Breast cancer automatically diagnosed

QinetiQ scientists claim to have developed the world’s first automatic breast cancer diagnostic system – PathScore.

Existing method of diagnosing breast cancer in women requires pathologists to look at tissue samples through a microscope and visually interpret complex microscope images.

This is not only time-consuming but is also subjective and the problem is exacerbated by a worldwide shortage of pathologists.

Currently, diagnosis is achieved from a combination of radiological, surgical and pathological assessment of tissue samples. This requires detailed evaluation of tissue to provide information such as tumour size and type. On the basis of this information, a course of treatment is then planned for a patient. QinetiQ’s PathScore system speeds up the diagnostic process by making an automatic assessment of the condition of the patient.

In addition. the system could be used to create a patient database in which the scoring of prognostic features of breast cancer is comparable from patient to patient. Such a database would be invaluable in clinical trials of both new and existing therapies.

The QinetiQ team has also developed a content sensitive image compression technique to enable transmission of the very large quantity of images over realistic bandwidths in realistic times. This allows pathologists to give a valid second opinion from viewing compressed images on a remote computer.

Until now, use of compressed images for such a purpose has been controversial because commercial image compression algorithms such as JPEG often lose important detail when achieving data compression. QinetiQ’s system, developed in 2000, achieves high compression ratio without loss of diagnostically important information, making telepathology a viable option.

Qinetiq claim that the performance of the system is similar to that of a human pathologist.

On the web