Morphometrical methods

Ladefoged O, Roswall K, Larsen J-J. 1994. Acetone potentiation and influence on the reversibility of 2,5-hexanedione-induced neurotoxicity studied with behavioral and morphometric methods in rats. Pharmacol Toxicol 74 294-299. 

To circumvent the above limitations, the morphometric method can be used (Riischoff et al., 1990). It consists of automatic or semiquantitative measurement of the area occupied by the silver-stained structures within the nuclear profile with computer-assisted image analysis (Trerb et al., 1995). This method is faster, more accurate and reproducible, and shows less interobserver variation. Moreover, in contrast to the counting technique, the morphometric method is predictive of patient survival, independent of the clinical stage of the disease (Ofner et al., 1995a,b). 

Lee, R.M.K.W., Forrest, J.B., Garfield, R.E. and Daniel, E.E. (1983b) Comparison of blood vessel wall dimensions in normotensive and hypertensive rats by histometric and morphometric methods. Blood Vessels 20 245-254. 

Morphometric methods showed that low temperature (4°C for 7 days) inhibited bleomycin lung toxicity in the rat (Berend 1983). 

All taxa misidentifled by both morphometric methods were also listed among the set of daisy s failed identifications suggesting its results represent a more consistently conservative estimate of identification certainty than either of the morphometric approaches. If the question being asked was whether these two groups were present in the sample, the quality of DAISY S answer would be indistinguishable from those of the much more labor-intensive morphometric methods. Even in terms of estimating relative abundances of these taxa, the DAISY estimates do not differ markedly from the morphometric results (see Table 10.1). These results are not unusual, but rather typical of performance levels that can be expected of neural net approaches to the group... 

For most interspecific smdies this latter feature works to the net s - and the systematist s -advantage. DAISY S performance in the two-group analysis, while less precise than those of the morphometric methods nnder the same sample size constraints, was well within the range of acceptability for routine assessments of both richness and diversity. Given the speed with which DAISY was able to complete its analyses and its fully automated nature, PSOM neural nets have clear and practical advantages over both qualitative and morphometric approaches for most routine group-identification needs. 

---