Diagnos Inc.

Diagnos Inc.

June 13, 2008 08:38 ET

DIAGNOS and Queens University to Develop New Test to Increase Reliability of Cancer Diagnosis

BROSSARD, QUEBEC--(Marketwire - June 13, 2008) - DIAGNOS inc. ("DIAGNOS" or the "Corporation") (TSX VENTURE:ADK), a leader in the use of artificial intelligence ("AI") and advanced knowledge extraction techniques, announces that, in collaboration with the Cancer Research Institute, Division of Cancer Biology and Genetics, Queens University, it is developing an automated method of increasing the reliability of cancer diagnosis accuracy. Presently pathologists have difficulty in making a confident pathological diagnosis of lymphoma based on small biopsy samples. Our system will improve the quality of diagnosis, establish a standard for Canadian laboratories and help pathologists improve their interpretive and diagnostic confidence. We have specifically chosen follicular lymphoma as it is one of the more difficult lymphoma subtypes to diagnose accurately and confidently. We are convinced that our system will significantly enhance the productivity, the quality of diagnosis and will dramatically decrease the cost to the health care system specifically all laboratories that test for Canadian hospitals.

The first phase will be in collaboration with the Cancer Research Institute in developing an automated approach to distinguish follicular lymphoma from benign lymphoid infiltrates based on small samples. Our recent trials have demonstrated a high degree of accuracy. This has been achieved by developing new algorithms that have shown the potential of increasing the reliability of cancer diagnosis. Our technical team has developed computational tools for automated cancer diagnosis that operate on quantitative measures. We expect to increase the accuracy of our results by augmenting our morphological analysis with immunohistochemistry.

Presently, lymph node pathologists have difficulty making a confident pathological diagnosis of lymphoma based on small biopsy samples, their preference being in order of at least one centimetre in diameter such as those obtained when an entire lymph node is excised. Recently and with increasing frequency, they are getting tiny cores of tissue that are obtained using cutting biopsy needles. Needle biopsies seem expedient to clinicians and patients but, when dealing with the possibility of lymphoma, such samples present challenges for pathologists due to the increased possibility of error.

In an article published last year in the American Journal of Clinical Pathology, entitled: "The reliability of lymphoma diagnosis in small tissue samples is heavily influenced by lymphoma subtype", follicular lymphoma was one of the more difficult lymphoma subtypes to diagnose accurately and confidently in small samples. The objective of the current project is to develop technology and an automated system to assist pathologists in making lymphoma diagnoses with greater accuracy and using smaller samples in distinguishing follicular lymphoma from non-lymphomatous processes. The algorithm will be adapted subsequently to address the distinction between different lymphoma types and between lymphoma and other cancers.

The second phase will be to apply our technology to the discovery and ascertainment of "biomarkers". Biomarkers are macromolecules, typically proteins or nucleic acids, whose presence is useful in predicting clinical outcomes, such as response to a particular treatment. Immunohistology, also called "immunostaining"or "immunohistochemistry", is commonly used in biomarker detection since it is inexpensive, widely available and readily applicable to conventionally collected and processed tumour biopsy samples. Immunohistology results are typically scored by visual inspection with a microscope. We expect based on present results that our system will do better in scoring the immunostaining results in an objective and more quantitative manner.

For further information, please visit our Website at www.diagnos.com.

The TSX Venture Exchange has not reviewed and does not accept responsibility for the adequacy or accuracy of this release.

Contact Information