Imsight Technology

Pathology: Products and services of Imsight involve three advantageous fields of radiology, pathology, and radiotherapy and two emerging fields of ophthalmology and dermatology. Its products are widely available in hospitals, independent image and inspection centers

Cervical-Sight Cervical Liquid-Based Cell AI Screening System: Cervical-Sight uses a deep learning technique based on convolutional neural networks to enable automated analysis of cervical fluid-based cytopathic images. It allows rapid screening, able to locate and classify diseased cells according to TBS guidelines, and detection of microbial infections. Diagnosis is made through intelligence analysis and ultimately generate test reports.

Advantages: (1) Improve the efficiency of digital pathology cloud annotation system; (2) Can handle up to 2000 cases daily; (3) Sensitivity 90%, specificity 85%; (4) Can act as a first line for pathology screening.

Issues: There is a large demand for cervical cancer screening but at the same time there is a critical shortage of pathologists. While convention diagnosis method mainly depends on the pathologist’s experience, the labour intensive task could increase the risk of missed diagnosis and misdiagnosis.

Features: (1) Automatically identify 15 types of abnormalities; (2) Whole slide screening to reduce the rate of missed detection; (3) Issue standard report according to TBS guidelines; (4) Can rapidly locate lesions therefore improving screening efficiency.

Breast-Sight A.I. detection of lymph node metastasis in breast cancer: Detection of lymphatic metastasis based on pathological images is a key indicator of breast cancer staging diagnosis. Breast-Sight can automatically and accurately delineate the metastasis area on pathological section, measure the size of the metastatic lesion and therefore facilitate doctors in locating metastasis area and determining the staging to prevent any missed diagnosis.

Advantages: (1) Diagnosis rate of 99.5% for benign and malignant;
(2) Accuracy rate of 85.33% in locating metastasis region.

Issues: Breast cancer metastasis detection through tissues slide is time consuming. It could be hard to reproduce quantitative result and such result is susceptible to the pathologist’s experience.

Features

(1) Information on benign or malignant;

(2) Locate metastasis regions;

(3) Information on staging.