公开(公告)号：WO2004090779A3

公开(公告)日：2005-08-11

申请号：PCT/EP2004003613

申请日：2004-04-06

Applicant: ROCHE DIAGNOSTICS GMBH ,  HOFFMANN LA ROCHE ,  PETRICH WOLFANG ,  WERNER GERHARD ,  HEIN HEINZ-MICHAEL ,  KUHR HANS-JUERGEN

Inventor： PETRICH WOLFANG ,  WERNER GERHARD ,  HEIN HEINZ-MICHAEL ,  KUHR HANS-JUERGEN

IPC: G06F19/00

CPC classification number: G16H10/40 ,  G16H50/50

Abstract: The invention relates to analysis methods for diagnosing diseases on human and animal samples. Said invention also relates to an evaluation method for diagnosing the individual stages of a disease in such a way that it is possible to display the progression thereof. Said invention also makes it possible to identify diseases in an early manner and to carry out therapeutic controls. The inventive method consists in carrying out actually known multivariable evaluation methods for classifying samples. Nevertheless, the invention is characterised in that no sample is allocated to a certain class, but it is classified in a data record based on the interpolation between different classes.

Abstract translation: 本发明涉及的是适合于疾病的人类和动物样品中的诊断分析方法的字段。 本发明还公开了通过其使得可以诊断个体疾病阶段，使得疾病的过程可以被成像的评价方法。 这使得疾病和治疗监测的早期检测可以做到的等 本发明的方法使基获利本领域中已知的用于样品的多变量分类的评价方法。 然而，根据本发明的没有样品的一类分配，但在一组从不同类之间的内插得到的数据进行了分类不像在现有技术中发生的。

公开(公告)号：CA2521921A1

公开(公告)日：2004-10-21

申请号：CA2521921

申请日：2004-04-06

Applicant: HOFFMANN LA ROCHE

Inventor： KUHR HANS-JUERGEN ,  WERNER GERHARD ,  PETRICH WOLFANG ,  HEIN HEINZ-MICHAEL

IPC: G06F19/00

Abstract: The invention relates to analysis methods for diagnosing diseases on human a nd animal samples. Said invention also relates to an evaluation method for diagnosing the individual stages of a disease in such a way that it is possible to display the progression thereof. Said invention also makes it possible to identify diseases in an early manner and to carry out therapeuti c controls. The inventive method consists in carrying out actually known multivariable evaluation methods for classifying samples. Nevertheless, the invention is characterised in that no sample is allocated to a certain class , but it is classified in a data record based on the interpolation between different classes.