MECOS combines automate the analysis of complex morphological microscopic objects of biomaterials. Harvesters automatically microscopy specimens, form and analyze digital images of specimens (DIS) in accordance with the requirements of known regulated manual microscopy techniques.
Automatic techniques form the results of the class "preclassification" with the provision in DIS of complete information about the specimen for control and possible adjustment of automatic results by the user.
The AMSM© automatic adaptation technology used in combines maximizes the quality and speed of production of DISs. Analyzes of the DISs in accordance with the tasks of the techniques are performed by automatic analyzers of combines - trained neural networks.
The use of MECOS combines significantly increases labor productivity and accuracy of analyzes.
MECOS harvesters offer users new opportunities for cooperation with colleagues of the professional community and with the artificial intelligence system of the manufacturer of combines.
The user of the harvester can export the DISs and the analysis results from the laboratory database to the harvester page in the collective cloud resource http://mecosvirt.ru. The mecosvirt.ru resource allows to receive both individual telemedicine consultations of colleagues and access to collective community events for collective servicing of the consultation flow, professional development, quality control, creation of archives of diagnostic cases, conferences.
The class of preclassification of automatic results of MECOS combines, as well as other modern automatic analyzers of complex types of specimens, is determined by the fact that at the development stage it is impossible to perform comprehensive training of analyzers due to the huge variety of biomaterials and objects of analysis. Incomplete training determines the possibility of using such analyzers only as an auxiliary tool with mandatory control of the results by the user.
For the first time, MECOS harvester technologies allow continuing the training of automatic analyzers during operation. Users, performing regular control of preclassification results, can send DISs and verified analysis results to the MecosMonitor manufacturer's cloud resource for inclusion in the training base of automatic analyzers of techniques. The artificial intelligence system LOSCAM© from MecosMonitor periodically modernizes the automatic analyzers of the operated combines, performing further training of neural networks using an ever-increasing training sample created by a team of users in the MecosMonitor database. LOSCAM creates conditions for regular "professional development" of automatic techniques in order to achieve the Class of 1 opinion harvester without the control of the user.