The procedure of sharing the diagnostic medical reports and the scanned images of patients among doctors in a secured fashion with corresponding suggestions for joint treatment creates greatest care for the patients through quicker and crucial decisions. Intrusion Detection is the essential part of Information and Network Security domain in opposition against illegal access or malicious attacks. In this context, a new mechanism has been projected to detect offline Intrusion Detection System over the medical images. The proposed scheme deals with Image Blocks matching for the intrusive and normal pattern detection. From the experimental results, the proposed scheme detects the attacks namely: R2L, U2R and DoS. The observation shows that the proposed model offers high efficient detection and accuracy. This model gives realistic results over the Biomedical Images (Ex. DICOM).