View Full Version : Tumour classification via ANN's

H.j. Woltring, Fax/tel +31.40.413 744
10-08-1992, 02:22 AM
Dear Biomch-L readers,

The following notice was distributed last month via the AIMED list. The
principal author, David W. Piraino at the Cleveland Clinic Foundation will
be out of email touch for the next three weeks. Fortunately, one of his
colleagues at the CCF is a subscriber to Biomch-L and has kindly consented
to act as a go-between for the time being.

Persons interested in the notice below may get in touch with Brian L. Davis
Ph.D., Dept. of Biomedical Engineering (Wb3), Cleveland Clinic Foundation
at . Alternatively, Dr Piraino can be contacted at:
Tel. +1(216)444-4845, Fax +1(216)445-9445.



Date: Fri, 18 Sep 92 11:32:50 EDT
From: humphrey@nlm.nih.gov (Susanne M Humphrey)
Sender: AI Medicine Digest [Volume 4, nr. 9]
To: ai-medicine@medisg.Stanford.EDU

The following file in archives:


contains the citation:

UI - 92126706
EM - 9205
AU - Piraino DW ; Amartur SC ; Richmond BJ ; Schils JP ; Thome JM ;
Belhobek GH ; Schlucter MD
TI - Application of an artificial neural network in radiographic
AB - The description of 44 cases of bone tumors was used by an
artificial neural network to rank the likelihood of 55 possible
pathologic diagnoses. The performance of the artificial neural
network was compared with the performance of experienced (3 or
more years of radiology training) residents and inexperienced
(less than 1 year of radiology training) residents. The
artificial neural network was trained using descriptions of 110
radiographs of bone tumors with known diagnoses. The descriptions
of a separate set of 44 cases were used to test the neural
network. The neural network ranked 55 possible pathologic
diagnoses on a scale from 1 to 55. Experienced and inexperienced
residents also ranked the possible diagnoses in the same 44
cases. Inexperienced residents had a significantly lower mean
proportion of diagnoses ranked first or second than did the
neural network. Experienced residents had a significantly higher
proportion of correct diagnoses ranked first than did the
neural network. Experienced residents had a significantly higher
proportion of correct diagnoses ranked first than did the
network. Otherwise, a significant difference between the
performance of the network and experienced or inexperienced
residents was not identified. These results demonstrate that
artificial neural networks can be trained to classify bone
tumors. Whether neural network performance in classification of
bone tumors can be made accurate enough to assist radiologists in
clinical practice remains an open question. These preliminary
results indicate that further investigation of this technology
for interpretation assistance is warranted.
MH - Bone Neoplasms/RADIOGRAPHY ; Human ; *Neural Networks (Computer)
; *Radiographic Image Interpretation, Computer-Assisted
AD - Department of Radiology, Cleveland Clinic Foundation, OH
SO - J Digit Imaging 1991 Nov;4(4):226-32

I do not know if this article publishes the actual algorithm, but the
author address is given.