The clinical role of extracellular bioimpedance tomography (Gastro-Mida(x)) in the diagnosis of colorectal diseases

Minerva Med. 2004 Dec;95(6):541-56.
[Article in English, Italian]

Abstract

Aim: Gastro-Mida(x) is a non-invasive diagnostic medical device which carries out an extracellular bioimpedance tomographic analysis of the colorectal region with a neural technique interpretation using artificial neural networks to diagnose colorectal diseases. The aim of this spontaneous study carried out in various centres in Italy was to train Gastro-Mida(x) to distinguish normal patients from those with colorectal diseases.

Methods: Patients who were referred to the Endoscopy Units of the 20 centres involved in the study to undergo colonoscopy from September 2002 to December 2003 were included in the study. At least 1 day before colonoscopy, patients underwent Gastro-Mida(x). The training of the bioimpedance tomography in this study considered patients negative (normal) when the colorectal endoscopy was normal or revealed only the presence of haemorrhoids. Patients were considered positive in the presence of diverticula (diverticulosis or diverticulitis), polyps, cancer. For positive patients with more than one pathology, neoplastic or preneoplastic lesions were considered more important than the others and the final diagnosis was made in accordance with this rule.

Results: Overall, the 20 centres enrolled 947 patients (males/females: 477/470; average age+/-SD: 57.44+/-13.85 years). The specificity registered after training was 80% (95% CI: 76.5-83.1) and sensitivity was 83.89% (95% CI: 79.5-87.4). The sensitivity of bioimpedance tomography in the subgroup of patients with cancer (N=68) was 88.2% (95% CI: 78.4-93.9). As to polyps, sensitivity was calculated as a whole and according to size classification. The sensitivity of bioimpedance tomography in diagnosing patients with polyps was therefore 78.3% (95% CI: 71.9-83.4). However, sensitivity increased to 83.7% (95% CI: 74.1-90.2) and 83% (95% CI: 70.7-90.8) when the bioimpedance tomography's ability to identify patients with at least one polyp > or = 6 mm and > or = 10 mm in size was taken into consideration.

Conclusions: Gastro-Mida(x) has proven to be a simple, reliable and accurate instrument, once training is completed, in the most common colorectal diseases. The device can therefore be proposed for both diagnosis and screening of colorectal diseases.

Publication types

  • Evaluation Study
  • Multicenter Study

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Colonic Diseases / diagnosis*
  • Colonic Neoplasms / diagnosis
  • Electric Impedance*
  • Electrodes
  • Female
  • Humans
  • Intestinal Polyps / diagnosis
  • Italy
  • Male
  • Middle Aged
  • Neural Networks, Computer*
  • Rectal Diseases / diagnosis*
  • Rectal Neoplasms / diagnosis
  • Sensitivity and Specificity
  • Tomography / instrumentation
  • Tomography / methods*