[Prognostic analysis of the hypertensive putaminal hemorrhage in relation to the hematoma volume estimated by CT scanning (author's transl)].
Motoyuki Kaku,Y. Matsukado
DOI: https://doi.org/10.2176/NMC.20.1115
1980-11-01
Neurologia medico-chirurgica
Abstract:In conventional neurological grading and CT classification of hypertensive cerebral hemorrhages, the size or volume of the hematoma was often regarded as one of the prognostic factors of surgical treatment, although estimation of the hematoma volume during the operative procedure showed considerable inaccuracy. To assess the reliability of measuring hematoma volume from CT scans, a phantom study was performed. The Hounsfield Unit (H.U.) of the experimental hematoma was evaluated with respect to the partial volume phenomenon, the peripheral space phenomenon and the shadow effect of CT imaging. A plaster box containing a mixture of agar, milk and water, which was found to reproduce reasonably well the corresponding H.U. of the human head, was prepared as the experimental model, and whole blood or diluted blood was injected into the box. CT scanning was performed with EMI 1010 (120 Kv, 33 mA, 10 mm slice) and the volume of the injected blood was estimated by measuring the number of pixels, which were 10 H.U. over the H.U. of the phantom brain on the print out data. The estimation was easily programmed in a small-size additional computer and calculation of total pixels could be performed simultaneously. Deviation of the estimated hematoma volume from the injected blood volume was less than 4.9%. Forty cases of acute hypertensive putaminal hemorrhage with no evidence of intraventricular hemorrhage were examined and the volume of the hematoma was calculated. Almost all of the CT scans were performed within 3 days after the hemorrhage. The H.U. of the putaminal region of the contralateral side was used as a standard value, which was found to be 35 H.U. on the average and the extent of the hematoma was thus determined from the area of 45 H.U. or more. The hematoma volume estimated by CT was significantly less than those in previous reports which were estimated by surgical evacuation. However, it showed remarkable correlation with the neurological grading of Kanaya. The average amount of the hematoma was 10 ml in grade 1, 20 ml in grade 2, 40 ml in grade 3, 60 ml in grade 4a, and 90 ml in grade 4b and grade 5. Furthermore, there was a significant relationship between ADL scores in the follow-up study and the estimated hematoma volume. In cases with a Kanaya CT classification of III a or IV a or cases with the high density area extending into the corona radiata, the ADL score did not reach one or two in the follow-up study, unless the hematoma was less than 30 ml. When the hematoma volume was estimated to be more than 65 ml, there were no cases of survival.