Prediction of posterior fossa tumor type in children by means of magnetic resonance image properties, spectroscopy, and neural networks

Abstract

Recent studies have explored characteristics of brain tumors by means of magnetic resonance spectroscopy (MRS) to increase diagnostic accuracy and improve understanding of tumor biology. In this study, a computer-based neural network was developed to combine MRS data (ratios of N-acetyl-aspartate, choline, and creatine) with 10 characteristics of tumor tissue obtained from magnetic resonance (MR) studies, as well as tumor size and the patient's age and sex, in hopes of further improving diagnostic accuracy.Data were obtained in 33 children presenting with posterior fossa tumors. The cases were analyzed by a neuroradiologist, who then predicted the tumor type from among three categories (primitive neuroectodermal tumor, astrocytoma, or ependymoma/other) based only on the data obtained via MR imaging. These predictions were compared with those made by neural networks that had analyzed different combinations of the data. The neuroradiologist correctly predicted the tumor type in 73% of the cases, whereas four neural networks using different datasets as inputs were 58 to 95% correct. The neural network that used only the three spectroscopy ratios had the least predictive ability. With the addition of data including MR imaging characteristics, age, sex, and tumor size, the network's accuracy improved to 72%, consistent with the predictions of the neuroradiologist who was using the same information. Use of only the analog data (leaving out information obtained from MR imaging), resulted in 88% accuracy. A network that used all of the data was able to identify 95% of the tumors correctly. It is concluded that a neural network provided with imaging data, spectroscopic data, and a limited amount of clinical information can predict pediatric posterior fossa tumor type with remarkable accuracy.

1H-MRS is useful to reinforce the suspicion of primary central nervous system lymphoma prior to surgery

“…These results (between 71 % and 89 %) are in the range reported in the literature for classifying tumours by spectroscopy [22][23][24]. There were several limitations in our study.…”

Diagnostic assessment of brain tumours and non-neoplastic brain disorders in vivo using proton nuclear magnetic resonance spectroscopy and artificial neural networks

“…Secondly, cells cultured from human brain tumours express characteristic metabolite phenotypes (Florian et al 1996) and, thirdly, unphysiological metabolites such as serine, acetate and free amino acids, lipid and other macromolecules are present in tuberculomas and abscesses Remy et al 1995;Gupta et al 1996). Proton MRS, however, has not been very successful in distinguishing between dierent tumour types, when conventional spectral analysis methods, such as statistical analysis of metabolite ratios (Kugel et al 1992;Arle et al 1997) or principal-component analysis (Hagberg et al 1995) are used. Given the heterogeneity of tumours and their growth pattern, this is not surprising, as these methods usually have an operator- Table 3 Percentage sensitivity, speci®city, positive predictive value (PPV) and negative predictive value (NPV) in tasks one and two.…”

“…All network predictions for a given training set were obtained by cross-validation analysis. This method is considered to be an eective technique for network validation (Astion et al 1993;Arle et al 1997). Over-training of the ANN was avoided by stopping the training immediately when learning started to slow down rapidly.…”

Grundlagen der 1 H-MR-Spektroskopie intrakranieller Tumoren

“…In der Spektroskopie befindet sich die untersuchte Probe (bei der In-vivo-MRS das ausgewählte Volumen) in einem homogenen Magnetfeld ohne von außen angelegte Gradienten. Das Energiespektrum bildet dann intramolekulare Feldän-derungen ab, die durch Wechselwirkungen des Kernspins mit der Elektronenhülle und des benachbarten Lanfermann H, et al 1 …”

“…und in vivo[16, 23, 25-27, 51, 78, 80] analysiert, relative und absolute Konzentrationen[38,82,83] ermittelt und letztlich die erhobenen Daten mit der Histologie korreliert[1,57,61,69,81]. Die Ergebnisse waren heterogen und stimmten z.T.…”