zahtevna in naporna. Dinamicna strategija skrajša cas preiskave na polovico. Temelji na razširitvi obmocja praga DSO, ki se pojavi v podrocjih znižane DSO, in zato v teh podrocjih uporablja vecje korake pri spreminjanju dražljaja (do 10 dB). TOP strategija skrajša preiskavo na petino casa potrebnega pri normalni strategiji. Vrednosti DSO izmerjene na neki tocki uporablja (upoštevaje oddaljenost) tudi za dolocitev praga sosednjih tock.
NAMEN IN HIPOTEZE. Z analizo rezultatov, ki smo jih dobili z normalno, dinamicno in TOP strategijo pri istih preiskovancih, smo želeli opredeliti razlike med temi strategijami v dolocanju MD, LV, števila vseh tock z izmerjenim odklonom od normalnih vrednosti, števila tock z najvecjo znacilnostjo odklona od normalne vrednosti p<0,005 in razlike v obcutljivosti in specificnosti za splošne kriterije razlocevanja normalnih izvidov preiskave vidnega polja od bolezenskih. Preverjali smo tri hipoteze.
1. Normalna, dinamicna in TOP strategija se pri dolocanju vrednosti MD pomembno ne razlikujejo. TOP strategija izmeri pomembno nižje vrednosti LV kot drugi dve strategiji.
2. V primerjavi z normalno in dinamicno strategijo prikaže TOP strategija izpade glede na število vseh tock z odklonom pomembno širše, glede na število tock z najvecjo znacilnostjo odklona (p<0,005) pa so podrocja v globini izpadov ožja.
3. V obcutljivosti in specificnosti TOP, dinamicne in normalne strategije ni pomembnih razlik.
METODE. V preiskavo smo vkljucili 22 vidnih polj normalnih oci 17 preiskovancev (povprecna starost 33±15 let) in 22 vidnih polj 17 preiskovancev (povprecna starost 47±16 let) z razlicnimi vrstami in stopnjami vidnih okvar (glavkom, vnetne bolezni žilnice in mrežnice, žilne motnje, okvara vidnega živca, odstop mrežnice). Pri tem smo upoštevali eticna dolocila za raziskave na ljudeh in pridobili soglasje Komisije za medicinsko etiko Ministrstva za zdravstvo Republike Slovenije (št. dokumenta 68/06/99). Pred vstopom v raziskavo so imeli vsi opravljeno vsaj eno perimetricno meritev na tem mestu. Vsako vidno polje smo izmerili s TOP, normalno in dinamicno strategijo na perimetru Octopus 101 v standardnih pogojih za to napravo. Vrstni red strategij smo enakomerno menjavali. Odmori med preiskavami so znašali vsaj 25 minut. Vse tri meritve smo opravili v najvec 24 urah. Rezultate smo statisticno ovrednotili z neparametricnim Friedmanovim testom za primerjavo vec odvisnih vzorcev, ob znacilnem ali mejno znacilnem rezultatu Friedmanovega testa smo razlike med posameznimi pari preverili z Wilcoxonovim testom predznacenih rangov. Obcutljivost in specificnost strategij smo dolocali loceno glede na tri splošne kriterije za locevanje normalnih izvidov od bolezenskih (MD>2 dB, LV>6 dB2, prisotnost vsaj 7 tock z odklonom, od tega vsaj tri v skupini). Obcutljivost in specificnost smo statisticno ovrednotili s hi-kvadrat testom prirejenim za oceno parnih vzorcev.
REZULTATI. Pri obeh skupinah med povprecnimi izmerjenimi vrednostmi MD ni bilo statisticno znacilnih razlik (p>0,05). Najvecje posamezne vrednosti razlik v MD so ustrezale razlikam, ki jih lahko pricakujemo zaradi variabilnosti rezultatov med locenimi meritvami. Vrednosti LV s TOP strategijo so bile pri skupini z izpadi vidnega polja povprecno za 11±14 dB2 nižje od LV pri dinamicni strategiji (p<0,01) in 9,8±16 dB2 nižje od LV pri normalni (p<0,061) pri tem pa je ocena za povprecno razliko v LV med TOP in normalno strategijo znašala -17 dB2 < µ < -2,7 dB2 (p<0,05). Pri normalnih vidnih poljih so bile razlike v LV majhne in statisticno neznacilne. Dinamicna strategija izmeri pri skupini vidnih polj z izpadi statisticno znacilno manj tock z odklonom kot drugi dve strategiji - povprecna razlika je bila 3 tocke (p<0,05). V številu tock z najvecjo znacilnostjo odklona izmerjene vrednosti od normalne (p<0,005) med strategijami ni bilo znacilnih razlik. V obcutljivosti in specificnosti pri vseh treh kriterijih med strategijami ni bilo statisticno znacilnih razlik.
ZAKLJUCKI.
1. Top, dinamicna in normalna strategija se v dolocanju MD ne razlikujejo pomembno. TOP strategija izmeri pri vidnih poljih z izpadi pomembno nižje vrednosti LV kot drugi dve strategiji.
2. Glede na število vseh tock z odklonom prikaže dinamicna strategija izpade ožje kot drugi dve strategiji, med katerima ni pomembnih razlik v številu tock z odklonom. Glede na število tock z najvecjo znacilnostjo odklona (p<0,005) prikažejo podrocja v globini izpadov vse tri strategije enako dobro.
3. V obcutljivosti in specificnosti se pri uporabi splošnih kriterijev za razlikovanje normalnega od bolezenskega perimetricnega izvida TOP, dinamicna in normalna strategija znacilno ne razlikujejo.
