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

Citation
H. Poptani et al., Diagnostic assessment of brain tumours and non-neoplastic brain disorders in vivo using proton nuclear magnetic resonance spectroscopy and artificialneural networks, J CANC RES, 125(6), 1999, pp. 343-349
Citations number
31
Categorie Soggetti
Onconogenesis & Cancer Research
Journal title
JOURNAL OF CANCER RESEARCH AND CLINICAL ONCOLOGY
ISSN journal
01715216 → ACNP
Volume
125
Issue
6
Year of publication
1999
Pages
343 - 349
Database
ISI
SICI code
0171-5216(199906)125:6<343:DAOBTA>2.0.ZU;2-O
Abstract
Purpose: Experiments were carried out to assess the potential of artificial neural network (ANN) analysis in the differential diagnosis of brain tumou rs (low- and high-grade gliomas) from non-neoplastic focal brain lesions (t uberculomas and abscesses), using proton magnetic resonance spectroscopy (H -1 MRS) as input data. Methods: Single-voxel stimulated echo acquisition mo de (STEAM) (echo time of 20 ms) spectra were acquired from 138 subjects inc luding 15 with low-grade gliomas, 47 with high-grade gliomas, 18 with tuber culomas, 18 with abscesses and 40 healthy controls. Two neural networks wer e constructed using the spectral points from 0.6 to 3.4 parts per million. In the first network construction, the ANN had to differentiate between tum ours from infections, while the second network had to differentiate between all five histological classes. Results: ANN analysis gave a histologically correct diagnosis for low- and high-grade gliomas with an accuracy of 73% and 98% respectively. None of the 62 tumours was diagnosed as an infectious lesion. Among the non-neoplastic lesions, ANN classification was correct i n 89% of tuberculomas and in 83% of brain abscesses. The specificity of ANN diagnosis was 98%, 92%, 99%, and 100% for low-grade gliomas, high-grade gl iomas, tuberculomas and abscesses respectively. Conclusion: The present dat a show the clinical utility of noninvasive 1H MRS by automated ANN analysis in the diagnosis of tumour and non-tumour cerebral disorders.