Non-invasive and quantitative method for making the diagnosis of congenital heart disease having shunt Analysis of dye dilution curve recorded by earpiece, following injection into the peripheral vein

• Published in: Journal of Nippon Medical School Vol. 49; no. 3; pp. 403 - 420
• Main Author: Kawagoe, Shigeru
• Format: Journal Article
• Language: Japanese
• Published: Japan The Medical Association of Nippon Medical School 1982
• Subjects: Adolescent; Adult; Child; congenital heart disease with shunt; Diagnosis, Differential; dye dilution curve; Dye Dilution Technique; Heart Defects, Congenital - diagnosis; Humans; Middle Aged; non-invasive procedure; quantitative diagnosis; theory of quantification
• ISSN: 0048-0444; 1884-0108
• DOI: 10.1272/jnms1923.49.403

The author has developed a non-invasive method to make an accurate diagnosis of congenital heart disease having shunt. Namely, Indocyanin green was rapidly injected into the antecubital vein and the dye dilution curve was recorded with an ear piece. The parameters of the dilution curve were then subjected to relevant mathematical analysis, including the study of correlations between the given pairs of parameters and the usefulness of the parameters was assessed. Finally theory of quantification, which was applicable even to non-linear relationship, was employed and a discriminant function was derived. It was this function that was instrumental in making differential diagnosis. The method was tested in 53 cases of atrial septal defect (ASD), 27 cases of ventricular septal defect (VSD), 15 cases of patent ductus arteriosus (PDA) and 26 cases of normal controls without organic heart disease. Diagnosis was confirmed by cardiac catheterization and/or surgery. Using this procedure, correct diagnosis was made in all 23 cases with VSD, 12 cases with PDA and 20 cases with normal controls. Only five out of 42 cases with ASD were erroneously diagnosed. The total rate of correct diagnosis was 94.8%. As to the external samples, 4 cases of VSD, 3 cases of PDA and 6 normal controls were correctly diagnosed, and erroneous diagnosis was made only in 2 out of 11 cases of ASD. The total percentage of correct diagnosis was 91.7%. These results imply that this method is of practical value in clinical medicine.