Automatically determining a brock score

Inventors

Warier, Prashant • Modi, Ankit • Putha, Preetham • Vanapalli, Prakash • Challa, Vikash • DEVI, Ranjana • JAIN, Ritvik

Assignees

Qure AI Technologies Pvt Ltd

Abstract

Disclosed is a system and a method for determining a brock score. A CT scan image may be resampled into a plurality of slices using a bilinear interpolation. A nodule may be detected on one or more of the plurality of slices. A region of interest associated with the nodule may be identified using an image processing technique. Further, a nodule segmentation may be performed to remove an area surrounding the region of interest. Subsequently, a plurality of characteristics associated with the nodule may be identified automatically using a deep learning model. Finally, a brock score for the patient may be determined based on the plurality of characteristics and demographic data of the patient.

Core Innovation

A method to determine a Brock score for a patient uses a chest-region Computed Tomography (CT) scan image and processes the image in real-time. The method receives the CT scan image, resamples the CT scan image into a plurality of slices, and detects a nodule on one or more of the plurality of slices based on an analysis of the plurality of slices using a first deep learning.

The method identifies a region of interest associated with the nodule using an image processing technique and performs nodule segmentation by masking the region of interest. The masking removes an area including at least one of black areas and fatty tissues surrounding the region of interest, where the nodule segmentation is performed using the first deep learning model.

Next, the method automatically determines a plurality of characteristics associated with the region of interest using a second deep learning model trained using historical characteristics. The plurality of characteristics comprises a location of the nodule, a texture of the nodule, a size of the nodule, a speculation of the nodule, and number of nodules present in the CT scan image.

Finally, the method determines the Brock score for the patient automatically in real-time based on the plurality of characteristics and demographics data of the patient. The demographics data includes patient age and patient gender, and the method can further include patient name and the patient’s medical family history, with an optional comparison to a Brock threshold in dependent implementations.

Claims Coverage

The independent claims (three) cover a real-time pipeline for determining a Brock score using chest-region CT processing, nodule detection and segmentation with a first deep learning model, extraction of nodule characteristics with a second deep learning model, and automatic real-time scoring based on the extracted characteristics and patient demographics data. Across the independent claims, the main inventive features repeat: CT resampling into slices using bilinear interpolation, first deep learning nodule detection, ROI identification and masking-based segmentation removing black areas and fatty tissues, second deep learning extraction of location/texture/size/speculation/number, and Brock score determination based on characteristics and demographics.

The independent claim set consistently provides real-time Brock score determination from chest-region CT by combining bilinear interpolation slice resampling, first deep learning nodule detection, ROI identification with masking-based segmentation removing black areas and fatty tissues, second deep learning extraction of nodule location/texture/size/speculation/number, and a final scoring step based on those characteristics together with patient demographics data.

Stated Advantages

Documented Applications