Schematization of the integration framework. The anatomic pathology laboratory information system (AP-LIS) and the external artificial intelligence (AI)-based decision support system (AI-DSS) containing a set of pre-trained deep-learning (DL) models communicate with each other via Health Level 7 (HL7) messaging. The communication occurs through the following sequence of events: (1) a laboratory order message (OML^O33) is sent from the AP-LIS to the AI-DSS; (2) an acknowledgment (ACK) message is sent from the AI-DSS to the AP-LIS upon reception of the OML^O33 message; (3) the AI-DSS processes the OML^O33 message retrieving information on the DL model name and the path to the whole-slide image (WSI) to analyze (bold orange); (4) DL model inference is run; (5) a QuPath project is built to visualize DL models results; (6) a laboratory observation message (OUL^R21) is sent from the AI-DSS to the AP-LIS. Here, DL model results are transmitted to the AP-LIS as OBX segments (bold green); (7) an ACK message is sent from the AP-LIS to the AI-DSS upon reception of the OUL^R21 message