4. Z izjemo nižjih vrednosti LV pri TOP strategiji so razlike med rezultati TOP, dinamicne in normalne strategije majhne, zato je uporaba dinamicne in TOP strategije, glede na prednosti ki jih prinaša skrajšanje casa preiskave, klinicno utemeljena.
«»
[Abstract / English version]
Evaluation of fast perimetric strategies
Author: Jernej Pajek
Mentor: doc. dr. sc. Marko Hawlina, dr. med
BACKGROUND. Automated perimetry is a diagnostic method which measures the threshold of differential light sensitivity (DLS) on multiple test points in visual field and is based on presenting stimuli of varied brightness and recording the responses of the patient. The determination of statistical visual field indices is based on measured values. The two most important indices are mean defect (MD) of measured values and loss variance (LV). LV increases with the heterogeneity of DLS in the visual field and is therefore a good indicator of well defined visual field depressions (scotomas). Software controlling the perimeter measures the DLS threshold by employing different test strategies. The normal strategy is routinely used and represents a standard. It changes the brightness of presented stimuli in decreasing steps of 4, 2 and 1 dB, until the threshold is determined with a 1 dB accuracy. Using the normal strategy the examination takes 15 minutes on average and is therefore demanding and fatiguing for the patient. The duration was shortened by development of fast perimetric strategies (dynamic strategy and TOP strategy). The dynamic strategy cuts the examination time in half. It is based on the expansion of threshold zone appearing in the area of lowered DLS and for this reason it changes the brightness of the stimulus in this areas in bigger steps (up to 10 dB). TOP strategy shortens the examination by 80%. Measured value of a certain point is used by the TOP strategy also for the determination of threshold of the neighboring points regarding the in-between distances.
AIMS AND HYPOTHESES. By analyzing the results of normal, dynamic and TOP strategy measured on the same examinees, we wanted to study the differences between the strategies in determination of MD, LV, the differences in determination of number of all points with a deficit and number of points with a significant deficit of p<0,005. We also studied the differences in sensitivity and specificity applying general criteria for discrimination between normal and pathological test results. Three hypotheses were tested.
1. There are no significant differences in determination of MD values when using normal, dynamic or TOP strategy. LV values measured by TOP strategy are significantly lower than LV values measured by the other two strategies.
2. Regarding the number of all points with a deficit measured by the TOP strategy the visual field defects are significantly wider compared to the dynamic or normal strategy. Regarding the number of points with a significant deficit of p<0,005 the most depressed areas in visual field defects are shown significantly narrower with the TOP strategy compared to dynamic or normal strategy.
3. There are no significant differences in sensitivity and specificity between TOP, dynamic and normal strategy when applying general criteria for discrimination between normal and pathological visual fields.
METHODS. 22 normal visual fields of 17 subjects (mean age 33±15 years) and 22 visual fields with defects of 17 patients (mean age 47±16 years) having different types and degrees of visual lesions (glaucoma, inflammatory retinal and choroidal disease, vascular lesions, optic nerve lesion, retinal detachment) were examined. Ethical guidelines for researches on human subjects were followed and official agreement was provided by the Board of medical ethics at the Health ministry of Republic of Slovenia (document No. 68/06/99). All subjects concluded at least one perimetric examination before participating in our research. All visual fields were examined using all three strategies on Octopus 101 perimeter using the standard conditions for this apparatus. The sequence of the strategies was equally alternated for all three strategies. Breaks between examinations were at least 25 minutes long. All three measurements were conducted in no more than 24 hours. Results were statistically evaluated by the Friedman nonparametric technique for comparison of multiple related samples. The differences between individual pairs of strategies were statistically evaluated by Wilcoxon rank-sum technique if the Friedman test result was significant or borderline significant. Three general criteria for discrimination between normal and pathological test results were used for determination of sensitivity and specificity (MD>2 dB, LV>6 dB2, presence of at least 7 points with a deficit and at least three of them forming a group). Sensitivity and specificity were evaluated with the chi-square test for matched binary data.
RESULTS. There were no statistically significant differences in measured MD values between the strategies for both groups of subjects. Maximum values of differences in MD were similar to the values of differences expected on the basis of normal variability of MD. Abnormal visual fields measured by TOP strategy showed 11±14 dB2 lower LV values compared to dynamic strategy (p<0,01) and 9,8±16 dB2 lower LV values compared to normal strategy (p<0,061). Evaluation of average difference in LV between TOP and normal strategy in population of abnormal visual fields was -17 dB2 <µ< -2,7 dB2 (p<0,05). In the group of normal visual fields were the differences in LV between strategies small and statistically insignificant. In the abnormal visual fields group the dynamic strategy measured fewer points with the deficit compared to the other two strategies. The average difference was 3 points (p<0,05). There were no significant differences between strategies in the number of points with a deficit of p<0,005 and also no statistical differences in sensitivity and specificity for all used criteria.
CONCLUSIONS.
1. There are no significant differences between the TOP, dynamic and normal strategy in measuring the MD values. The values of LV measured by the TOP strategy are significantly lower compared to other two strategies.
2. Regarding the number of all points with a deficit the visual field depressions measured by the dynamic strategy are smaller compared to the TOP and normal strategy, which do not differ significantly in this regard. Regarding the number of points with a significant deficit of p<0,005 all three strategies show most depressed areas of visual fields equally well.
3. There are no significant differences in sensitivity and specificity between the TOP, dynamic and normal strategy when applying general criteria for differentiation between the normal and pathological visual fields.
4. With exception of the lower LV values measured by TOP strategy, the differences between TOP, dynamic and normal strategy are small and the application of TOP and dynamic strategy is therefore clinically justified, regarding the benefits of shortening the examination time.
|
|
 |
Išči po nalogah
Brskaj po nalogah
Prijava na obvestila
|
