Project Details
Description
Almost all women in Denmark participate in vaginal cytological screening programmes, that is, frequent analyses by microscope of cells (smear) from the vagina. A major problem is that cyto-technicians may misinterpret or overlook ill cells when looking into the microscope. At present about 650,000 smears are examined annually. There are, however, almost 500 new cases of cervical cancer per year.
The project objective is to develop an automated screening system, as an aid for the cyto-technicians, in order to increase the diagnostic certainty.
The Dept. of Automation focuses on developing a robot that will feed a sample to a microscope, move it under the lens during the image processing phase, and take it away again for storage in a rack. Another task for the department is to develop algorithms that classify cells -- by means of, say, shape, colour, and size -- into the groups: OK and not OK. The latter group should be further divided into mild, moderate, and severe cancer (dysplasia).
So far, three Master's projects have been completed (Voss, 1996; Post & gjerløv, 1998; Byriel, 1999).
The project objective is to develop an automated screening system, as an aid for the cyto-technicians, in order to increase the diagnostic certainty.
The Dept. of Automation focuses on developing a robot that will feed a sample to a microscope, move it under the lens during the image processing phase, and take it away again for storage in a rack. Another task for the department is to develop algorithms that classify cells -- by means of, say, shape, colour, and size -- into the groups: OK and not OK. The latter group should be further divided into mild, moderate, and severe cancer (dysplasia).
So far, three Master's projects have been completed (Voss, 1996; Post & gjerløv, 1998; Byriel, 1999).
| Status | Finished |
|---|---|
| Effective start/end date | 01/02/1997 → 31/12/2000 |