Method and apparatus for separating cell objects for analysis

• US 5,218,645 A
• Filed: 03/29/1991
• Issued: 06/08/1993
• Est. Priority Date: 03/29/1991
• Status: Expired due to Term

• Current Assignee: Cell Analysis Systems, Inc. (Becton, Dickinson & Co)
• Original Assignee: Cell Analysis Systems, Inc. (Becton, Dickinson & Co)
• Inventors: Bacus, James V.
• Primary Examiner(s): COUSO, JOSE L
• Application Number:
  US07/677,213
• Time in Patent Office: 802 Days
• Field of Search: 382/6, 382/48, 364/413.08, 364/413.1, 358/107, 358/93, 356/39
• US Class Current: 382/133
• CPC Class Codes: G01N 15/1433 using image recognition
• CPC Class Codes: G06V 20/69 Microscopic objects, e.g. biological cells or cellular parts

Abstract

A method and apparatus for analyzing the cell objects of a cell sample for the diagnosis and treatment of possible cancer is disclosed. An optical image of the cell sample is digitized and stored in a frame buffer. The digital image is then displayed on a video monitor. The apparatus can automatically select cell objects from the sample for analysis. Also, an operator can influence the selection of cell objects by manipulating a computer pointing device (mouse) to draw lines on the displayed image. When a line is drawn through a cell object, the cell object can be separated into portions for analysis. When a line is drawn around a region of interest in the image, the region of interest can be separately analyzed to the exclusion of the rest of the image.

Claims

1. In a cell analysis apparatus, a method of analyzing, with a predetermined analysis routine, an image of a cell sample including at least one cell object surrounded by image background, said method comprising:

creating a search mask of said cell sample image by storing image representing picture element values in a plurality of storage locations;

displaying a representation of said cell sample to an operator;

assigning a unique boundary value to selected ones of said storage locations of said picture element values stored in said search mask at storage locations selected by an operator in a line from said image background through said at least one cell object and returning to said image background;

identifying at least one boundary of a region of said cell sample image by a search routine in which said unique value is recognized as being a region boundary; and

analyzing by said analysis routine the regions delimited by said at least one boundary identified in said identifying step.

2. A method in accordance with claim 1 wherein said displayed representation comprises a plurality of picture elements each corresponding to one of said search mask storage locations and said assigning step comprises:

drawing at least one line on said displayed representation; and

assigning said unique value to the ones of said search mask storage locations corresponding to ones of said picture elements making up said at least one line.

3. A method in accordance with claim 1 wherein said search mask creating step comprises storing a gray level picture element representation of said cell sample image and said displaying step comprises displaying said search mask on a video monitor.

4. In a cell analysis apparatus comprising a storage array having a plurality of storage locations for storing values, a method of analyzing a cell sample including at least one cell object on a background, said method comprising:

producing an optical image of said cell sample;

separating said optical image into a plurality of picture elements each representing a portion of said optical image and storing in each of said storage locations a gray level value representing one picture element of said optical image;

storing a predetermined background value in each of said storage locations corresponding to a background picture element;

displaying on a video monitor a video image generated from said values stored in said storage locations;

storing a unique boundary value in ones of said storage locations selected by an operator of said apparatus, said storing boundary value step comprising, selecting by means of a moveable cursor a line starting point in a first background portion of said displayed video image, moving said cursor to trace a line across a cell object portion of said displayed video image, selecting an ending point for said line in a background portion of said displayed video image;

identifying from the values stored by said storage locations at least one boundary derived from said stored predetermined background values and said stored unique boundary values; and

analyzing a portion of said cell sample within said at least one boundary.

5. A method in accordance with claim 4 wherein said predetermined background value storing step comprises:

comparing the value of each storage location to a predetermined cell object threshold value to identify storage locations corresponding to picture elements representing optical image background; and

changing the value stored by each storage location identified in said comparing step to the predetermined background value.

6. A method in accordance with claim 4 comprising storing said unique boundary value in ones of said storage locations corresponding to picture elements traversed by said line.

7. In a cell analysis apparatus comprising a storage array having a plurality of storage locations for storing values, a method of analyzing a cell sample including at least one cell object on a background, said method comprising:

producing an optical image of said cell sample;

separating said optical image into a plurality of picture elements each representing a portion of said optical image and storing in each of said storage locations a gray level value representing one picture element of said optical image;

storing a predetermined background value in each of said storage locations corresponding to a background picture element;

displaying on a video monitor a video image generated from said values stored in said storage locations;

storing a unique boundary value in ones of said storage locations corresponding to the picture elements of a line drawn at the direction of an operator on said video image, said line extending through an image of said cell object from background to background;

performing a raster search of said storage locations to identify a first storage location storing a gray level value corresponding to a cell object picture element;

performing a neighbor search from said first storage location to identify a boundary delimiting a portion of said cell object, said boundary being defined by storage locations storing said unique boundary value and by storage locations storing said predetermined background value; and

analyzing said portion of said cell sample within said boundary.

8. In an automated analysis apparatus a method of analyzing a cell sample comprising at least one cell object, said method comprising:

displaying on a video monitor a video image of said cell sample;

drawing a line through said at least one cell object in said video image by an operator;

selecting by said apparatus a cell object portion of said cell object for analysis, at least one portion of a boundary of said selected cell object portion being defined by said line; and

analyzing said selected cell object portion by said apparatus in accordance with a predetermined analysis routine.

9. A method in accordance with claim 8 wherein said drawing step comprises:

manipulating a computer pointing device to move an image cursor across said cell object; and

drawing on said monitor a line along a path defined by the image cursor.

10. A method in accordance with claim 8 comprising:

separating an optical image of said cell sample into a plurality of picture elements each representing a portion of said optical image and storing in each of a plurality of storage locations a gray level value representing one picture element of said optical image;

storing, responsive to said drawing step a unique boundary signifying value in ones of said storage locations defined by the position of said line on said at least one cell object; and

said selecting step comprises identifying ones of said storage locations storing said unique boundary signifying value.

11. Apparatus for analyzing a cell sample comprising at least one cell object, said apparatus comprising:

means for displaying a video image of said cell sample;

means for drawing a line through said at least one cell object on said video image;

means for selecting a cell object portion of said at least one cell object for analysis, at least one boundary of said selected cell object portion being defined by said line drawn on said video image; and

means for analyzing said selected cell object portion in accordance with a predetermined analysis routine.