HCMA AI is a healthcare-AI company empowering digital medicine with responsible AI and advanced technology. MAIND PATH is our research platform for the computational analysis of whole-slide pathology images.
Whole-slide images are digitised pathology slides. A single slide can reach several gigabytes, and a research study can involve thousands of them. MAIND PATH gives our research team one place to manage that workload: it catalogues each slide, checks whether it is technically fit for analysis, splits it into image tiles, extracts numerical features with deep-learning models, and trains and evaluates classification models on the result.
The platform is operated by HCMA AI for its own research and development. It is used by our internal research team, and it is not a consumer product and not open to public sign-up.
Every slide is registered with its source, organ and disease group, then checked against a documented eligibility standard before it can enter a study.
Slides are divided into tissue tiles and passed through foundation models to produce compact numerical representations for downstream analysis.
Multiple-instance-learning models are trained on those representations and reported with the metrics needed to judge whether a result is trustworthy.
Our current studies are built on publicly available, de-identified research datasets, including TCGA (The Cancer Genome Atlas), GTEx (Genotype-Tissue Expression) and BRACS (BReAst Carcinoma Subtyping). These datasets are released for research use and carry no direct patient identifiers.
Where we work with material contributed by a partner institution, that material is de-identified by the contributing institution before it reaches us, and it is handled under the written agreement with that institution.
MAIND PATH uses Google Drive as the storage backend for its own research files. The platform is connected to a single Google account that belongs to HCMA AI, and it reads and writes only inside the folders that account owns.
The platform never asks you to connect your Google account. There is no "Sign in with Google" anywhere in MAIND PATH, and the platform does not read, request or receive Google account data belonging to visitors, partners or any third party.
Our Privacy Policy explains in full which Google API scope the platform requests, why it is needed, what is stored, how long it is kept, and how access can be withdrawn.
MAIND PATH is a research and development platform. It is not a medical device, it has not been reviewed or cleared by any medical-device regulator, and its output must not be used to diagnose, treat, or make any clinical decision about a patient. Every result it produces requires interpretation and confirmation by a qualified pathologist working from the original material.
Questions about this platform, our data handling, or these pages can be sent to histopathology.2026.3.5@gmail.com.