OUR VISION
“A health system in which pathological anatomy is increasingly digitized and integrated.”
“A health system in which pathological anatomy is increasingly digitized and integrated.”
OUR MISSION
“Improve the quality, accuracy and timing of diagnoses through the use of digital pathology, technology solutions, and AI.”
“Improve the quality, accuracy and timing of diagnoses through the use of digital pathology, technology solutions, and AI.”
CLOUD
PATHOLOGY
GROUP (CPG)
our history
Cloud Pathology Group (CPG) was born in Milan in 2012 with the aim of improving the quality of diagnoses using the practice of Digital Pathology for the operational uses of public and private healthcare facilities. Through the proposal of integrated services that make use of own and/or third-party hardware, software and technologies. CPG offers solutions that benefit the work of technicians and doctors in their daily lives.
In September 2015, CPG presented its development project "A New Pathology" within the "Horizon 2020 - Research and Innovation Framework Programme" of the European Commission, obtaining the Seal of Excellence Certificate.
In 2018, CPG started operations with the New Ospedale degli Infermi in Biella, supporting the department in the digitalization project of the clinical-histological routine.
In 2019 CPG was awarded the DiOncoGen Innovative Diagnostics project, with European funds, through the Sicily Region, for the development of a diagnosis support application with AI algorithms, for colorectal cancer.
In 2020 CPG created the first Italian R&D hub of AI applied to Pathological Anatomy in Catania with the collaboration of the ISTI-CNR of Pisa.
In 2020 CPG was awarded the tender for the construction and management of a Network of Digital Pathology Laboratories centered on the Insubria area, which includes the hospitals of Busto Arsizio, Varese and Como.
In 2022, CPG begins to make the developed AI diagnostic support solutions available to its customers.
FROM TRADITIONAL
TO DIGITAL
For structures equipped with a pathological anatomy laboratory, CPG offers an integrated, open and flexible solution that allows to optimize production processes very quickly, without risks and difficulties, making available the use of the most advanced digital technologies in SaaS mode (Software as a Service).
RESEARCH AND
DEVELOPMENT
CPG has here developed an artificial intelligence model that exploits image analysis. It is a computer vision model that uses a neural network to classify areas of interest within a histological slide.
Knowledge
The acquisition of clinical knowledge for neural network training using deep learning methods was delegated by qualified medical personnel familiar with the pathological anatomy environment. The knowledge consists of manual annotations made on the digitized slides. This process allows information to be stored in a database and cataloged to facilitate learning. The importance of clinical staff is fundamental for the realization of the computer vision model, as learning the operational context for artificial intelligence is the most critical process and which makes the difference on the application in the routine.
Technique
The design of the artificial intelligence model was carried out using programming frameworks chosen among the most used by the big companies in the tech world.
The computer vision model is thus both reliable from a mathematical point of view and scalable on the various clinical fields for which the annotation system is useful.
Clinic area
The clinical area initially chosen is that of the colon and rectum, of interest thanks to the high diffusion of pathologies that concern them.
Achieved results
The results obtained by the artificial intelligence model are promising. The classification of areas of interest is precise and accurate with both statistical values above 90%. The differentiation of classes to which they belong is limited to the main features found within a histological slide, such as adenocarcinomas, dysplasias, lymphocyte aggregates.
Future evolutions
However, new areas of histological and cytological interest are candidates to follow it, taking advantage of the work done and the model created previously.
Operational flexibility
Our solution can be integrated into any clinical operating environment.
The results obtained from the model output can be viewed as areas of interest on a digital histology slide viewer.
Scarica “Soluzione innovativa nella diagnosi assistita dei tumori al colon retto”
Scarica “Soluzione innovativa nella diagnosi
assistita dei tumori al colon retto”
LINK VELOCI
DOVE SIAMO
Cloud Pathology Group srl
Via Lanzone, 4, 20123 Milano
c.f./p.iva 07753790968
